Discussion:
[EE] R4850-G2 CAN Bus without any clue...
Tobias Gogolin
2018-11-12 10:30:40 UTC
Permalink
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as base for
a quick charger for my electric trike. Regrettably the published pdf about
its operation are not very in-depth and I assume it needs some CAN bus
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038


I appreciate any hints of how to proceed. I know the unit is alive and
boots up and even puts a small test voltage out, and the 3 status LED tell
that its alive, but it must want some configuration still. I already got a
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...

Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------

Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman
Richard Prosser
2018-11-12 20:30:09 UTC
Permalink
Hi Tobias,
It looks like a standard Telecom type battery charger. From the datasheet I
don't see anything indicating that it needs comms to startup but you may
need comms to adjust the voltage & current limit to your requirement. It's
probably preset to 53.5V or so with maximum current & power limit etc.

What Leds are showing? - a single green would appear to indicate normal
operation.

What happens with a resistive load (eg 4x12V automotive bulbs in series, a
220V incandecent lamp or a heater element)? With 3kW output capability it
should be able to provide 50A or so current. Note that the output
protection is foldback type so short circuit current will be less (~25A
from the datasheet).

Re the CAN bus, I can't help. We used to use RS485 comms on our rectifiers
(otherwise pretty similar). Probably there's a data sheet somewhere on the
web giving info on what's required for this or a similar product. If you
can find info on the commands for a similar but non-identical Huawei
rectifier it may assist as companies tend to retain the same protocol &
commands etc.. I don't think there's a Telecom standard for this but there
may be something these days. Or the commands may be encrypted in some way.
:-( in which case hardware adjustment may be the only option if you need to
change the voltage.

Once you register on the Huawei website there seems to be a lot of
datasheets available. It just may take a while to find what you need.

RP
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as base for
a quick charger for my electric trike. Regrettably the published pdf about
its operation are not very in-depth and I assume it needs some CAN bus
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive and
boots up and even puts a small test voltage out, and the 3 status LED tell
that its alive, but it must want some configuration still. I already got a
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Tobias Gogolin
2018-11-12 22:31:20 UTC
Permalink
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to begin with
to add its juice to? Similar to a cell phone battery protection circuit
that may need to be woken up by charging its internal battery beyond a
certain threshold if the battery has been sitting for too long, this supply
may not dare to power up what appears to be a dead battery? Could be an
option set within some internal config register though... I could test
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not indicated must
belong to the negative terminal of the output DC because that negative PCB
contact is recessed as if for hot plugging, I doubt it has any significance
in regular operation tough?

The green LED is the one that remains on, but if I probe around and bridge
that ~1 Volt feeler voltage that Voltage goes zero and the red LED comes on.
I have put it in a case with a DPS5020 50V 20A digital power supply to
provide a side function while I develop my own charge pump and controller.
So there would be a capacitive load if the supply was to do a test run just
like that... I thought about trying Huawei directly, so registering is a
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a exciting -
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the banks gets
even close to a save voltage limit...
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the datasheet I
don't see anything indicating that it needs comms to startup but you may
need comms to adjust the voltage & current limit to your requirement. It's
probably preset to 53.5V or so with maximum current & power limit etc.
What Leds are showing? - a single green would appear to indicate normal
operation.
What happens with a resistive load (eg 4x12V automotive bulbs in series, a
220V incandecent lamp or a heater element)? With 3kW output capability it
should be able to provide 50A or so current. Note that the output
protection is foldback type so short circuit current will be less (~25A
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our rectifiers
(otherwise pretty similar). Probably there's a data sheet somewhere on the
web giving info on what's required for this or a similar product. If you
can find info on the commands for a similar but non-identical Huawei
rectifier it may assist as companies tend to retain the same protocol &
commands etc.. I don't think there's a Telecom standard for this but there
may be something these days. Or the commands may be encrypted in some way.
:-( in which case hardware adjustment may be the only option if you need to
change the voltage.
Once you register on the Huawei website there seems to be a lot of
datasheets available. It just may take a while to find what you need.
RP
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as base
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published pdf
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some CAN bus
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive and
boots up and even puts a small test voltage out, and the 3 status LED
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I already got
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------

Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/
Richard Prosser
2018-11-13 08:08:27 UTC
Permalink
Yes, it could be that it requires an output voltage before it starts as
part of the startup. If so, it can probably be disabled once you know the
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V' bus. There
are likely to be output capacitors and without precharge the input current
surge will damage the contacts. The fact that the precharge is on the neg.
rail is another clue that it's a telecom charger - most telecom circuits
are positive earth to minimise corrosion damage.

i guess you know that fast charging virtually any sort of battery will
reduce its life.

RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to begin with
to add its juice to? Similar to a cell phone battery protection circuit
that may need to be woken up by charging its internal battery beyond a
certain threshold if the battery has been sitting for too long, this supply
may not dare to power up what appears to be a dead battery? Could be an
option set within some internal config register though... I could test
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not indicated must
belong to the negative terminal of the output DC because that negative PCB
contact is recessed as if for hot plugging, I doubt it has any significance
in regular operation tough?
The green LED is the one that remains on, but if I probe around and bridge
that ~1 Volt feeler voltage that Voltage goes zero and the red LED comes on.
I have put it in a case with a DPS5020 50V 20A digital power supply to
provide a side function while I develop my own charge pump and controller.
So there would be a capacitive load if the supply was to do a test run just
like that... I thought about trying Huawei directly, so registering is a
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a exciting -
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the banks gets
even close to a save voltage limit...
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup but you may
need comms to adjust the voltage & current limit to your requirement.
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power limit etc.
What Leds are showing? - a single green would appear to indicate normal
operation.
What happens with a resistive load (eg 4x12V automotive bulbs in series,
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output capability it
should be able to provide 50A or so current. Note that the output
protection is foldback type so short circuit current will be less (~25A
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet somewhere on
the
Post by Richard Prosser
web giving info on what's required for this or a similar product. If you
can find info on the commands for a similar but non-identical Huawei
rectifier it may assist as companies tend to retain the same protocol &
commands etc.. I don't think there's a Telecom standard for this but
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted in some
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option if you need
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a lot of
datasheets available. It just may take a while to find what you need.
RP
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as base
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published pdf
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some CAN bus
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive and
boots up and even puts a small test voltage out, and the 3 status LED
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I already
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http
Tobias Gogolin
2018-11-14 15:00:31 UTC
Permalink
Meanwhile esteemed Wizards (danke Russel...)
I did some tests, and while I got no power out yet I found a following
scheme:
If I power up with no prepared voltage on the output, just the green LED
comes on
If I then power up 48 or 53V the Orange LED starts flashing, and the Red
one comes on

If I power up with prepared 48V on the output, the green LED comes on and
the orange flashes
If I then remove the prepared input the output just goes down too and the
red LED comes on


If I power up with prepared 53 or 58V on the output, the green LED comes
alone the first few times (now the orange flashes
always too) If I then remove the prepared input the output just goes down
too and the red LED comes on

any voltage change while on is attested with the red coming on...

So I gather there is no other than having to protocol via CAN bus?
Post by Richard Prosser
Yes, it could be that it requires an output voltage before it starts as
part of the startup. If so, it can probably be disabled once you know the
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V' bus. There
are likely to be output capacitors and without precharge the input current
surge will damage the contacts. The fact that the precharge is on the neg.
rail is another clue that it's a telecom charger - most telecom circuits
are positive earth to minimise corrosion damage.
i guess you know that fast charging virtually any sort of battery will
reduce its life.
RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to begin
with
Post by Tobias Gogolin
to add its juice to? Similar to a cell phone battery protection circuit
that may need to be woken up by charging its internal battery beyond a
certain threshold if the battery has been sitting for too long, this
supply
Post by Tobias Gogolin
may not dare to power up what appears to be a dead battery? Could be an
option set within some internal config register though... I could test
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not indicated must
belong to the negative terminal of the output DC because that negative
PCB
Post by Tobias Gogolin
contact is recessed as if for hot plugging, I doubt it has any
significance
Post by Tobias Gogolin
in regular operation tough?
The green LED is the one that remains on, but if I probe around and
bridge
Post by Tobias Gogolin
that ~1 Volt feeler voltage that Voltage goes zero and the red LED comes on.
I have put it in a case with a DPS5020 50V 20A digital power supply to
provide a side function while I develop my own charge pump and
controller.
Post by Tobias Gogolin
So there would be a capacitive load if the supply was to do a test run
just
Post by Tobias Gogolin
like that... I thought about trying Huawei directly, so registering is a
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a
exciting -
Post by Tobias Gogolin
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the banks
gets
Post by Tobias Gogolin
even close to a save voltage limit...
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup but you
may
Post by Tobias Gogolin
Post by Richard Prosser
need comms to adjust the voltage & current limit to your requirement.
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power limit etc.
What Leds are showing? - a single green would appear to indicate normal
operation.
What happens with a resistive load (eg 4x12V automotive bulbs in
series,
Post by Tobias Gogolin
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output capability
it
Post by Tobias Gogolin
Post by Richard Prosser
should be able to provide 50A or so current. Note that the output
protection is foldback type so short circuit current will be less (~25A
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet somewhere on
the
Post by Richard Prosser
web giving info on what's required for this or a similar product. If
you
Post by Tobias Gogolin
Post by Richard Prosser
can find info on the commands for a similar but non-identical Huawei
rectifier it may assist as companies tend to retain the same protocol &
commands etc.. I don't think there's a Telecom standard for this but
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted in some
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option if you
need
Post by Tobias Gogolin
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a lot of
datasheets available. It just may take a while to find what you need.
RP
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as
base
Post by Tobias Gogolin
Post by Richard Prosser
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published pdf
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some CAN
bus
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive
and
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
boots up and even puts a small test voltage out, and the 3 status LED
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I already
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------

Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailma
Manu Abraham
2018-11-14 15:42:48 UTC
Permalink
CAN bus commands are simple. It uses a mailbox approach. Generally
power supplies use SMBUS (I2C variant) on PC motherboards, PMBUS
(Serial variant) on larger boards or interconnects, CANBUS for large
backplanes. The CANBUS usage can be understood, as the PSU could be
controlled over still longer wires.

That said, how are you sending the commands ?

Please note that CAN Bus requires a termination of 120 ohms IIRC. If
not, there will be echoes on the bus and communication issues.
Looking at your SMPS manual, the GREEN led states everything is fine.
Flashing ORANGE/YELLOW states that bad CAN BUS communication. You
might need to look at the BUS termination, probably. As I said
earlier, the SMPS is a network/telecom oriented SMPS, I guess in this
situation, I might guess it is more intended for PoE applications. Now
that 48V is the standard. When you are setting to anything more than
48V, that is considered overvoltage, which is what the manual states
too.

Not surprising isn't it ?

Cheers,
Manu
Post by Tobias Gogolin
Meanwhile esteemed Wizards (danke Russel...)
I did some tests, and while I got no power out yet I found a following
If I power up with no prepared voltage on the output, just the green LED
comes on
If I then power up 48 or 53V the Orange LED starts flashing, and the Red
one comes on
If I power up with prepared 48V on the output, the green LED comes on and
the orange flashes
If I then remove the prepared input the output just goes down too and the
red LED comes on
If I power up with prepared 53 or 58V on the output, the green LED comes
alone the first few times (now the orange flashes
always too) If I then remove the prepared input the output just goes down
too and the red LED comes on
any voltage change while on is attested with the red coming on...
So I gather there is no other than having to protocol via CAN bus?
Post by Richard Prosser
Yes, it could be that it requires an output voltage before it starts as
part of the startup. If so, it can probably be disabled once you know the
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V' bus. There
are likely to be output capacitors and without precharge the input current
surge will damage the contacts. The fact that the precharge is on the neg.
rail is another clue that it's a telecom charger - most telecom circuits
are positive earth to minimise corrosion damage.
i guess you know that fast charging virtually any sort of battery will
reduce its life.
RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to begin
with
Post by Tobias Gogolin
to add its juice to? Similar to a cell phone battery protection circuit
that may need to be woken up by charging its internal battery beyond a
certain threshold if the battery has been sitting for too long, this
supply
Post by Tobias Gogolin
may not dare to power up what appears to be a dead battery? Could be an
option set within some internal config register though... I could test
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not indicated must
belong to the negative terminal of the output DC because that negative
PCB
Post by Tobias Gogolin
contact is recessed as if for hot plugging, I doubt it has any
significance
Post by Tobias Gogolin
in regular operation tough?
The green LED is the one that remains on, but if I probe around and
bridge
Post by Tobias Gogolin
that ~1 Volt feeler voltage that Voltage goes zero and the red LED comes on.
I have put it in a case with a DPS5020 50V 20A digital power supply to
provide a side function while I develop my own charge pump and
controller.
Post by Tobias Gogolin
So there would be a capacitive load if the supply was to do a test run
just
Post by Tobias Gogolin
like that... I thought about trying Huawei directly, so registering is a
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a
exciting -
Post by Tobias Gogolin
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the banks
gets
Post by Tobias Gogolin
even close to a save voltage limit...
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup but you
may
Post by Tobias Gogolin
Post by Richard Prosser
need comms to adjust the voltage & current limit to your requirement.
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power limit etc.
What Leds are showing? - a single green would appear to indicate normal
operation.
What happens with a resistive load (eg 4x12V automotive bulbs in
series,
Post by Tobias Gogolin
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output capability
it
Post by Tobias Gogolin
Post by Richard Prosser
should be able to provide 50A or so current. Note that the output
protection is foldback type so short circuit current will be less (~25A
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet somewhere on
the
Post by Richard Prosser
web giving info on what's required for this or a similar product. If
you
Post by Tobias Gogolin
Post by Richard Prosser
can find info on the commands for a similar but non-identical Huawei
rectifier it may assist as companies tend to retain the same protocol &
commands etc.. I don't think there's a Telecom standard for this but
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted in some
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option if you
need
Post by Tobias Gogolin
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a lot of
datasheets available. It just may take a while to find what you need.
RP
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as
base
Post by Tobias Gogolin
Post by Richard Prosser
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published pdf
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some CAN
bus
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive
and
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
boots up and even puts a small test voltage out, and the 3 status LED
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I already
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http
Tobias Gogolin
2018-11-14 16:12:56 UTC
Permalink
Well yes, I have an Arduino CAN bus shield (MCP2515) with selectable
termination. It's not even wired yet, because I wouldn't know what to
listen for, or how to talk to - saying what?
I am not sure does CAN bus have some kind of discovery protocol?
Post by Manu Abraham
CAN bus commands are simple. It uses a mailbox approach. Generally
power supplies use SMBUS (I2C variant) on PC motherboards, PMBUS
(Serial variant) on larger boards or interconnects, CANBUS for large
backplanes. The CANBUS usage can be understood, as the PSU could be
controlled over still longer wires.
That said, how are you sending the commands ?
Please note that CAN Bus requires a termination of 120 ohms IIRC. If
not, there will be echoes on the bus and communication issues.
Looking at your SMPS manual, the GREEN led states everything is fine.
Flashing ORANGE/YELLOW states that bad CAN BUS communication. You
might need to look at the BUS termination, probably. As I said
earlier, the SMPS is a network/telecom oriented SMPS, I guess in this
situation, I might guess it is more intended for PoE applications. Now
that 48V is the standard. When you are setting to anything more than
48V, that is considered overvoltage, which is what the manual states
too.
Not surprising isn't it ?
Cheers,
Manu
Post by Tobias Gogolin
Meanwhile esteemed Wizards (danke Russel...)
I did some tests, and while I got no power out yet I found a following
If I power up with no prepared voltage on the output, just the green LED
comes on
If I then power up 48 or 53V the Orange LED starts flashing, and the Red
one comes on
If I power up with prepared 48V on the output, the green LED comes on and
the orange flashes
If I then remove the prepared input the output just goes down too and the
red LED comes on
If I power up with prepared 53 or 58V on the output, the green LED comes
alone the first few times (now the orange flashes
always too) If I then remove the prepared input the output just goes
down
Post by Tobias Gogolin
too and the red LED comes on
any voltage change while on is attested with the red coming on...
So I gather there is no other than having to protocol via CAN bus?
Post by Richard Prosser
Yes, it could be that it requires an output voltage before it starts as
part of the startup. If so, it can probably be disabled once you know
the
Post by Tobias Gogolin
Post by Richard Prosser
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V' bus.
There
Post by Tobias Gogolin
Post by Richard Prosser
are likely to be output capacitors and without precharge the input
current
Post by Tobias Gogolin
Post by Richard Prosser
surge will damage the contacts. The fact that the precharge is on the
neg.
Post by Tobias Gogolin
Post by Richard Prosser
rail is another clue that it's a telecom charger - most telecom
circuits
Post by Tobias Gogolin
Post by Richard Prosser
are positive earth to minimise corrosion damage.
i guess you know that fast charging virtually any sort of battery will
reduce its life.
RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to begin
with
Post by Tobias Gogolin
to add its juice to? Similar to a cell phone battery protection
circuit
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
that may need to be woken up by charging its internal battery beyond
a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
certain threshold if the battery has been sitting for too long, this
supply
Post by Tobias Gogolin
may not dare to power up what appears to be a dead battery? Could be
an
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
option set within some internal config register though... I could
test
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not indicated
must
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
belong to the negative terminal of the output DC because that
negative
Post by Tobias Gogolin
Post by Richard Prosser
PCB
Post by Tobias Gogolin
contact is recessed as if for hot plugging, I doubt it has any
significance
Post by Tobias Gogolin
in regular operation tough?
The green LED is the one that remains on, but if I probe around and
bridge
Post by Tobias Gogolin
that ~1 Volt feeler voltage that Voltage goes zero and the red LED
comes
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
on.
I have put it in a case with a DPS5020 50V 20A digital power supply
to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
provide a side function while I develop my own charge pump and
controller.
Post by Tobias Gogolin
So there would be a capacitive load if the supply was to do a test
run
Post by Tobias Gogolin
Post by Richard Prosser
just
Post by Tobias Gogolin
like that... I thought about trying Huawei directly, so registering
is a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a
exciting -
Post by Tobias Gogolin
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the banks
gets
Post by Tobias Gogolin
even close to a save voltage limit...
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup but
you
Post by Tobias Gogolin
Post by Richard Prosser
may
Post by Tobias Gogolin
Post by Richard Prosser
need comms to adjust the voltage & current limit to your
requirement.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power limit
etc.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
What Leds are showing? - a single green would appear to indicate
normal
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
operation.
What happens with a resistive load (eg 4x12V automotive bulbs in
series,
Post by Tobias Gogolin
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output
capability
Post by Tobias Gogolin
Post by Richard Prosser
it
Post by Tobias Gogolin
Post by Richard Prosser
should be able to provide 50A or so current. Note that the output
protection is foldback type so short circuit current will be less
(~25A
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet
somewhere on
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
the
Post by Richard Prosser
web giving info on what's required for this or a similar product.
If
Post by Tobias Gogolin
Post by Richard Prosser
you
Post by Tobias Gogolin
Post by Richard Prosser
can find info on the commands for a similar but non-identical
Huawei
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
rectifier it may assist as companies tend to retain the same
protocol &
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
commands etc.. I don't think there's a Telecom standard for this
but
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted in
some
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option if you
need
Post by Tobias Gogolin
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a lot of
datasheets available. It just may take a while to find what you
need.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
RP
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as
base
Post by Tobias Gogolin
Post by Richard Prosser
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published
pdf
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some
CAN
Post by Tobias Gogolin
Post by Richard Prosser
bus
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
hacking before it even releases its awesome power! I published
in the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
thread that helped me to find this device including a look
inside the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is
alive
Post by Tobias Gogolin
Post by Richard Prosser
and
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
boots up and even puts a small test voltage out, and the 3
status LED
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I
already
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know
what to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen,
die uns
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------

Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/lis
Manu Abraham
2018-11-14 16:54:39 UTC
Permalink
This is exactly what I've asked; How you are sending the commands. It
would be meaningless to say that you are using a CAN
transceiver/micro-controller combo to do so. Obviously, without which
you cannot do so.

My question was exactly what are you command sequences or based on
what knowledge are you sending those commands, since I could not see
any on the manual. If you are asking about the protocol
implementation/command structure how the micro controller interprets
the CAN messages, that's done at the firmware level on the
micro-controller. It could vary from case to case. Only the vendor can
help you on protocol commands for bus monitoring/control. I don't
think there would be a one standard command for all PSU's. Usually,
the PSU vendor would have their own application to interface with the
communication bus. In some corner cases, some vendors might share with
you the commands that interface to the micro-controller.

To get a basic idea of the interface, you can have a look at this.
http://www.keil.com/appnotes/files/apnt_236.pdf

If you want to try hacking the command/protocol structure/interface,
maybe you might find a starting point here.
https://www.blackhat.com/docs/asia-15/materials/asia-15-Evenchick-Hopping-On-The-Can-Bus.pdf

The best would be to contact the vendor, if you are looking for a
hassle free path.

Cheers,
Manu
Post by Tobias Gogolin
Well yes, I have an Arduino CAN bus shield (MCP2515) with selectable
termination. It's not even wired yet, because I wouldn't know what to
listen for, or how to talk to - saying what?
I am not sure does CAN bus have some kind of discovery protocol?
Post by Manu Abraham
CAN bus commands are simple. It uses a mailbox approach. Generally
power supplies use SMBUS (I2C variant) on PC motherboards, PMBUS
(Serial variant) on larger boards or interconnects, CANBUS for large
backplanes. The CANBUS usage can be understood, as the PSU could be
controlled over still longer wires.
That said, how are you sending the commands ?
Please note that CAN Bus requires a termination of 120 ohms IIRC. If
not, there will be echoes on the bus and communication issues.
Looking at your SMPS manual, the GREEN led states everything is fine.
Flashing ORANGE/YELLOW states that bad CAN BUS communication. You
might need to look at the BUS termination, probably. As I said
earlier, the SMPS is a network/telecom oriented SMPS, I guess in this
situation, I might guess it is more intended for PoE applications. Now
that 48V is the standard. When you are setting to anything more than
48V, that is considered overvoltage, which is what the manual states
too.
Not surprising isn't it ?
Cheers,
Manu
Post by Tobias Gogolin
Meanwhile esteemed Wizards (danke Russel...)
I did some tests, and while I got no power out yet I found a following
If I power up with no prepared voltage on the output, just the green LED
comes on
If I then power up 48 or 53V the Orange LED starts flashing, and the Red
one comes on
If I power up with prepared 48V on the output, the green LED comes on and
the orange flashes
If I then remove the prepared input the output just goes down too and the
red LED comes on
If I power up with prepared 53 or 58V on the output, the green LED comes
alone the first few times (now the orange flashes
always too) If I then remove the prepared input the output just goes
down
Post by Tobias Gogolin
too and the red LED comes on
any voltage change while on is attested with the red coming on...
So I gather there is no other than having to protocol via CAN bus?
Post by Richard Prosser
Yes, it could be that it requires an output voltage before it starts as
part of the startup. If so, it can probably be disabled once you know
the
Post by Tobias Gogolin
Post by Richard Prosser
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V' bus.
There
Post by Tobias Gogolin
Post by Richard Prosser
are likely to be output capacitors and without precharge the input
current
Post by Tobias Gogolin
Post by Richard Prosser
surge will damage the contacts. The fact that the precharge is on the
neg.
Post by Tobias Gogolin
Post by Richard Prosser
rail is another clue that it's a telecom charger - most telecom
circuits
Post by Tobias Gogolin
Post by Richard Prosser
are positive earth to minimise corrosion damage.
i guess you know that fast charging virtually any sort of battery will
reduce its life.
RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to begin
with
Post by Tobias Gogolin
to add its juice to? Similar to a cell phone battery protection
circuit
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
that may need to be woken up by charging its internal battery beyond
a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
certain threshold if the battery has been sitting for too long, this
supply
Post by Tobias Gogolin
may not dare to power up what appears to be a dead battery? Could be
an
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
option set within some internal config register though... I could
test
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not indicated
must
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
belong to the negative terminal of the output DC because that
negative
Post by Tobias Gogolin
Post by Richard Prosser
PCB
Post by Tobias Gogolin
contact is recessed as if for hot plugging, I doubt it has any
significance
Post by Tobias Gogolin
in regular operation tough?
The green LED is the one that remains on, but if I probe around and
bridge
Post by Tobias Gogolin
that ~1 Volt feeler voltage that Voltage goes zero and the red LED
comes
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
on.
I have put it in a case with a DPS5020 50V 20A digital power supply
to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
provide a side function while I develop my own charge pump and
controller.
Post by Tobias Gogolin
So there would be a capacitive load if the supply was to do a test
run
Post by Tobias Gogolin
Post by Richard Prosser
just
Post by Tobias Gogolin
like that... I thought about trying Huawei directly, so registering
is a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a
exciting -
Post by Tobias Gogolin
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the banks
gets
Post by Tobias Gogolin
even close to a save voltage limit...
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup but
you
Post by Tobias Gogolin
Post by Richard Prosser
may
Post by Tobias Gogolin
Post by Richard Prosser
need comms to adjust the voltage & current limit to your
requirement.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power limit
etc.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
What Leds are showing? - a single green would appear to indicate
normal
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
operation.
What happens with a resistive load (eg 4x12V automotive bulbs in
series,
Post by Tobias Gogolin
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output
capability
Post by Tobias Gogolin
Post by Richard Prosser
it
Post by Tobias Gogolin
Post by Richard Prosser
should be able to provide 50A or so current. Note that the output
protection is foldback type so short circuit current will be less
(~25A
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet
somewhere on
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
the
Post by Richard Prosser
web giving info on what's required for this or a similar product.
If
Post by Tobias Gogolin
Post by Richard Prosser
you
Post by Tobias Gogolin
Post by Richard Prosser
can find info on the commands for a similar but non-identical
Huawei
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
rectifier it may assist as companies tend to retain the same
protocol &
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
commands etc.. I don't think there's a Telecom standard for this
but
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted in
some
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option if you
need
Post by Tobias Gogolin
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a lot of
datasheets available. It just may take a while to find what you
need.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
RP
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as
base
Post by Tobias Gogolin
Post by Richard Prosser
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published
pdf
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some
CAN
Post by Tobias Gogolin
Post by Richard Prosser
bus
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
hacking before it even releases its awesome power! I published
in the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
thread that helped me to find this device including a look
inside the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is
alive
Post by Tobias Gogolin
Post by Richard Prosser
and
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
boots up and even puts a small test voltage out, and the 3
status LED
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I
already
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know
what to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen,
die uns
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mi
Tobias Gogolin
2018-11-15 10:19:51 UTC
Permalink
Ok very true, wonder what the chances are to convince the vendor how
conveniently the demand for his product would increase if the
communications protocol to his device was 'leaked' or even outright made
public! We wouldn't by any chance have some inside contact to Huawei in
this fraternity?

I found a forum post where some ham operators where talking about the same
issue but apparently he had been fortunate enough to get one configured so
it would start up by itself ( http://forum.db3om.de/ftopic24798.html )
Alternatively if this first option fails and Huawei doesn't play nice would
it be realistic to hope that if somebody connected a CAN sniffer and I (or
somebody even more prepped for the task) somehow got a dump of a
configuration session with the device it might be possible to deduct the
important parts of the configuration protocol? I seen similar things being
done to try to open up an after market for Tesla Motors components...
What are the chances that if I find the device to has a command shell type
of terminal mode via CAN?
Unlikely they would have some kind of security encryption scheme for this
type of device?

Thanks Manu for the valuable links! I definitely want to learn everything
there is and start using the bus on my own vehicle designs, so I ordered
more CAN modules too...
Post by Manu Abraham
This is exactly what I've asked; How you are sending the commands. It
would be meaningless to say that you are using a CAN
transceiver/micro-controller combo to do so. Obviously, without which
you cannot do so.
My question was exactly what are you command sequences or based on
what knowledge are you sending those commands, since I could not see
any on the manual. If you are asking about the protocol
implementation/command structure how the micro controller interprets
the CAN messages, that's done at the firmware level on the
micro-controller. It could vary from case to case. Only the vendor can
help you on protocol commands for bus monitoring/control. I don't
think there would be a one standard command for all PSU's. Usually,
the PSU vendor would have their own application to interface with the
communication bus. In some corner cases, some vendors might share with
you the commands that interface to the micro-controller.
To get a basic idea of the interface, you can have a look at this.
http://www.keil.com/appnotes/files/apnt_236.pdf
If you want to try hacking the command/protocol structure/interface,
maybe you might find a starting point here.
https://www.blackhat.com/docs/asia-15/materials/asia-15-Evenchick-Hopping-On-The-Can-Bus.pdf
The best would be to contact the vendor, if you are looking for a
hassle free path.
Cheers,
Manu
Post by Tobias Gogolin
Well yes, I have an Arduino CAN bus shield (MCP2515) with selectable
termination. It's not even wired yet, because I wouldn't know what to
listen for, or how to talk to - saying what?
I am not sure does CAN bus have some kind of discovery protocol?
Post by Manu Abraham
CAN bus commands are simple. It uses a mailbox approach. Generally
power supplies use SMBUS (I2C variant) on PC motherboards, PMBUS
(Serial variant) on larger boards or interconnects, CANBUS for large
backplanes. The CANBUS usage can be understood, as the PSU could be
controlled over still longer wires.
That said, how are you sending the commands ?
Please note that CAN Bus requires a termination of 120 ohms IIRC. If
not, there will be echoes on the bus and communication issues.
Looking at your SMPS manual, the GREEN led states everything is fine.
Flashing ORANGE/YELLOW states that bad CAN BUS communication. You
might need to look at the BUS termination, probably. As I said
earlier, the SMPS is a network/telecom oriented SMPS, I guess in this
situation, I might guess it is more intended for PoE applications. Now
that 48V is the standard. When you are setting to anything more than
48V, that is considered overvoltage, which is what the manual states
too.
Not surprising isn't it ?
Cheers,
Manu
Post by Tobias Gogolin
Meanwhile esteemed Wizards (danke Russel...)
I did some tests, and while I got no power out yet I found a
following
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
If I power up with no prepared voltage on the output, just the green
LED
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
comes on
If I then power up 48 or 53V the Orange LED starts flashing, and the
Red
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
one comes on
If I power up with prepared 48V on the output, the green LED comes
on and
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
the orange flashes
If I then remove the prepared input the output just goes down too
and the
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
red LED comes on
If I power up with prepared 53 or 58V on the output, the green LED
comes
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
alone the first few times (now the orange flashes
always too) If I then remove the prepared input the output just goes
down
Post by Tobias Gogolin
too and the red LED comes on
any voltage change while on is attested with the red coming on...
So I gather there is no other than having to protocol via CAN bus?
Post by Richard Prosser
Yes, it could be that it requires an output voltage before it
starts as
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
part of the startup. If so, it can probably be disabled once you
know
Post by Tobias Gogolin
Post by Manu Abraham
the
Post by Tobias Gogolin
Post by Richard Prosser
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V'
bus.
Post by Tobias Gogolin
Post by Manu Abraham
There
Post by Tobias Gogolin
Post by Richard Prosser
are likely to be output capacitors and without precharge the input
current
Post by Tobias Gogolin
Post by Richard Prosser
surge will damage the contacts. The fact that the precharge is on
the
Post by Tobias Gogolin
Post by Manu Abraham
neg.
Post by Tobias Gogolin
Post by Richard Prosser
rail is another clue that it's a telecom charger - most telecom
circuits
Post by Tobias Gogolin
Post by Richard Prosser
are positive earth to minimise corrosion damage.
i guess you know that fast charging virtually any sort of battery
will
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
reduce its life.
RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to
begin
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
with
Post by Tobias Gogolin
to add its juice to? Similar to a cell phone battery protection
circuit
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
that may need to be woken up by charging its internal battery
beyond
Post by Tobias Gogolin
Post by Manu Abraham
a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
certain threshold if the battery has been sitting for too long,
this
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
supply
Post by Tobias Gogolin
may not dare to power up what appears to be a dead battery?
Could be
Post by Tobias Gogolin
Post by Manu Abraham
an
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
option set within some internal config register though... I could
test
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not
indicated
Post by Tobias Gogolin
Post by Manu Abraham
must
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
belong to the negative terminal of the output DC because that
negative
Post by Tobias Gogolin
Post by Richard Prosser
PCB
Post by Tobias Gogolin
contact is recessed as if for hot plugging, I doubt it has any
significance
Post by Tobias Gogolin
in regular operation tough?
The green LED is the one that remains on, but if I probe around
and
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
bridge
Post by Tobias Gogolin
that ~1 Volt feeler voltage that Voltage goes zero and the red
LED
Post by Tobias Gogolin
Post by Manu Abraham
comes
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
on.
I have put it in a case with a DPS5020 50V 20A digital power
supply
Post by Tobias Gogolin
Post by Manu Abraham
to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
provide a side function while I develop my own charge pump and
controller.
Post by Tobias Gogolin
So there would be a capacitive load if the supply was to do a
test
Post by Tobias Gogolin
Post by Manu Abraham
run
Post by Tobias Gogolin
Post by Richard Prosser
just
Post by Tobias Gogolin
like that... I thought about trying Huawei directly, so
registering
Post by Tobias Gogolin
Post by Manu Abraham
is a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a
exciting -
Post by Tobias Gogolin
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the
banks
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
gets
Post by Tobias Gogolin
even close to a save voltage limit...
On Mon, Nov 12, 2018 at 9:31 PM Richard Prosser <
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup
but
Post by Tobias Gogolin
Post by Manu Abraham
you
Post by Tobias Gogolin
Post by Richard Prosser
may
Post by Tobias Gogolin
Post by Richard Prosser
need comms to adjust the voltage & current limit to your
requirement.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power
limit
Post by Tobias Gogolin
Post by Manu Abraham
etc.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
What Leds are showing? - a single green would appear to
indicate
Post by Tobias Gogolin
Post by Manu Abraham
normal
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
operation.
What happens with a resistive load (eg 4x12V automotive bulbs
in
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
series,
Post by Tobias Gogolin
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output
capability
Post by Tobias Gogolin
Post by Richard Prosser
it
Post by Tobias Gogolin
Post by Richard Prosser
should be able to provide 50A or so current. Note that the
output
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
protection is foldback type so short circuit current will be
less
Post by Tobias Gogolin
Post by Manu Abraham
(~25A
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet
somewhere on
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
the
Post by Richard Prosser
web giving info on what's required for this or a similar
product.
Post by Tobias Gogolin
Post by Manu Abraham
If
Post by Tobias Gogolin
Post by Richard Prosser
you
Post by Tobias Gogolin
Post by Richard Prosser
can find info on the commands for a similar but non-identical
Huawei
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
rectifier it may assist as companies tend to retain the same
protocol &
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
commands etc.. I don't think there's a Telecom standard for
this
Post by Tobias Gogolin
Post by Manu Abraham
but
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted
in
Post by Tobias Gogolin
Post by Manu Abraham
some
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option
if you
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
need
Post by Tobias Gogolin
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a
lot of
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
datasheets available. It just may take a while to find what you
need.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
RP
On Mon, 12 Nov 2018 at 23:32, Tobias Gogolin <
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2
rectifier as
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
base
Post by Tobias Gogolin
Post by Richard Prosser
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the
published
Post by Tobias Gogolin
Post by Manu Abraham
pdf
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs
some
Post by Tobias Gogolin
Post by Manu Abraham
CAN
Post by Tobias Gogolin
Post by Richard Prosser
bus
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
hacking before it even releases its awesome power! I
published
Post by Tobias Gogolin
Post by Manu Abraham
in the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
thread that helped me to find this device including a look
inside the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
really sophisticated design of this Access Power Rectifier
here.
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is
alive
Post by Tobias Gogolin
Post by Richard Prosser
and
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
boots up and even puts a small test voltage out, and the 3
status LED
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I
already
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know
what to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen,
die uns
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list
archive
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list
archive
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Manu Abraham
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------

Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://
Manu Abraham
2018-11-15 16:53:16 UTC
Permalink
From your post, it was difficult to understand your english, I am by
no means a german speaker, sorry about that.

From what you wrote: "prepared 48V output", I was under the belief you
were able to get a 48V output. With your latest post, I understood
incorrectly, what you tried to imply there.

If you want to talk to the vendor, the best that you could ask them to
provide the software to configure the power supply. Couldn't be any
simpler.

I doubt, even if anyone from the vendor were to be on the list,
providing a closed protocol of a product, would be the last thing
anyone would remotely try to do.
Why don't you get the actual necessary software from the vendor and
use it directly ? That would be the simplest to do. If really
necessary, you can snoop on the communication to get the basic command
infrastructure to work on some bare metal platform.

There is talk about secure CAN in the industry, so that vendors can
lock people out, or IOW keep their automobiles safer, another way you
look at it. A coin has two sides to it; not one.

That said, a PSU controller is unlikely to be a device with large
amounts of memory, given the cost of the PSU, it is unlikely that way
either. The lesser memory implies that the amount of parsing /
tokenizing would be just minimal. More tokens imply more storage of
words. Most likely, you could open the power supply, have a look at
what controller it uses and so on. If you are lucky, you could even
find that controller datasheet and get a fair understanding of what
happens under the hood inside that PSU. The next good thing would be
to know, what the PSU was really intended for. Then you could try to
gain info, how it is run in that context. Please note, that all this
takes time and and times, could be frustrating. I don't think, there's
an easy way out.

If I were you, I would've got the actual software/platform, snooped on
the bus, logged each command, to be later utilized into a dictionary
for RE. the comms. syntax. I think it would be a very interesting
project. If you've used Linux and Digital Video hardware, probably you
would've known. Most of the Linux DVB drivers were written that way.
The rev. engg'd stuff would be then compared with actual datasheets
(if available) and the quality of the driver improved eventually. ;-)
It's painful, but if you enjoy doing such things, very much an
enjoyable project. It will take you a long way. In the process, you
will be giving away knowledge to other people in the same way too.

Let us people on this list know, how you get along. Not that anyone
could help, due to lack of time, hardware and the list could go on ..
But you could be of help to someone who comes along later on another
day like this.


Cheers,
Manu
Post by Tobias Gogolin
Ok very true, wonder what the chances are to convince the vendor how
conveniently the demand for his product would increase if the
communications protocol to his device was 'leaked' or even outright made
public! We wouldn't by any chance have some inside contact to Huawei in
this fraternity?
I found a forum post where some ham operators where talking about the same
issue but apparently he had been fortunate enough to get one configured so
it would start up by itself ( http://forum.db3om.de/ftopic24798.html )
Alternatively if this first option fails and Huawei doesn't play nice would
it be realistic to hope that if somebody connected a CAN sniffer and I (or
somebody even more prepped for the task) somehow got a dump of a
configuration session with the device it might be possible to deduct the
important parts of the configuration protocol? I seen similar things being
done to try to open up an after market for Tesla Motors components...
What are the chances that if I find the device to has a command shell type
of terminal mode via CAN?
Unlikely they would have some kind of security encryption scheme for this
type of device?
Thanks Manu for the valuable links! I definitely want to learn everything
there is and start using the bus on my own vehicle designs, so I ordered
more CAN modules too...
Post by Manu Abraham
This is exactly what I've asked; How you are sending the commands. It
would be meaningless to say that you are using a CAN
transceiver/micro-controller combo to do so. Obviously, without which
you cannot do so.
My question was exactly what are you command sequences or based on
what knowledge are you sending those commands, since I could not see
any on the manual. If you are asking about the protocol
implementation/command structure how the micro controller interprets
the CAN messages, that's done at the firmware level on the
micro-controller. It could vary from case to case. Only the vendor can
help you on protocol commands for bus monitoring/control. I don't
think there would be a one standard command for all PSU's. Usually,
the PSU vendor would have their own application to interface with the
communication bus. In some corner cases, some vendors might share with
you the commands that interface to the micro-controller.
To get a basic idea of the interface, you can have a look at this.
http://www.keil.com/appnotes/files/apnt_236.pdf
If you want to try hacking the command/protocol structure/interface,
maybe you might find a starting point here.
https://www.blackhat.com/docs/asia-15/materials/asia-15-Evenchick-Hopping-On-The-Can-Bus.pdf
The best would be to contact the vendor, if you are looking for a
hassle free path.
Cheers,
Manu
Post by Tobias Gogolin
Well yes, I have an Arduino CAN bus shield (MCP2515) with selectable
termination. It's not even wired yet, because I wouldn't know what to
listen for, or how to talk to - saying what?
I am not sure does CAN bus have some kind of discovery protocol?
Post by Manu Abraham
CAN bus commands are simple. It uses a mailbox approach. Generally
power supplies use SMBUS (I2C variant) on PC motherboards, PMBUS
(Serial variant) on larger boards or interconnects, CANBUS for large
backplanes. The CANBUS usage can be understood, as the PSU could be
controlled over still longer wires.
That said, how are you sending the commands ?
Please note that CAN Bus requires a termination of 120 ohms IIRC. If
not, there will be echoes on the bus and communication issues.
Looking at your SMPS manual, the GREEN led states everything is fine.
Flashing ORANGE/YELLOW states that bad CAN BUS communication. You
might need to look at the BUS termination, probably. As I said
earlier, the SMPS is a network/telecom oriented SMPS, I guess in this
situation, I might guess it is more intended for PoE applications. Now
that 48V is the standard. When you are setting to anything more than
48V, that is considered overvoltage, which is what the manual states
too.
Not surprising isn't it ?
Cheers,
Manu
Post by Tobias Gogolin
Meanwhile esteemed Wizards (danke Russel...)
I did some tests, and while I got no power out yet I found a
following
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
If I power up with no prepared voltage on the output, just the green
LED
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
comes on
If I then power up 48 or 53V the Orange LED starts flashing, and the
Red
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
one comes on
If I power up with prepared 48V on the output, the green LED comes
on and
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
the orange flashes
If I then remove the prepared input the output just goes down too
and the
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
red LED comes on
If I power up with prepared 53 or 58V on the output, the green LED
comes
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
alone the first few times (now the orange flashes
always too) If I then remove the prepared input the output just goes
down
Post by Tobias Gogolin
too and the red LED comes on
any voltage change while on is attested with the red coming on...
So I gather there is no other than having to protocol via CAN bus?
Post by Richard Prosser
Yes, it could be that it requires an output voltage before it
starts as
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
part of the startup. If so, it can probably be disabled once you
know
Post by Tobias Gogolin
Post by Manu Abraham
the
Post by Tobias Gogolin
Post by Richard Prosser
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V'
bus.
Post by Tobias Gogolin
Post by Manu Abraham
There
Post by Tobias Gogolin
Post by Richard Prosser
are likely to be output capacitors and without precharge the input
current
Post by Tobias Gogolin
Post by Richard Prosser
surge will damage the contacts. The fact that the precharge is on
the
Post by Tobias Gogolin
Post by Manu Abraham
neg.
Post by Tobias Gogolin
Post by Richard Prosser
rail is another clue that it's a telecom charger - most telecom
circuits
Post by Tobias Gogolin
Post by Richard Prosser
are positive earth to minimise corrosion damage.
i guess you know that fast charging virtually any sort of battery
will
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
reduce its life.
RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to
begin
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
with
Post by Tobias Gogolin
to add its juice to? Similar to a cell phone battery protection
circuit
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
that may need to be woken up by charging its internal battery
beyond
Post by Tobias Gogolin
Post by Manu Abraham
a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
certain threshold if the battery has been sitting for too long,
this
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
supply
Post by Tobias Gogolin
may not dare to power up what appears to be a dead battery?
Could be
Post by Tobias Gogolin
Post by Manu Abraham
an
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
option set within some internal config register though... I could
test
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not
indicated
Post by Tobias Gogolin
Post by Manu Abraham
must
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
belong to the negative terminal of the output DC because that
negative
Post by Tobias Gogolin
Post by Richard Prosser
PCB
Post by Tobias Gogolin
contact is recessed as if for hot plugging, I doubt it has any
significance
Post by Tobias Gogolin
in regular operation tough?
The green LED is the one that remains on, but if I probe around
and
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
bridge
Post by Tobias Gogolin
that ~1 Volt feeler voltage that Voltage goes zero and the red
LED
Post by Tobias Gogolin
Post by Manu Abraham
comes
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
on.
I have put it in a case with a DPS5020 50V 20A digital power
supply
Post by Tobias Gogolin
Post by Manu Abraham
to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
provide a side function while I develop my own charge pump and
controller.
Post by Tobias Gogolin
So there would be a capacitive load if the supply was to do a
test
Post by Tobias Gogolin
Post by Manu Abraham
run
Post by Tobias Gogolin
Post by Richard Prosser
just
Post by Tobias Gogolin
like that... I thought about trying Huawei directly, so
registering
Post by Tobias Gogolin
Post by Manu Abraham
is a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a
exciting -
Post by Tobias Gogolin
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the
banks
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
gets
Post by Tobias Gogolin
even close to a save voltage limit...
On Mon, Nov 12, 2018 at 9:31 PM Richard Prosser <
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup
but
Post by Tobias Gogolin
Post by Manu Abraham
you
Post by Tobias Gogolin
Post by Richard Prosser
may
Post by Tobias Gogolin
Post by Richard Prosser
need comms to adjust the voltage & current limit to your
requirement.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power
limit
Post by Tobias Gogolin
Post by Manu Abraham
etc.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
What Leds are showing? - a single green would appear to
indicate
Post by Tobias Gogolin
Post by Manu Abraham
normal
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
operation.
What happens with a resistive load (eg 4x12V automotive bulbs
in
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
series,
Post by Tobias Gogolin
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output
capability
Post by Tobias Gogolin
Post by Richard Prosser
it
Post by Tobias Gogolin
Post by Richard Prosser
should be able to provide 50A or so current. Note that the
output
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
protection is foldback type so short circuit current will be
less
Post by Tobias Gogolin
Post by Manu Abraham
(~25A
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet
somewhere on
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
the
Post by Richard Prosser
web giving info on what's required for this or a similar
product.
Post by Tobias Gogolin
Post by Manu Abraham
If
Post by Tobias Gogolin
Post by Richard Prosser
you
Post by Tobias Gogolin
Post by Richard Prosser
can find info on the commands for a similar but non-identical
Huawei
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
rectifier it may assist as companies tend to retain the same
protocol &
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
commands etc.. I don't think there's a Telecom standard for
this
Post by Tobias Gogolin
Post by Manu Abraham
but
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted
in
Post by Tobias Gogolin
Post by Manu Abraham
some
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option
if you
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
need
Post by Tobias Gogolin
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a
lot of
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
datasheets available. It just may take a while to find what you
need.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
RP
On Mon, 12 Nov 2018 at 23:32, Tobias Gogolin <
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2
rectifier as
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
base
Post by Tobias Gogolin
Post by Richard Prosser
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the
published
Post by Tobias Gogolin
Post by Manu Abraham
pdf
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs
some
Post by Tobias Gogolin
Post by Manu Abraham
CAN
Post by Tobias Gogolin
Post by Richard Prosser
bus
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
hacking before it even releases its awesome power! I
published
Post by Tobias Gogolin
Post by Manu Abraham
in the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
thread that helped me to find this device including a look
inside the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
really sophisticated design of this Access Power Rectifier
here.
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is
alive
Post by Tobias Gogolin
Post by Richard Prosser
and
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
boots up and even puts a small test voltage out, and the 3
status LED
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I
already
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know
what to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen,
die uns
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list
archive
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list
archive
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Manu Abraham
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options
Manu Abraham
2018-11-15 17:54:28 UTC
Permalink
Actually, a quick search yielded me the following, which is where that
PSU is used:
https://fccid.io/ANATEL/03368-13-03257/Manual/E55D6368-99CD-4125-9693-256F6FCC50A1/PDF

Which states:

Networked monitoring TP482000B-L20B1
supports remote operations and software upgrades over a Web UI

Searching again, led me to the vendor support web site. So, they have
some sort of support for it.
Only thing is, you need to ask them the Right Thing, the Right Way, at
the Right Time.

e.huawei.com/ae/account/login
or
goo.gl/utqHnq

So, definitely, you need to create an account with them and eventually
talk to them.
Let us know of your success on the topic.

Hope it helps.

Cheers,
Manu
Post by Manu Abraham
From your post, it was difficult to understand your english, I am by
no means a german speaker, sorry about that.
From what you wrote: "prepared 48V output", I was under the belief you
were able to get a 48V output. With your latest post, I understood
incorrectly, what you tried to imply there.
If you want to talk to the vendor, the best that you could ask them to
provide the software to configure the power supply. Couldn't be any
simpler.
I doubt, even if anyone from the vendor were to be on the list,
providing a closed protocol of a product, would be the last thing
anyone would remotely try to do.
Why don't you get the actual necessary software from the vendor and
use it directly ? That would be the simplest to do. If really
necessary, you can snoop on the communication to get the basic command
infrastructure to work on some bare metal platform.
There is talk about secure CAN in the industry, so that vendors can
lock people out, or IOW keep their automobiles safer, another way you
look at it. A coin has two sides to it; not one.
That said, a PSU controller is unlikely to be a device with large
amounts of memory, given the cost of the PSU, it is unlikely that way
either. The lesser memory implies that the amount of parsing /
tokenizing would be just minimal. More tokens imply more storage of
words. Most likely, you could open the power supply, have a look at
what controller it uses and so on. If you are lucky, you could even
find that controller datasheet and get a fair understanding of what
happens under the hood inside that PSU. The next good thing would be
to know, what the PSU was really intended for. Then you could try to
gain info, how it is run in that context. Please note, that all this
takes time and and times, could be frustrating. I don't think, there's
an easy way out.
If I were you, I would've got the actual software/platform, snooped on
the bus, logged each command, to be later utilized into a dictionary
for RE. the comms. syntax. I think it would be a very interesting
project. If you've used Linux and Digital Video hardware, probably you
would've known. Most of the Linux DVB drivers were written that way.
The rev. engg'd stuff would be then compared with actual datasheets
(if available) and the quality of the driver improved eventually. ;-)
It's painful, but if you enjoy doing such things, very much an
enjoyable project. It will take you a long way. In the process, you
will be giving away knowledge to other people in the same way too.
Let us people on this list know, how you get along. Not that anyone
could help, due to lack of time, hardware and the list could go on ..
But you could be of help to someone who comes along later on another
day like this.
Cheers,
Manu
Post by Tobias Gogolin
Ok very true, wonder what the chances are to convince the vendor how
conveniently the demand for his product would increase if the
communications protocol to his device was 'leaked' or even outright made
public! We wouldn't by any chance have some inside contact to Huawei in
this fraternity?
I found a forum post where some ham operators where talking about the same
issue but apparently he had been fortunate enough to get one configured so
it would start up by itself ( http://forum.db3om.de/ftopic24798.html )
Alternatively if this first option fails and Huawei doesn't play nice would
it be realistic to hope that if somebody connected a CAN sniffer and I (or
somebody even more prepped for the task) somehow got a dump of a
configuration session with the device it might be possible to deduct the
important parts of the configuration protocol? I seen similar things being
done to try to open up an after market for Tesla Motors components...
What are the chances that if I find the device to has a command shell type
of terminal mode via CAN?
Unlikely they would have some kind of security encryption scheme for this
type of device?
Thanks Manu for the valuable links! I definitely want to learn everything
there is and start using the bus on my own vehicle designs, so I ordered
more CAN modules too...
Post by Manu Abraham
This is exactly what I've asked; How you are sending the commands. It
would be meaningless to say that you are using a CAN
transceiver/micro-controller combo to do so. Obviously, without which
you cannot do so.
My question was exactly what are you command sequences or based on
what knowledge are you sending those commands, since I could not see
any on the manual. If you are asking about the protocol
implementation/command structure how the micro controller interprets
the CAN messages, that's done at the firmware level on the
micro-controller. It could vary from case to case. Only the vendor can
help you on protocol commands for bus monitoring/control. I don't
think there would be a one standard command for all PSU's. Usually,
the PSU vendor would have their own application to interface with the
communication bus. In some corner cases, some vendors might share with
you the commands that interface to the micro-controller.
To get a basic idea of the interface, you can have a look at this.
http://www.keil.com/appnotes/files/apnt_236.pdf
If you want to try hacking the command/protocol structure/interface,
maybe you might find a starting point here.
https://www.blackhat.com/docs/asia-15/materials/asia-15-Evenchick-Hopping-On-The-Can-Bus.pdf
The best would be to contact the vendor, if you are looking for a
hassle free path.
Cheers,
Manu
Post by Tobias Gogolin
Well yes, I have an Arduino CAN bus shield (MCP2515) with selectable
termination. It's not even wired yet, because I wouldn't know what to
listen for, or how to talk to - saying what?
I am not sure does CAN bus have some kind of discovery protocol?
Post by Manu Abraham
CAN bus commands are simple. It uses a mailbox approach. Generally
power supplies use SMBUS (I2C variant) on PC motherboards, PMBUS
(Serial variant) on larger boards or interconnects, CANBUS for large
backplanes. The CANBUS usage can be understood, as the PSU could be
controlled over still longer wires.
That said, how are you sending the commands ?
Please note that CAN Bus requires a termination of 120 ohms IIRC. If
not, there will be echoes on the bus and communication issues.
Looking at your SMPS manual, the GREEN led states everything is fine.
Flashing ORANGE/YELLOW states that bad CAN BUS communication. You
might need to look at the BUS termination, probably. As I said
earlier, the SMPS is a network/telecom oriented SMPS, I guess in this
situation, I might guess it is more intended for PoE applications. Now
that 48V is the standard. When you are setting to anything more than
48V, that is considered overvoltage, which is what the manual states
too.
Not surprising isn't it ?
Cheers,
Manu
Post by Tobias Gogolin
Meanwhile esteemed Wizards (danke Russel...)
I did some tests, and while I got no power out yet I found a
following
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
If I power up with no prepared voltage on the output, just the green
LED
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
comes on
If I then power up 48 or 53V the Orange LED starts flashing, and the
Red
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
one comes on
If I power up with prepared 48V on the output, the green LED comes
on and
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
the orange flashes
If I then remove the prepared input the output just goes down too
and the
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
red LED comes on
If I power up with prepared 53 or 58V on the output, the green LED
comes
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
alone the first few times (now the orange flashes
always too) If I then remove the prepared input the output just goes
down
Post by Tobias Gogolin
too and the red LED comes on
any voltage change while on is attested with the red coming on...
So I gather there is no other than having to protocol via CAN bus?
Post by Richard Prosser
Yes, it could be that it requires an output voltage before it
starts as
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
part of the startup. If so, it can probably be disabled once you
know
Post by Tobias Gogolin
Post by Manu Abraham
the
Post by Tobias Gogolin
Post by Richard Prosser
correct commands (as you note).
The precharge is used so that it can be hot plugged into a '48V'
bus.
Post by Tobias Gogolin
Post by Manu Abraham
There
Post by Tobias Gogolin
Post by Richard Prosser
are likely to be output capacitors and without precharge the input
current
Post by Tobias Gogolin
Post by Richard Prosser
surge will damage the contacts. The fact that the precharge is on
the
Post by Tobias Gogolin
Post by Manu Abraham
neg.
Post by Tobias Gogolin
Post by Richard Prosser
rail is another clue that it's a telecom charger - most telecom
circuits
Post by Tobias Gogolin
Post by Richard Prosser
are positive earth to minimise corrosion damage.
i guess you know that fast charging virtually any sort of battery
will
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
reduce its life.
RP
Post by Tobias Gogolin
Thanks Richard!
If it is supposed to be a charger maybe it expects a voltage to
begin
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
with
Post by Tobias Gogolin
to add its juice to? Similar to a cell phone battery protection
circuit
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
that may need to be woken up by charging its internal battery
beyond
Post by Tobias Gogolin
Post by Manu Abraham
a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
certain threshold if the battery has been sitting for too long,
this
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
supply
Post by Tobias Gogolin
may not dare to power up what appears to be a dead battery?
Could be
Post by Tobias Gogolin
Post by Manu Abraham
an
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
option set within some internal config register though... I could
test
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
giving it some voltage even before powering up...
by the way there is a pre-charge pin, which although not
indicated
Post by Tobias Gogolin
Post by Manu Abraham
must
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
belong to the negative terminal of the output DC because that
negative
Post by Tobias Gogolin
Post by Richard Prosser
PCB
Post by Tobias Gogolin
contact is recessed as if for hot plugging, I doubt it has any
significance
Post by Tobias Gogolin
in regular operation tough?
The green LED is the one that remains on, but if I probe around
and
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
bridge
Post by Tobias Gogolin
that ~1 Volt feeler voltage that Voltage goes zero and the red
LED
Post by Tobias Gogolin
Post by Manu Abraham
comes
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
on.
I have put it in a case with a DPS5020 50V 20A digital power
supply
Post by Tobias Gogolin
Post by Manu Abraham
to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
provide a side function while I develop my own charge pump and
controller.
Post by Tobias Gogolin
So there would be a capacitive load if the supply was to do a
test
Post by Tobias Gogolin
Post by Manu Abraham
run
Post by Tobias Gogolin
Post by Richard Prosser
just
Post by Tobias Gogolin
like that... I thought about trying Huawei directly, so
registering
Post by Tobias Gogolin
Post by Manu Abraham
is a
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
good idea for sure!
On my current fattest battery 3000W would be almost 2C, so its a
exciting -
Post by Tobias Gogolin
if I managed to direct that much power some day - the idea to try
considering 10-15 minutes stops, basically ending if any of the
banks
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
gets
Post by Tobias Gogolin
even close to a save voltage limit...
On Mon, Nov 12, 2018 at 9:31 PM Richard Prosser <
Post by Richard Prosser
Hi Tobias,
It looks like a standard Telecom type battery charger. From the
datasheet I
Post by Richard Prosser
don't see anything indicating that it needs comms to startup
but
Post by Tobias Gogolin
Post by Manu Abraham
you
Post by Tobias Gogolin
Post by Richard Prosser
may
Post by Tobias Gogolin
Post by Richard Prosser
need comms to adjust the voltage & current limit to your
requirement.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
It's
Post by Richard Prosser
probably preset to 53.5V or so with maximum current & power
limit
Post by Tobias Gogolin
Post by Manu Abraham
etc.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
What Leds are showing? - a single green would appear to
indicate
Post by Tobias Gogolin
Post by Manu Abraham
normal
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
operation.
What happens with a resistive load (eg 4x12V automotive bulbs
in
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
series,
Post by Tobias Gogolin
a
Post by Richard Prosser
220V incandecent lamp or a heater element)? With 3kW output
capability
Post by Tobias Gogolin
Post by Richard Prosser
it
Post by Tobias Gogolin
Post by Richard Prosser
should be able to provide 50A or so current. Note that the
output
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
protection is foldback type so short circuit current will be
less
Post by Tobias Gogolin
Post by Manu Abraham
(~25A
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
from the datasheet).
Re the CAN bus, I can't help. We used to use RS485 comms on our
rectifiers
Post by Richard Prosser
(otherwise pretty similar). Probably there's a data sheet
somewhere on
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
the
Post by Richard Prosser
web giving info on what's required for this or a similar
product.
Post by Tobias Gogolin
Post by Manu Abraham
If
Post by Tobias Gogolin
Post by Richard Prosser
you
Post by Tobias Gogolin
Post by Richard Prosser
can find info on the commands for a similar but non-identical
Huawei
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
rectifier it may assist as companies tend to retain the same
protocol &
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
commands etc.. I don't think there's a Telecom standard for
this
Post by Tobias Gogolin
Post by Manu Abraham
but
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
there
Post by Richard Prosser
may be something these days. Or the commands may be encrypted
in
Post by Tobias Gogolin
Post by Manu Abraham
some
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
way.
Post by Richard Prosser
:-( in which case hardware adjustment may be the only option
if you
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
need
Post by Tobias Gogolin
to
Post by Richard Prosser
change the voltage.
Once you register on the Huawei website there seems to be a
lot of
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
datasheets available. It just may take a while to find what you
need.
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
RP
On Mon, 12 Nov 2018 at 23:32, Tobias Gogolin <
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2
rectifier as
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
base
Post by Tobias Gogolin
Post by Richard Prosser
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the
published
Post by Tobias Gogolin
Post by Manu Abraham
pdf
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs
some
Post by Tobias Gogolin
Post by Manu Abraham
CAN
Post by Tobias Gogolin
Post by Richard Prosser
bus
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
hacking before it even releases its awesome power! I
published
Post by Tobias Gogolin
Post by Manu Abraham
in the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
thread that helped me to find this device including a look
inside the
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
really sophisticated design of this Access Power Rectifier
here.
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is
alive
Post by Tobias Gogolin
Post by Richard Prosser
and
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
boots up and even puts a small test voltage out, and the 3
status LED
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I
already
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
got
Post by Richard Prosser
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know
what to
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen,
die uns
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
im
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list
archive
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list
archive
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Post by Richard Prosser
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Richard Prosser
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht
allmählich so
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von
Thor
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
Post by Richard Prosser
Post by Tobias Gogolin
Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der
Augenblick
Post by Tobias Gogolin
Post by Manu Abraham
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die
uns im
Post by Tobias Gogolin
Post by Manu Abraham
Post by Tobias Gogolin
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/
Manu Abraham
2018-11-15 19:34:22 UTC
Permalink
Following Bob's post, noticed that I had forgotten to trim my previous posts ..

Got a bit more interested to dig deeper into it.

The PSU is additionally used here as well:

goo.gl/hGL4Nc
goo.gl/RP8HRT

Scrolling down, It states the following additionally:

2. Technology support
For all the EPON and terminal units, Ethernet switch, and other
products, HUANET promises:
Technology support for life: Free call and E-mail technology supports
for each sold network products for life.
Software upgrade for life: Provide upgrade software in the form of
E-mail or CD for clients and clients can download it in customer
service web

The CAN bus is connected to a interface module which interfaces
between Ethernet ports and the CAN bus.
The Ethernet port seems to be accessed a web UI to control/monitor the PSU
https://support.huawei.com/enterprise/en/doc/EDOC1100003664
You need to register with their support site.

If nothing works, you can buy that and ask them to support it with
their Web UI. ;-)
Hopefully, if everything goes well.
Just the cost of 2 PSU's. After that, you can fiddle around to your
heart's content.

Hope that helps.

Cheers,
Manu
Post by Manu Abraham
Actually, a quick search yielded me the following, which is where that
https://fccid.io/ANATEL/03368-13-03257/Manual/E55D6368-99CD-4125-9693-256F6FCC50A1/PDF
Networked monitoring TP482000B-L20B1
supports remote operations and software upgrades over a Web UI
Searching again, led me to the vendor support web site. So, they have
some sort of support for it.
Only thing is, you need to ask them the Right Thing, the Right Way, at
the Right Time.
e.huawei.com/ae/account/login
or
goo.gl/utqHnq
So, definitely, you need to create an account with them and eventually
talk to them.
Let us know of your success on the topic.
Hope it helps.
Cheers,
Manu
Post by Manu Abraham
From your post, it was difficult to understand your english, I am by
no means a german speaker, sorry about that.
From what you wrote: "prepared 48V output", I was under the belief you
were able to get a 48V output. With your latest post, I understood
incorrectly, what you tried to imply there.
If you want to talk to the vendor, the best that you could ask them to
provide the software to configure the power supply. Couldn't be any
simpler.
I doubt, even if anyone from the vendor were to be on the list,
providing a closed protocol of a product, would be the last thing
anyone would remotely try to do.
Why don't you get the actual necessary software from the vendor and
use it directly ? That would be the simplest to do. If really
necessary, you can snoop on the communication to get the basic command
infrastructure to work on some bare metal platform.
There is talk about secure CAN in the industry, so that vendors can
lock people out, or IOW keep their automobiles safer, another way you
look at it. A coin has two sides to it; not one.
That said, a PSU controller is unlikely to be a device with large
amounts of memory, given the cost of the PSU, it is unlikely that way
either. The lesser memory implies that the amount of parsing /
tokenizing would be just minimal. More tokens imply more storage of
words. Most likely, you could open the power supply, have a look at
what controller it uses and so on. If you are lucky, you could even
find that controller datasheet and get a fair understanding of what
happens under the hood inside that PSU. The next good thing would be
to know, what the PSU was really intended for. Then you could try to
gain info, how it is run in that context. Please note, that all this
takes time and and times, could be frustrating. I don't think, there's
an easy way out.
If I were you, I would've got the actual software/platform, snooped on
the bus, logged each command, to be later utilized into a dictionary
for RE. the comms. syntax. I think it would be a very interesting
project. If you've used Linux and Digital Video hardware, probably you
would've known. Most of the Linux DVB drivers were written that way.
The rev. engg'd stuff would be then compared with actual datasheets
(if available) and the quality of the driver improved eventually. ;-)
It's painful, but if you enjoy doing such things, very much an
enjoyable project. It will take you a long way. In the process, you
will be giving away knowledge to other people in the same way too.
Let us people on this list know, how you get along. Not that anyone
could help, due to lack of time, hardware and the list could go on ..
But you could be of help to someone who comes along later on another
day like this.
Cheers,
Manu
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Tobias Gogolin
2018-11-16 00:39:17 UTC
Permalink
Thanks again Manu

I have searched a lot and entered into the world of these communication
power supply cabinets

I found that they usually contain LiFePo batteries, and the supplies can
charge as well as supply the continuous operational current. With half a
million hours MTBF (
https://www.cdiweb.com/datasheets/huawei/R4850G2%20Rectifier%20Data%20Sheet%2005-.pdf
) they are build like y should want to build them and they are probably
worth every dime of the hundreds of monetary units they usually cost! That
I got mine for €60 inc. shipping could get me suspicious that it might be a
reject, the ebay seller specializes on server components clearance and he
has no reliable access to get that deal again...
The only other supply in that power range I have found so far (a HP ESP120
supply mentioned in the German HAM thread I posted about) is huge and
weighs 4 times as much! No wonder there is a lot of interest in this little
Huawei miracle!

I am going to advance on my CAN Arduino setup and it might already exist or
its probably easy enough to write a sniffer that logs onto an SD card. Then
if I find somebody to send my supply to to try it in its original
environment, i might integrate the sniffer to log everything happening on
its bus.
By the way that reminds me there are only 2 signal wires (CAN +/-)
specified on the pinout I found but in reality there are 4 fingers in that
section, leaving 2 unknown ones...

Chances are somebody already hacked the protocol, just that it is hidden
away in some Russian or Chinese forum and it will just take time till
somebody finds it and brings it into this language space...
For example I found about the System Management Unit SMU02B via the videos
of this Alejandro (si entiendo espaniol ;)
https://www.youtube.com/results?search_query=huawei+smu02b
He was kind enough to respond but he doubts that there is a way to operate
the rectifier outside its usual nest of controllers but he knows about
peoples interest in his videos because I am not the only one asking these
questions...
Post by Manu Abraham
Following Bob's post, noticed that I had forgotten to trim my previous posts ..
Got a bit more interested to dig deeper into it.
goo.gl/hGL4Nc
goo.gl/RP8HRT
2. Technology support
For all the EPON and terminal units, Ethernet switch, and other
Technology support for life: Free call and E-mail technology supports
for each sold network products for life.
Software upgrade for life: Provide upgrade software in the form of
E-mail or CD for clients and clients can download it in customer
service web
The CAN bus is connected to a interface module which interfaces
between Ethernet ports and the CAN bus.
The Ethernet port seems to be accessed a web UI to control/monitor the PSU
https://support.huawei.com/enterprise/en/doc/EDOC1100003664
You need to register with their support site.
If nothing works, you can buy that and ask them to support it with
their Web UI. ;-)
Hopefully, if everything goes well.
Just the cost of 2 PSU's. After that, you can fiddle around to your
heart's content.
Hope that helps.
Cheers,
Manu
Post by Manu Abraham
Actually, a quick search yielded me the following, which is where that
https://fccid.io/ANATEL/03368-13-03257/Manual/E55D6368-99CD-4125-9693-256F6FCC50A1/PDF
Post by Manu Abraham
Networked monitoring TP482000B-L20B1
supports remote operations and software upgrades over a Web UI
Searching again, led me to the vendor support web site. So, they have
some sort of support for it.
Only thing is, you need to ask them the Right Thing, the Right Way, at
the Right Time.
e.huawei.com/ae/account/login
or
goo.gl/utqHnq
So, definitely, you need to create an account with them and eventually
talk to them.
Let us know of your success on the topic.
Hope it helps.
Cheers,
Manu
Post by Manu Abraham
From your post, it was difficult to understand your english, I am by
no means a german speaker, sorry about that.
From what you wrote: "prepared 48V output", I was under the belief you
were able to get a 48V output. With your latest post, I understood
incorrectly, what you tried to imply there.
If you want to talk to the vendor, the best that you could ask them to
provide the software to configure the power supply. Couldn't be any
simpler.
I doubt, even if anyone from the vendor were to be on the list,
providing a closed protocol of a product, would be the last thing
anyone would remotely try to do.
Why don't you get the actual necessary software from the vendor and
use it directly ? That would be the simplest to do. If really
necessary, you can snoop on the communication to get the basic command
infrastructure to work on some bare metal platform.
There is talk about secure CAN in the industry, so that vendors can
lock people out, or IOW keep their automobiles safer, another way you
look at it. A coin has two sides to it; not one.
That said, a PSU controller is unlikely to be a device with large
amounts of memory, given the cost of the PSU, it is unlikely that way
either. The lesser memory implies that the amount of parsing /
tokenizing would be just minimal. More tokens imply more storage of
words. Most likely, you could open the power supply, have a look at
what controller it uses and so on. If you are lucky, you could even
find that controller datasheet and get a fair understanding of what
happens under the hood inside that PSU. The next good thing would be
to know, what the PSU was really intended for. Then you could try to
gain info, how it is run in that context. Please note, that all this
takes time and and times, could be frustrating. I don't think, there's
an easy way out.
If I were you, I would've got the actual software/platform, snooped on
the bus, logged each command, to be later utilized into a dictionary
for RE. the comms. syntax. I think it would be a very interesting
project. If you've used Linux and Digital Video hardware, probably you
would've known. Most of the Linux DVB drivers were written that way.
The rev. engg'd stuff would be then compared with actual datasheets
(if available) and the quality of the driver improved eventually. ;-)
It's painful, but if you enjoy doing such things, very much an
enjoyable project. It will take you a long way. In the process, you
will be giving away knowledge to other people in the same way too.
Let us people on this list know, how you get along. Not that anyone
could help, due to lack of time, hardware and the list could go on ..
But you could be of help to someone who comes along later on another
day like this.
Cheers,
Manu
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------

Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu
Manu Abraham
2018-11-16 04:59:23 UTC
Permalink
AFAICS, I see only 2 pins. :-) It is also marked very clearly. You
need to look very carefully.

Cheers,
Manu
Post by Tobias Gogolin
By the way that reminds me there are only 2 signal wires (CAN +/-)
specified on the pinout I found but in reality there are 4 fingers in that
section, leaving 2 unknown ones...
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Tobias Gogolin
2018-11-17 23:54:39 UTC
Permalink
You have one of those units? while there are potential pin ranges on both
sides only 4 physical contact fingers (PCB) exist on mine leaving 2 that
are not identified on the pinout sheet. I'll try tracing them when i open
the unit next. Alejandro from the videos stated that possibly a minimal
configuration is just the backplane, he also voiced the hypothesis that
Huawei might have added meassures to discurrage aftermarket uses to avoid
robbery from their installations...
Post by Manu Abraham
AFAICS, I see only 2 pins. :-) It is also marked very clearly. You
need to look very carefully.
Cheers,
Manu
Post by Tobias Gogolin
By the way that reminds me there are only 2 signal wires (CAN +/-)
specified on the pinout I found but in reality there are 4 fingers in
that
Post by Tobias Gogolin
section, leaving 2 unknown ones...
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Manu Abraham
2018-11-18 08:08:31 UTC
Permalink
No, I do not have one of those, but the datasheet appears to be
precisely clear on that fact. Unable to see the source of your
contention.

Cheers,
Manu
Post by Tobias Gogolin
You have one of those units? while there are potential pin ranges on both
sides only 4 physical contact fingers (PCB) exist on mine leaving 2 that
are not identified on the pinout sheet. I'll try tracing them when i open
the unit next. Alejandro from the videos stated that possibly a minimal
configuration is just the backplane, he also voiced the hypothesis that
Huawei might have added meassures to discurrage aftermarket uses to avoid
robbery from their installations...
Post by Manu Abraham
AFAICS, I see only 2 pins. :-) It is also marked very clearly. You
need to look very carefully.
Cheers,
Manu
Post by Tobias Gogolin
By the way that reminds me there are only 2 signal wires (CAN +/-)
specified on the pinout I found but in reality there are 4 fingers in
that
Post by Tobias Gogolin
section, leaving 2 unknown ones...
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Tobias Gogolin
2018-11-23 14:03:43 UTC
Permalink
Looks like i have a Hong Kong Huawei specialist who has a CAN box to
program these and has clients using Huawei rectifiers up to hundrets of kW.
So far no confirmation they sell the CAN devices... Updated forum
Post by Manu Abraham
No, I do not have one of those, but the datasheet appears to be
precisely clear on that fact. Unable to see the source of your
contention.
Cheers,
Manu
Post by Tobias Gogolin
You have one of those units? while there are potential pin ranges on both
sides only 4 physical contact fingers (PCB) exist on mine leaving 2 that
are not identified on the pinout sheet. I'll try tracing them when i open
the unit next. Alejandro from the videos stated that possibly a minimal
configuration is just the backplane, he also voiced the hypothesis that
Huawei might have added meassures to discurrage aftermarket uses to avoid
robbery from their installations...
Post by Manu Abraham
AFAICS, I see only 2 pins. :-) It is also marked very clearly. You
need to look very carefully.
Cheers,
Manu
Post by Tobias Gogolin
By the way that reminds me there are only 2 signal wires (CAN +/-)
specified on the pinout I found but in reality there are 4 fingers in
that
Post by Tobias Gogolin
section, leaving 2 unknown ones...
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Tobias Gogolin
2018-12-02 14:39:34 UTC
Permalink
Hello friends,
I know that at least those who contributed to this challenge so far are
probably going to be happy with me that the milestone 'operating the
RG4850G2 has been accomplished! I attribute the 'ease' of this achievement
to the help of some users of the Endless-Sphere forum! Mind you so far the
achievement is knowing what it needs to turn on (pulling those two until
recently unknown pin to V- = GND).
We have also located a Chinese maker that has build and markets a simple
CAN calibrator, but there are also users that have declared the intention
to create an open source implementation and seem to know how to get there.
All of this has been documented in the forum and if you like to look it up
(and more exciting part of the larger project) the link via
https://www.patreon.com/SunTripToGo should be good ...
Happy Sunday...
Post by Tobias Gogolin
Looks like i have a Hong Kong Huawei specialist who has a CAN box to
program these and has clients using Huawei rectifiers up to hundrets of kW.
So far no confirmation they sell the CAN devices... Updated forum
Post by Manu Abraham
No, I do not have one of those, but the datasheet appears to be
precisely clear on that fact. Unable to see the source of your
contention.
Cheers,
Manu
Post by Tobias Gogolin
You have one of those units? while there are potential pin ranges on
both
Post by Tobias Gogolin
sides only 4 physical contact fingers (PCB) exist on mine leaving 2 that
are not identified on the pinout sheet. I'll try tracing them when i
open
Post by Tobias Gogolin
the unit next. Alejandro from the videos stated that possibly a minimal
configuration is just the backplane, he also voiced the hypothesis that
Huawei might have added meassures to discurrage aftermarket uses to
avoid
Post by Tobias Gogolin
robbery from their installations...
Post by Manu Abraham
AFAICS, I see only 2 pins. :-) It is also marked very clearly. You
need to look very carefully.
Cheers,
Manu
Post by Tobias Gogolin
By the way that reminds me there are only 2 signal wires (CAN +/-)
specified on the pinout I found but in reality there are 4 fingers
in
Post by Tobias Gogolin
Post by Manu Abraham
that
Post by Tobias Gogolin
section, leaving 2 unknown ones...
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------

Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/m
Manu Abraham
2018-11-13 08:54:32 UTC
Permalink
That's no battery charger. That's a standard SMPS of a Redundant Power
Supply class commonly found in servers with the Backplane class or
telecommunication/network equipment. The CAN bus is simply meant to
monitor/fine tweak the PWM output. I doubt you will be able to vary
much output except for a few volts. Looked at the thread that you
posted, it states that you can vary about 10V. The PCB edge that you
see on the power supply nicely goes and sits into a card slot, much
like how you put add on cards into your PC.

Looking further into that thread, it appears that it is a Hot Swap RPS
class SMPS. By default, it appears that the output is set to 48V. Have
a similar RPS on one of my stacked network switches, though that power
supply smaller in size and is slightly of a lower rating. You can have
2 power supplies on the switch: pull one of them while the switch is
still running. It is meant for very high uptime equipment where a dead
power supply should not take the entire network down.

Cheers,

Manu
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as base for
a quick charger for my electric trike. Regrettably the published pdf about
its operation are not very in-depth and I assume it needs some CAN bus
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive and
boots up and even puts a small test voltage out, and the 3 status LED tell
that its alive, but it must want some configuration still. I already got a
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick wo
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor Heyerdahl
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/
Sean Breheny
2018-11-13 22:22:07 UTC
Permalink
Regarding hot swap power supplies. I was once in a server room at work
because I had to help spec out power drops and wiring for some servers and
UPSes. I had nothing to do with the software side of the server operation.

I needed to get the model number off the label on one of the two redundant
supplies in this running server. The label could only be seen if the supply
was outside of the server chassis. I figured that since it was redundant I
could pull it out while the server was running.

Sure enough, I pulled it out and the server kept running - the only change
being that the fan speed on the remaining supply increased since it had to
handle the entire load by itself.

I copied the model number and put the power supply back in. As soon as it
made full contact with the backplane, I heard the loud snap of a high
current arc (presumably the working power supply suddenly charging the
output caps on the ps that I was plugging-in) and the status LED on the
working supply went into the red faulted condition and the server shut
down. It promptly rebooted and both supplies were now working again.

I quietly left the room and closed the door, feeling like Charlie Chaplain
in the scene where he gets a job at a shipyard and removes the block of
wood which was preventing the incomplete ship from sliding down into the
water.

The moral of the story is that these supplies do handle the situation where
a supply is removed or fails open, but not necessarily true hot swap.
Post by Manu Abraham
That's no battery charger. That's a standard SMPS of a Redundant Power
Supply class commonly found in servers with the Backplane class or
telecommunication/network equipment. The CAN bus is simply meant to
monitor/fine tweak the PWM output. I doubt you will be able to vary
much output except for a few volts. Looked at the thread that you
posted, it states that you can vary about 10V. The PCB edge that you
see on the power supply nicely goes and sits into a card slot, much
like how you put add on cards into your PC.
Looking further into that thread, it appears that it is a Hot Swap RPS
class SMPS. By default, it appears that the output is set to 48V. Have
a similar RPS on one of my stacked network switches, though that power
supply smaller in size and is slightly of a lower rating. You can have
2 power supplies on the switch: pull one of them while the switch is
still running. It is meant for very high uptime equipment where a dead
power supply should not take the entire network down.
Cheers,
Manu
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as base
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published pdf
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some CAN bus
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive and
boots up and even puts a small test voltage out, and the 3 status LED
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I already got
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http:/
Manu Abraham
2018-11-14 06:08:20 UTC
Permalink
Hi Sean,

One of my experiences was exactly like that. Only at that point that I
realized that vendors advertised RPS which is a different thing,
definitely they handle the power supply failure, then there is another
category which states Hot Swap RPS. Asking vendors then about the
issue, most of them stated that almost all HS RPS state that, not all
are true Hot swappable.

I was intrigued those days what caused the shutdown, like what you
stated. In my mind,blamed the power supply designers. Years down the
lane, came across articles from vendors and manufacturers on the
topic.

An interesting read:
http://www.ti.com/lit/an/slva673a/slva673a.pdf

The newer HS power supplies do seem to work as expected, they seem to have read:

https://pdfserv.maximintegrated.com/en/an/AN5807.pdf

Talk about you feeling like Charlie .. My experience, long back with a
server with an Intel L440GX motherboard being built with everything
Hot swap. We went on so far as to go for a true hardware RAID adapter
instead of the firmware based RAID on the L440GX. The Adaptec had
RAID0 for high availablity, configured with RAID 1 mirroring, for the
highest up time. The Hard disks were Fujitsu Ultra 160 SCSI 9GB drives
on bays that could be ejected pressing the Eject button. Everything
was running perfect with about 8 users on the network, the server
running ipchains and had a DMZ. We were happy that everything we had,
had Redundant marked on it. After building everything the eject
buttons did look cute enough. It did want some action, for live eject.
Anyway, what could go wrong, right ? Pressed the button, the HDD
smoothly ejected.

But the system froze. Rebooting the system after that, it had a Linux
ext2 filesystem. The filesystem got corrupted too, fsck at boot time
wept very badly citing many missing entries. The entire installation
process had to be redone. So much for the hot swap button! Wondered,
who designed such Hot swap hardware. I guess those days, it was all
about pure terminology going around as in marketing of the
terminology, rather than the true action of the properties of the
terminology.

Cheers,

Manu
Post by Sean Breheny
Regarding hot swap power supplies. I was once in a server room at work
because I had to help spec out power drops and wiring for some servers and
UPSes. I had nothing to do with the software side of the server operation.
I needed to get the model number off the label on one of the two redundant
supplies in this running server. The label could only be seen if the supply
was outside of the server chassis. I figured that since it was redundant I
could pull it out while the server was running.
Sure enough, I pulled it out and the server kept running - the only change
being that the fan speed on the remaining supply increased since it had to
handle the entire load by itself.
I copied the model number and put the power supply back in. As soon as it
made full contact with the backplane, I heard the loud snap of a high
current arc (presumably the working power supply suddenly charging the
output caps on the ps that I was plugging-in) and the status LED on the
working supply went into the red faulted condition and the server shut
down. It promptly rebooted and both supplies were now working again.
I quietly left the room and closed the door, feeling like Charlie Chaplain
in the scene where he gets a job at a shipyard and removes the block of
wood which was preventing the incomplete ship from sliding down into the
water.
The moral of the story is that these supplies do handle the situation where
a supply is removed or fails open, but not necessarily true hot swap.
Post by Manu Abraham
That's no battery charger. That's a standard SMPS of a Redundant Power
Supply class commonly found in servers with the Backplane class or
telecommunication/network equipment. The CAN bus is simply meant to
monitor/fine tweak the PWM output. I doubt you will be able to vary
much output except for a few volts. Looked at the thread that you
posted, it states that you can vary about 10V. The PCB edge that you
see on the power supply nicely goes and sits into a card slot, much
like how you put add on cards into your PC.
Looking further into that thread, it appears that it is a Hot Swap RPS
class SMPS. By default, it appears that the output is set to 48V. Have
a similar RPS on one of my stacked network switches, though that power
supply smaller in size and is slightly of a lower rating. You can have
2 power supplies on the switch: pull one of them while the switch is
still running. It is meant for very high uptime equipment where a dead
power supply should not take the entire network down.
Cheers,
Manu
Post by Tobias Gogolin
Hello estimated electronics wizards.
I am exploring to use a very cool 3kW Huawei R4850G2 rectifier as base
for
Post by Tobias Gogolin
a quick charger for my electric trike. Regrettably the published pdf
about
Post by Tobias Gogolin
its operation are not very in-depth and I assume it needs some CAN bus
hacking before it even releases its awesome power! I published in the
thread that helped me to find this device including a look inside the
really sophisticated design of this Access Power Rectifier here.
https://endless-sphere.com/forums/viewtopic.php?f=14&t=86038
I appreciate any hints of how to proceed. I know the unit is alive and
boots up and even puts a small test voltage out, and the 3 status LED
tell
Post by Tobias Gogolin
that its alive, but it must want some configuration still. I already got
a
Post by Tobias Gogolin
MCP2515 CAN bus shield and Arduino ready, but I wouldn't know what to
expect or try next...
Thanks for your time!
--
„Jetzt war der Augenblick gekommen wo wir denken mussten. Der Augenblick
wo
Post by Tobias Gogolin
wir aufspringen und entscheiden mußten, wenn wir nicht allmählich so
vergiftet werden wollten, daß wir blind die Plätze einnahmen, die uns im
führerlosen Zug der Älteren angeboten wurden.“
Aus dem ersten Kapitel von 'Fatu Hiva - Zurück zur Natur' von Thor
Heyerdahl
Post by Tobias Gogolin
-------------------------------------
Tobias Gogolin
Tel. D1 (49) 151 6600 7066
Tel. D2 (49) 2672 906 9191
skype: moontogo
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailma
James Cameron
2018-11-14 06:28:16 UTC
Permalink
Post by Manu Abraham
Talk about you feeling like Charlie .. My experience, long back with a
server with an Intel L440GX motherboard being built with everything
Hot swap. We went on so far as to go for a true hardware RAID adapter
instead of the firmware based RAID on the L440GX.
Reminds me of my time in Digital, Compaq, and HP. Often needed to
support hardware RAID adapters for OpenVMS and Digital UNIX.

Customer would joyfully say they purchased hardware RAID "because
better" only for me to point out to them that the CPU on the adapter
did run firmware, and so the firmware was also a source of bugs.

Woe to the customer who did not heed our advice to take a full backup
before a firmware upgrade. ;-)
--
James Cameron
http://quozl.netrek.org/
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Manu Abraham
2018-11-14 07:48:31 UTC
Permalink
Post by James Cameron
Post by Manu Abraham
Talk about you feeling like Charlie .. My experience, long back with a
server with an Intel L440GX motherboard being built with everything
Hot swap. We went on so far as to go for a true hardware RAID adapter
instead of the firmware based RAID on the L440GX.
Reminds me of my time in Digital, Compaq, and HP. Often needed to
support hardware RAID adapters for OpenVMS and Digital UNIX.
Customer would joyfully say they purchased hardware RAID "because
better" only for me to point out to them that the CPU on the adapter
did run firmware, and so the firmware was also a source of bugs.
True indeed. Completely in sync with you on that. The reason we went
for hardware RAID in that context: The server CPU in context was a
firewall, running ipchains serving users on the LAN as well 2 servers
on the DMZ. The firewall CPU would be continuously analyzing the
network packets all the time and we wanted to offload the RAID
software overhead from the Server CPU's. We had some SGI visual
workstations on the LAN, all of them came default with the Adaptec
RAID, but none of them had the hot swap bay. ;-)

It's always a huge battle between Hardware RAID and Software RAID. At
least in a few contexts with Linux, software RAID seems to be a better
option than hardware RAID.

Cheers,
Manu
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
RussellMc
2018-11-13 12:10:09 UTC
Permalink
Post by Tobias Gogolin
Hello estimated electronics wizards.
1. Lost in translation?
or
2. Intended to be "esteemed"
or
3. A considered slur? :-) :-) :-) :-)


Russell

[Just joking][of course].

2. I assume.
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
Loading...