[EE] USB cable blues

[Art]

I recently bought an RTL-SDR
(https://www.nooelec.com/store/nesdr-smart-sdr.html), a small radio
receiver that is powered by the USB connector on the punchbox. It
digitizes the received radio signal and sends the digital information to
the punchbox for processing via USB.

It turns out that the computer generates to much noise for the receiver
to operate in close proximity to the punchbox.

So, I have to move the receiver and/or the antenna further away from the
computer. I choose to move the receiver as low loss coax for the uhf
band is pricey.

So, I bought a high end 10 foot USB 2.0 extension cable, with 2 layers
of shielding, and emi suppressors (ferrite) on each end of the cable.
The cable has 24 gauge power bus conductors, mote than adequate for
powering my RTL-SDR receiver. The RTL-SDR receiver uses 370 ma, worst
case, so the 24 gauge power conductors in the cable should be adequate.

The RTL-SDR unit has an internal (smt type) emi absorber on the power
input terminal. The computer is a very dirty place in terms of rf noise,
and having rf noise on the USB lines is like rolling out the welcome
wagon for rf interference to the RTL-SDR receiver.

The new USB cable works fine for the smaller usb flash drives I have,
but it causes the RTL-SDR device and my USB HDD unit to fail when
connected through the new cable. Neither of the higher power devices are
recognized. The RTL-SDR requires the full USB 2.0 throughput, without
it, it can't send data fast enough to enable the software in the
punchbox to function. Both the HDD and the RTL-SDR work fine when
plugged (directly) into any of the ports on the computer (without the
extension cable).

I concluded that I need to roll my own cable. That way, I know the cable
itself is proper. I can buy the usb ends for the extension cable and
solder them to the cable.

However, I can't find the technical specs for the cable, so I don't know
which type of raw cable to buy....in order to insure my homebrew cable
works at it's full capacity (which is 480 MB/sec). There are USB
repeater cables that extend the max cable length out to 50 feet, but
they're expensive, they draw a lot of power and they don't work at the
full data rate for USB 2.0. And, none of the repeater cables have
shielding at the housing for the driver electronics.

I've thought about using an ethernet cable and putting usb connectors on
the ethernet cable. But, ethernet uses balanced transmission line
(twisted pair)...it might not work well for usb because USB doesn't use
a balanced line-so there is little benefit to be had by moving to
ethernet cable for usb data transmission.

I sure wish the RTL-SDR had an ehernet cable connector instead of a USB
connector!

My problem is I cannot find the usb 2.0 specification that outlines the
critical parameters for the cable itself. Any ideas?

Sorry to be so long winded::>

Art




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

[Jason White]

Have you tried using a different USB cable? Conventional USB cables with
their single layer of shielding should be more than adequate for your
application. I suspect that manufacturing your own cable may not the
easiest, cheapest, or fastest way of getting things to work.

I have a RTL-SDR type dongle and a 125MHz upconverter. In my experience it
is noisy and strong signals produce aliasing* every 2.4MHz (sampling
frequency), every ~27MHz (SDR crystal frequency) and every 125Mhz (from the
upconverter). The internal 30-40dB AGC typically will typically boost
signals past the point where aliasing occurs. As a result AGC needs to be
disabled and the the dongle's RF gain has to be manually set - typically
never exceeding 20dB.

* aliasing -> ghost signals, lines and spikes in the SDR's decoded spectrum

-Jason White


On Sun, Jul 8, 2018 at 3:09 PM, Art <***@myfairpoint.net> wrote:

> I recently bought an RTL-SDR
> (https://www.nooelec.com/store/nesdr-smart-sdr.html), a small radio
> receiver that is powered by the USB connector on the punchbox. It
> digitizes the received radio signal and sends the digital information to
> the punchbox for processing via USB.
>
> It turns out that the computer generates to much noise for the receiver
> to operate in close proximity to the punchbox.
>
> So, I have to move the receiver and/or the antenna further away from the
> computer. I choose to move the receiver as low loss coax for the uhf
> band is pricey.
>
> So, I bought a high end 10 foot USB 2.0 extension cable, with 2 layers
> of shielding, and emi suppressors (ferrite) on each end of the cable.
> The cable has 24 gauge power bus conductors, mote than adequate for
> powering my RTL-SDR receiver. The RTL-SDR receiver uses 370 ma, worst
> case, so the 24 gauge power conductors in the cable should be adequate.
>
> The RTL-SDR unit has an internal (smt type) emi absorber on the power
> input terminal. The computer is a very dirty place in terms of rf noise,
> and having rf noise on the USB lines is like rolling out the welcome
> wagon for rf interference to the RTL-SDR receiver.
>
> The new USB cable works fine for the smaller usb flash drives I have,
> but it causes the RTL-SDR device and my USB HDD unit to fail when
> connected through the new cable. Neither of the higher power devices are
> recognized. The RTL-SDR requires the full USB 2.0 throughput, without
> it, it can't send data fast enough to enable the software in the
> punchbox to function. Both the HDD and the RTL-SDR work fine when
> plugged (directly) into any of the ports on the computer (without the
> extension cable).
>
> I concluded that I need to roll my own cable. That way, I know the cable
> itself is proper. I can buy the usb ends for the extension cable and
> solder them to the cable.
>
> However, I can't find the technical specs for the cable, so I don't know
> which type of raw cable to buy....in order to insure my homebrew cable
> works at it's full capacity (which is 480 MB/sec). There are USB
> repeater cables that extend the max cable length out to 50 feet, but
> they're expensive, they draw a lot of power and they don't work at the
> full data rate for USB 2.0. And, none of the repeater cables have
> shielding at the housing for the driver electronics.
>
> I've thought about using an ethernet cable and putting usb connectors on
> the ethernet cable. But, ethernet uses balanced transmission line
> (twisted pair)...it might not work well for usb because USB doesn't use
> a balanced line-so there is little benefit to be had by moving to
> ethernet cable for usb data transmission.
>
> I sure wish the RTL-SDR had an ehernet cable connector instead of a USB
> connector!
>
> My problem is I cannot find the usb 2.0 specification that outlines the
> critical parameters for the cable itself. Any ideas?
>
> Sorry to be so long winded::>
>
> Art
>
>
>
>
> --
> http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
>



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

[Art]

Hi Jason,

I'm after weak signals on the vhf and uhf aircraft bands, so I need all
the shielding and suppression I can get. I've never needed an extension
usb cable, so I don't have anything to test with (other than the 10 foot
cable I just bought).

I do have a 6 foot cable on order, with double shield and a ferrite on
the braid. But, if I have to spend more money, I'd rather have something
that I know is right, which is why I hoped to build my own cable and be
done with it.

I know about the RTL-SDR's shortcomings and about the need for front end
bandpass filters. There is no way it will function without them-so many
users are oblivious to the ramification of out of band signals causing
overload. And, many of those that do know about overload assume a
digital signals can't be compromised by strong out of band signals.

BTW, the RTL-SDR is in a constant state of overload, even without the
AGC and the proper 50 db gain control setting. With only 40 db dynamic
range, the RTL-SDR doesn't even need out of band signals to cause non
linear mixer responses, and the degradation of the desired signals
reception.

Ultimately, I'd like to just put the thing outside, at the mast
location, but I need to do some USB cable testing before going any
further. A fiber optic link would be neat, although it's overkill for
the 50 to 100 feet transmission distance I'd like to ultimately have::>

Any idea where I can find specs for the proper usb (raw) cable?

Regards,

Art




On 07/08/2018 03:40 PM, Jason White wrote:
> Have you tried using a different USB cable? Conventional USB cables with
> their single layer of shielding should be more than adequate for your
> application. I suspect that manufacturing your own cable may not the
> easiest, cheapest, or fastest way of getting things to work.
>
> I have a RTL-SDR type dongle and a 125MHz upconverter. In my experience it
> is noisy and strong signals produce aliasing* every 2.4MHz (sampling
> frequency), every ~27MHz (SDR crystal frequency) and every 125Mhz (from the
> upconverter). The internal 30-40dB AGC typically will typically boost
> signals past the point where aliasing occurs. As a result AGC needs to be
> disabled and the the dongle's RF gain has to be manually set - typically
> never exceeding 20dB.
>
> * aliasing -> ghost signals, lines and spikes in the SDR's decoded spectrum
>
> -Jason White
>
>
> On Sun, Jul 8, 2018 at 3:09 PM, Art <***@myfairpoint.net> wrote:
>
>> I recently bought an RTL-SDR
>> (https://www.nooelec.com/store/nesdr-smart-sdr.html), a small radio
>> receiver that is powered by the USB connector on the punchbox. It
>> digitizes the received radio signal and sends the digital information to
>> the punchbox for processing via USB.
>>
>> It turns out that the computer generates to much noise for the receiver
>> to operate in close proximity to the punchbox.
>>
>> So, I have to move the receiver and/or the antenna further away from the
>> computer. I choose to move the receiver as low loss coax for the uhf
>> band is pricey.
>>
>> So, I bought a high end 10 foot USB 2.0 extension cable, with 2 layers
>> of shielding, and emi suppressors (ferrite) on each end of the cable.
>> The cable has 24 gauge power bus conductors, mote than adequate for
>> powering my RTL-SDR receiver. The RTL-SDR receiver uses 370 ma, worst
>> case, so the 24 gauge power conductors in the cable should be adequate.
>>
>> The RTL-SDR unit has an internal (smt type) emi absorber on the power
>> input terminal. The computer is a very dirty place in terms of rf noise,
>> and having rf noise on the USB lines is like rolling out the welcome
>> wagon for rf interference to the RTL-SDR receiver.
>>
>> The new USB cable works fine for the smaller usb flash drives I have,
>> but it causes the RTL-SDR device and my USB HDD unit to fail when
>> connected through the new cable. Neither of the higher power devices are
>> recognized. The RTL-SDR requires the full USB 2.0 throughput, without
>> it, it can't send data fast enough to enable the software in the
>> punchbox to function. Both the HDD and the RTL-SDR work fine when
>> plugged (directly) into any of the ports on the computer (without the
>> extension cable).
>>
>> I concluded that I need to roll my own cable. That way, I know the cable
>> itself is proper. I can buy the usb ends for the extension cable and
>> solder them to the cable.
>>
>> However, I can't find the technical specs for the cable, so I don't know
>> which type of raw cable to buy....in order to insure my homebrew cable
>> works at it's full capacity (which is 480 MB/sec). There are USB
>> repeater cables that extend the max cable length out to 50 feet, but
>> they're expensive, they draw a lot of power and they don't work at the
>> full data rate for USB 2.0. And, none of the repeater cables have
>> shielding at the housing for the driver electronics.
>>
>> I've thought about using an ethernet cable and putting usb connectors on
>> the ethernet cable. But, ethernet uses balanced transmission line
>> (twisted pair)...it might not work well for usb because USB doesn't use
>> a balanced line-so there is little benefit to be had by moving to
>> ethernet cable for usb data transmission.
>>
>> I sure wish the RTL-SDR had an ehernet cable connector instead of a USB
>> connector!
>>
>> My problem is I cannot find the usb 2.0 specification that outlines the
>> critical parameters for the cable itself. Any ideas?
>>
>> Sorry to be so long winded::>
>>
>> Art
>>
>>
>>
>>
>> --
>> 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

[Jean-Paul Louis]

Art,

If you are after weak signals on VHF/UHF, the RTL dongle is a poor choice at best. The ADC in the dongle is at best 10 bits, so your dynamic range is limited to about 60dB before you hit the quantizing noise. If you have any strong signal in the area you live in, like FM broadcast, those might also affect your receiving signal. I do not think that adding a long USB cable is going to help. A much better solution would be to add a bandpass filter to eliminate unwanted signals, but you do not seem willing to do that. Try also to EMI shield your RTL dongle.
Your idea of using the RTL dongle near the antenna on the mast will need some creative engineering, but a long USB cable will not cut it. Your idea of a fiber link is good, but you could also try a USB to WIFI point to point link.


Just my $0.02,

Jean-Paul
N1JPL

> On Jul 8, 2018, at 4:21 PM, Art <***@myfairpoint.net> wrote:
>
> Hi Jason,
>
> I'm after weak signals on the vhf and uhf aircraft bands, so I need all
> the shielding and suppression I can get. I've never needed an extension
> usb cable, so I don't have anything to test with (other than the 10 foot
> cable I just bought).
>
> I do have a 6 foot cable on order, with double shield and a ferrite on
> the braid. But, if I have to spend more money, I'd rather have something
> that I know is right, which is why I hoped to build my own cable and be
> done with it.
>
> I know about the RTL-SDR's shortcomings and about the need for front end
> bandpass filters. There is no way it will function without them-so many
> users are oblivious to the ramification of out of band signals causing
> overload. And, many of those that do know about overload assume a
> digital signals can't be compromised by strong out of band signals.
>
> BTW, the RTL-SDR is in a constant state of overload, even without the
> AGC and the proper 50 db gain control setting. With only 40 db dynamic
> range, the RTL-SDR doesn't even need out of band signals to cause non
> linear mixer responses, and the degradation of the desired signals
> reception.
>
> Ultimately, I'd like to just put the thing outside, at the mast
> location, but I need to do some USB cable testing before going any
> further. A fiber optic link would be neat, although it's overkill for
> the 50 to 100 feet transmission distance I'd like to ultimately have::>
>
> Any idea where I can find specs for the proper usb (raw) cable?
>
> Regards,
>
> Art
>
>
>
>
> On 07/08/2018 03:40 PM, Jason White wrote:
>> Have you tried using a different USB cable? Conventional USB cables with
>> their single layer of shielding should be more than adequate for your
>> application. I suspect that manufacturing your own cable may not the
>> easiest, cheapest, or fastest way of getting things to work.
>>
>> I have a RTL-SDR type dongle and a 125MHz upconverter. In my experience it
>> is noisy and strong signals produce aliasing* every 2.4MHz (sampling
>> frequency), every ~27MHz (SDR crystal frequency) and every 125Mhz (from the
>> upconverter). The internal 30-40dB AGC typically will typically boost
>> signals past the point where aliasing occurs. As a result AGC needs to be
>> disabled and the the dongle's RF gain has to be manually set - typically
>> never exceeding 20dB.
>>
>> * aliasing -> ghost signals, lines and spikes in the SDR's decoded spectrum
>>
>> -Jason White
>>
>>
>> On Sun, Jul 8, 2018 at 3:09 PM, Art <***@myfairpoint.net> wrote:
>>
>>> I recently bought an RTL-SDR
>>> (https://www.nooelec.com/store/nesdr-smart-sdr.html), a small radio
>>> receiver that is powered by the USB connector on the punchbox. It
>>> digitizes the received radio signal and sends the digital information to
>>> the punchbox for processing via USB.
>>>
>>> It turns out that the computer generates to much noise for the receiver
>>> to operate in close proximity to the punchbox.
>>>
>>> So, I have to move the receiver and/or the antenna further away from the
>>> computer. I choose to move the receiver as low loss coax for the uhf
>>> band is pricey.
>>>
>>> So, I bought a high end 10 foot USB 2.0 extension cable, with 2 layers
>>> of shielding, and emi suppressors (ferrite) on each end of the cable.
>>> The cable has 24 gauge power bus conductors, mote than adequate for
>>> powering my RTL-SDR receiver. The RTL-SDR receiver uses 370 ma, worst
>>> case, so the 24 gauge power conductors in the cable should be adequate.
>>>
>>> The RTL-SDR unit has an internal (smt type) emi absorber on the power
>>> input terminal. The computer is a very dirty place in terms of rf noise,
>>> and having rf noise on the USB lines is like rolling out the welcome
>>> wagon for rf interference to the RTL-SDR receiver.
>>>
>>> The new USB cable works fine for the smaller usb flash drives I have,
>>> but it causes the RTL-SDR device and my USB HDD unit to fail when
>>> connected through the new cable. Neither of the higher power devices are
>>> recognized. The RTL-SDR requires the full USB 2.0 throughput, without
>>> it, it can't send data fast enough to enable the software in the
>>> punchbox to function. Both the HDD and the RTL-SDR work fine when
>>> plugged (directly) into any of the ports on the computer (without the
>>> extension cable).
>>>
>>> I concluded that I need to roll my own cable. That way, I know the cable
>>> itself is proper. I can buy the usb ends for the extension cable and
>>> solder them to the cable.
>>>
>>> However, I can't find the technical specs for the cable, so I don't know
>>> which type of raw cable to buy....in order to insure my homebrew cable
>>> works at it's full capacity (which is 480 MB/sec). There are USB
>>> repeater cables that extend the max cable length out to 50 feet, but
>>> they're expensive, they draw a lot of power and they don't work at the
>>> full data rate for USB 2.0. And, none of the repeater cables have
>>> shielding at the housing for the driver electronics.
>>>
>>> I've thought about using an ethernet cable and putting usb connectors on
>>> the ethernet cable. But, ethernet uses balanced transmission line
>>> (twisted pair)...it might not work well for usb because USB doesn't use
>>> a balanced line-so there is little benefit to be had by moving to
>>> ethernet cable for usb data transmission.
>>>
>>> I sure wish the RTL-SDR had an ehernet cable connector instead of a USB
>>> connector!
>>>
>>> My problem is I cannot find the usb 2.0 specification that outlines the
>>> critical parameters for the cable itself. Any ideas?
>>>
>>> Sorry to be so long winded::>
>>>
>>> Art
>>>
>>>
>>>
>>>
>>> --
>>> 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

[Richard Prosser]

USB pcb data traces are normally spec'd at about 90ohms differential,
45ohms single ended. Cable requirements should be the same. Series
resistance isn't likely to be much of an issue either for realistic values
as there are often 22ohm damping resistors in series. Otherwise you can
buy a single port hub on the end of a 5 or 10 metre cable and that works
better for high speed USB than a simple extension cable.
RP

On 9 July 2018 at 08:21, Art <***@myfairpoint.net> wrote:

> Hi Jason,
>
> I'm after weak signals on the vhf and uhf aircraft bands, so I need all
> the shielding and suppression I can get. I've never needed an extension
> usb cable, so I don't have anything to test with (other than the 10 foot
> cable I just bought).
>
> I do have a 6 foot cable on order, with double shield and a ferrite on
> the braid. But, if I have to spend more money, I'd rather have something
> that I know is right, which is why I hoped to build my own cable and be
> done with it.
>
> I know about the RTL-SDR's shortcomings and about the need for front end
> bandpass filters. There is no way it will function without them-so many
> users are oblivious to the ramification of out of band signals causing
> overload. And, many of those that do know about overload assume a
> digital signals can't be compromised by strong out of band signals.
>
> BTW, the RTL-SDR is in a constant state of overload, even without the
> AGC and the proper 50 db gain control setting. With only 40 db dynamic
> range, the RTL-SDR doesn't even need out of band signals to cause non
> linear mixer responses, and the degradation of the desired signals
> reception.
>
> Ultimately, I'd like to just put the thing outside, at the mast
> location, but I need to do some USB cable testing before going any
> further. A fiber optic link would be neat, although it's overkill for
> the 50 to 100 feet transmission distance I'd like to ultimately have::>
>
> Any idea where I can find specs for the proper usb (raw) cable?
>
> Regards,
>
> Art
>
>
>
>
> On 07/08/2018 03:40 PM, Jason White wrote:
> > Have you tried using a different USB cable? Conventional USB cables with
> > their single layer of shielding should be more than adequate for your
> > application. I suspect that manufacturing your own cable may not the
> > easiest, cheapest, or fastest way of getting things to work.
> >
> > I have a RTL-SDR type dongle and a 125MHz upconverter. In my experience
> it
> > is noisy and strong signals produce aliasing* every 2.4MHz (sampling
> > frequency), every ~27MHz (SDR crystal frequency) and every 125Mhz (from
> the
> > upconverter). The internal 30-40dB AGC typically will typically boost
> > signals past the point where aliasing occurs. As a result AGC needs to be
> > disabled and the the dongle's RF gain has to be manually set - typically
> > never exceeding 20dB.
> >
> > * aliasing -> ghost signals, lines and spikes in the SDR's decoded
> spectrum
> >
> > -Jason White
> >
> >
> > On Sun, Jul 8, 2018 at 3:09 PM, Art <***@myfairpoint.net> wrote:
> >
> >> I recently bought an RTL-SDR
> >> (https://www.nooelec.com/store/nesdr-smart-sdr.html), a small radio
> >> receiver that is powered by the USB connector on the punchbox. It
> >> digitizes the received radio signal and sends the digital information to
> >> the punchbox for processing via USB.
> >>
> >> It turns out that the computer generates to much noise for the receiver
> >> to operate in close proximity to the punchbox.
> >>
> >> So, I have to move the receiver and/or the antenna further away from the
> >> computer. I choose to move the receiver as low loss coax for the uhf
> >> band is pricey.
> >>
> >> So, I bought a high end 10 foot USB 2.0 extension cable, with 2 layers
> >> of shielding, and emi suppressors (ferrite) on each end of the cable.
> >> The cable has 24 gauge power bus conductors, mote than adequate for
> >> powering my RTL-SDR receiver. The RTL-SDR receiver uses 370 ma, worst
> >> case, so the 24 gauge power conductors in the cable should be adequate.
> >>
> >> The RTL-SDR unit has an internal (smt type) emi absorber on the power
> >> input terminal. The computer is a very dirty place in terms of rf noise,
> >> and having rf noise on the USB lines is like rolling out the welcome
> >> wagon for rf interference to the RTL-SDR receiver.
> >>
> >> The new USB cable works fine for the smaller usb flash drives I have,
> >> but it causes the RTL-SDR device and my USB HDD unit to fail when
> >> connected through the new cable. Neither of the higher power devices are
> >> recognized. The RTL-SDR requires the full USB 2.0 throughput, without
> >> it, it can't send data fast enough to enable the software in the
> >> punchbox to function. Both the HDD and the RTL-SDR work fine when
> >> plugged (directly) into any of the ports on the computer (without the
> >> extension cable).
> >>
> >> I concluded that I need to roll my own cable. That way, I know the cable
> >> itself is proper. I can buy the usb ends for the extension cable and
> >> solder them to the cable.
> >>
> >> However, I can't find the technical specs for the cable, so I don't know
> >> which type of raw cable to buy....in order to insure my homebrew cable
> >> works at it's full capacity (which is 480 MB/sec). There are USB
> >> repeater cables that extend the max cable length out to 50 feet, but
> >> they're expensive, they draw a lot of power and they don't work at the
> >> full data rate for USB 2.0. And, none of the repeater cables have
> >> shielding at the housing for the driver electronics.
> >>
> >> I've thought about using an ethernet cable and putting usb connectors on
> >> the ethernet cable. But, ethernet uses balanced transmission line
> >> (twisted pair)...it might not work well for usb because USB doesn't use
> >> a balanced line-so there is little benefit to be had by moving to
> >> ethernet cable for usb data transmission.
> >>
> >> I sure wish the RTL-SDR had an ehernet cable connector instead of a USB
> >> connector!
> >>
> >> My problem is I cannot find the usb 2.0 specification that outlines the
> >> critical parameters for the cable itself. Any ideas?
> >>
> >> Sorry to be so long winded::>
> >>
> >> Art
> >>
> >>
> >>
> >>
> >> --
> >> 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

[AB Pearce - UKRI STFC]

> I've thought about using an ethernet cable and putting usb connectors on the ethernet cable.
> But, ethernet uses balanced transmission line (twisted pair)...it might not work well for usb
> because USB doesn't use a balanced line-so there is little benefit to be had by moving to ethernet cable for usb data transmission.

Why do you say USB is not a balanced signal? It quite definitely is a balanced signal, the data lines are called D+ and D-, and when laying out a PCB definitely need to be laid out as a differential pair.

I suspect your problem with the long cable is purely one of voltage drop in the cable if it really is drawing 350mA. I would try some method of powering the receiver locally. Rather than down the length of cable.



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

[Denny Esterline]

Not at all what you asked - but perhaps helpful....
I've had nothing but trouble from the longer "active" USB extensions. Many
devices do not play nice with them.
My usual solution is a _powered_ USB hub and a plain cable PC to hub.
As long as I use a good hub, I've had very few device balk at this solution.

-Denny


On Sun, Jul 8, 2018 at 3:25 PM, AB Pearce - UKRI STFC <
***@stfc.ac.uk> wrote:

> > I've thought about using an ethernet cable and putting usb connectors on
> the ethernet cable.
> > But, ethernet uses balanced transmission line (twisted pair)...it might
> not work well for usb
> > because USB doesn't use a balanced line-so there is little benefit to be
> had by moving to ethernet cable for usb data transmission.
>
> Why do you say USB is not a balanced signal? It quite definitely is a
> balanced signal, the data lines are called D+ and D-, and when laying out a
> PCB definitely need to be laid out as a differential pair.
>
> I suspect your problem with the long cable is purely one of voltage drop
> in the cable if it really is drawing 350mA. I would try some method of
> powering the receiver locally. Rather than down the length of cable.
>
>
>
> --
> 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

[Art]

TY all!!!!

Jean-Paul-The dynamic range is even worse than 60 db, it's an 8 bit
sample, which has 40 db dynamic range at best. My hope is to properly
filter out the out of band signals that cause it to overload so
frequently, and since I'll be listening to the noise floor a lot
(signals are weak at best), I'm hoping to squeak by with it, at least
long enough to see if aircraft band monitoring is useful. it’s a valid
concept. I expect to be listening to the noise floor a lot, so if I can
filter out the out of band signals that cause overload, I might be able
to get by.

Richard-90 ohms differential sure sounds like usb is indeed a balanced
line feed. I’ve done the calculations and 90 ohms is indeed obtainable
in close spaced twisted pair lines. In fact, for low voltage
applications, it is even possible to obtain 50 ohms impedance with a
twisted pair. Ethernet uses twisted pair transmission lines that are
differential and I have thought of homebrewing an ethernet cable with
USB connectors on it. Somehow, for some reason, I thought the usb was
just a common ground with a separate unbalanced input and output line
for the data! I found all kinds of info on ethernet cable, but found
almost nothing regarding the specs needed for USB 2.0 cable. If USB data
lines are indeed twisted pair, with the shield only used for shielding,
then I should be looking for cat 6 ethernet cable (maybe).

AB-It sure sounds like USB data lines are balanced feedline, twisted
pair! If you have any web references regarding design of PCB for USB
conductors, I would appreciate a pointer to the info. Is USB 2.0 full
duplex (which is possible with only 2 conductors).

Denny-I am going to modify one of my USB hubs to see if the resistance
of the power lines is making excessive voltage drop. However, the USB
cable I bought for the project is supposed to have 24 gauge power lines
in it, which amounts to less than 60 milliohms resistance over the 10
foot cable length-so droop from the load of less than 400 mA should not
be a problem. Making a powered hub is easy-I just haven’t done it yet. I
did evaluate whether the USB poort on my computer was a hub, which means
the available current to each USB is divided up between all the other
usb ports connected to the hub. I believe I have 2 usb connectors on the
back panel which are not bubs, and even those do not work.

All-I am aware that there is a time limit that actually limits the
maximum length of the USB cable to 16.5 feet, and if the sending side
doesn’t get a reply in a certain amount of time, data transmission is
reattempted. I also discovered that many implementations of USB drivers
do not do the verification of the round trip time. Honestly, I’d like to
have about 50 feet of cable, which gets me well away from the (rf)
noisey computer and let’s me put up a decent antenna on a mast beside
the house.

I should point out that all the usb ports on my computer drive the USB
HDD and the RTL-SDR well when they are plugged into the ports directly,
it's only when I try to use the extender cable that the data doesn't flow.

Again, if anyone is aware of any web resources that detail how usb works
and the technical specs for the necessary cable, please point me to
them!!! And, thanks to all who commented-this list never fails to bring
great info on technical topics!

Art

--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.ed

[Brent Brown]

Art,

Your first post (which I've lost for the moment) had a link to the SDR device you
have, and on that site was a picture of two devices plugged into a Raspberry Pi.

If you have a Pi laying around... put it and the SDR on the end of an Ethernet cable
or access via WiFi. The latest model Pi's (model 3 rev B?) can also run from PoE,
but require an adapter PCB plugged in on top to do so. Don't know what the
performance/noise would be like compared to running USB from your PC, just might
be an option to get the distance required.


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

[Art]

Thanks Brent,

The Raspberry Pi is popular and it might be less rf noisy than a full
size punchbox!

It certainly doesn't have as much electronics in it and it has far less
IO for sure.

Switching to the Pi might negate a significant amount of noise generated
in the computer and radiated by its I/O cables. I don't have a Pi, but
there is one in my future::> They're still getting cheaper::>

I'm not sure the Pi can do high speed USB 2.0 due to power limitations
and its ability to shed heat. The SDR will not work unless full speed
USB 2.0 is used, they wann you many times about this--and it's why I
bought a quality USB extender cable that fell well short of the rated
cable length of 16.5 feet for full speed USB 2.0.

Regards,

Art

On 07/09/2018 01:03 AM, Brent Brown wrote:
> Art,
>
> Your first post (which I've lost for the moment) had a link to the SDR device you
> have, and on that site was a picture of two devices plugged into a Raspberry Pi.
>
> If you have a Pi laying around... put it and the SDR on the end of an Ethernet cable
> or access via WiFi. The latest model Pi's (model 3 rev B?) can also run from PoE,
> but require an adapter PCB plugged in on top to do so. Don't know what the
> performance/noise would be like compared to running USB from your PC, just might
> be an option to get the distance required.
>
>



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

[Christopher Head]

On July 8, 2018 9:13:09 PM PDT, Art <***@myfairpoint.net> wrote:
>AB-It sure sounds like USB data lines are balanced feedline, twisted
>pair! If you have any web references regarding design of PCB for USB
>conductors, I would appreciate a pointer to the info. Is USB 2.0 full
>duplex (which is possible with only 2 conductors).

A USB 1 or 2 cable is four conductors: ground, +5V (roughly), and a half duplex mostly differential pair. At low speed 1.5 Mbps) and full speed (12 Mbps), the data lines are voltage based, with usually one being at about +3 and one at 0 (though both-at-0, referred to as single ended zero, is used to reset a device or mark the end of a packet). At high speed (480 Mbps), as I understand it, the voltage signalling is replaced with current signalling, but most other things remain the same (not sure what the SE0 state is replaced with). At super speed (3+ Gbps, USB3) there are more conductors and everything gets complicated.

The USB 2 specification (and many others) is freely available from usb.org.

If power is the issue, have you considered the hacky but easy solution of using an existing USB cable to ship data, cutting the +5V line, and injecting your own stiffly regulated +5V (sourced from some higher voltage) near the radio board? This would technically lead to non-compliant behaviour when the computer is off or asleep, but is unlikely to be a problem in practice.

--
Christopher Head

[Xiaofan Chen]

On Mon, Jul 9, 2018 at 2:50 PM, Christopher Head <***@chead.ca> wrote:
>
> The USB 2 specification (and many others) is freely available from usb.org.

Which is actually a pretty good document to read.

usb.org also has compliance testing software to download.
http://www.usb.org/developers/tools/

Vendors like Keysight offers electrical compliance test fixture
and test software but they are not cheap.

As for the power issue, I think powered hub is a good solution.


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

[Xiaofan Chen]

On Mon, Jul 9, 2018 at 10:00 PM, Xiaofan Chen <***@gmail.com> wrote:
> On Mon, Jul 9, 2018 at 2:50 PM, Christopher Head <***@chead.ca> wrote:
>>
>> The USB 2 specification (and many others) is freely available from usb.org.
>
> Which is actually a pretty good document to read.
>
> usb.org also has compliance testing software to download.
> http://www.usb.org/developers/tools/
>
> Vendors like Keysight offers electrical compliance test fixture
> and test software but they are not cheap.
>
> As for the power issue, I think powered hub is a good solution.
>

For cable, usb.org has the info as well.
http://www.usb.org/developers/compliance/cable/
http://www.usb.org/developers/compliance/cable/CabConn20.pdf

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

[Art]

I'd like to report that I found the mother lode(s) of technical
information for the operation of USB devices. I have learned so much in
the last few hours thanks to the new found info.

The breakthrough came when I stopped searching for 'usb balanced
feedline' and started searching for 'usb differential data
transmission'!!!! What a difference a few words make in an internet
search attempt!

Tomorrow's a new day, but at least I'll retire knowing that
documentation is out there and just waiting for me to find.

Regards,

Art



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