[EE] IR Transmission

[Harold Hallikainen]

Several years ago I designed a system that transmits audio and data over
IR. The audio is transmitted as frequency modulated RF over IR. The data
is transmitted is frequency shift keyed RF over IR. I've been thinking
about direct digital modulation of the IR. The IR LEDs would be pulsed.
The bit rate to carry the audio and digital would need to be in the area
of 5 Mbps. Does anyone have experience, thoughts, or references on this?
The IR is transmitted in a large room (movie theater auditorium). I'm
concerned about reflections causing intersymbol interference. What do
people think would be a good method of encoding the data (async, biphase,
etc.)? Are such systems existing? Are there any chip sets?

Thanks!

Harold

[Forrest Christian (List Account)]

If I did my calculations right, you're at about 30 meters per half bit.
Not sure about your environment, but it would seem like 30m isn't far
enough especially considering possible strong ir reflective surfaces. My
guess is that you'd end up with at least some dead areas, possibly a lot of
dead areas, where a simple on off bitstream couldn't be recovered. This is
even more true if you are relying on reflections to get the ir signal
through. If you decide to try this, please report back with your findings
as I'm kinda curious.

It would be interesting to investigate whether one of the rf digital
modulation methods (ofdm/qam, psk, etc) could be applied. Some of these
are multipath sensitive, some are not so much.

One possible ic resource is the power line modems such as the max2990.
These are generally ofdm modems in various frequency bands.


On Dec 26, 2017 12:26 PM, "Harold Hallikainen" <***@mai.hallikainen.org>
wrote:

Several years ago I designed a system that transmits audio and data over
IR. The audio is transmitted as frequency modulated RF over IR. The data
is transmitted is frequency shift keyed RF over IR. I've been thinking
about direct digital modulation of the IR. The IR LEDs would be pulsed.
The bit rate to carry the audio and digital would need to be in the area
of 5 Mbps. Does anyone have experience, thoughts, or references on this?
The IR is transmitted in a large room (movie theater auditorium). I'm
concerned about reflections causing intersymbol interference. What do
people think would be a good method of encoding the data (async, biphase,
etc.)? Are such systems existing? Are there any chip sets?

Thanks!

Harold


--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
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]

How about spread spectrum?

Quote from [1] "Applying spread spectrum technology to RF transmissions is
an extremely effective method of securing both wireless voice and data at
the hardware level. It is a very robust TX/RX technology that is inherently
immune to inter-symbol interference (ISI), reflections, noise and other
environmental factors, as well as jamming."

Spread spectrum transceiver chips are readily available however they
typically will operate either at a higher frequency or lower bandwidth than
you want. A down converter may be necessary to shift things back down to
the frequency range that you want.

[1] http://semiengineering.com/kc/knowledge_center/Spread-Spectrum/225

On Wednesday, December 27, 2017, Forrest Christian (List Account) <

Post by Forrest Christian (List Account)
If I did my calculations right, you're at about 30 meters per half bit.
Not sure about your environment, but it would seem like 30m isn't far
enough especially considering possible strong ir reflective surfaces. My
guess is that you'd end up with at least some dead areas, possibly a lot of
dead areas, where a simple on off bitstream couldn't be recovered. This is
even more true if you are relying on reflections to get the ir signal
through. If you decide to try this, please report back with your findings
as I'm kinda curious.
It would be interesting to investigate whether one of the rf digital
modulation methods (ofdm/qam, psk, etc) could be applied. Some of these
are multipath sensitive, some are not so much.
One possible ic resource is the power line modems such as the max2990.
These are generally ofdm modems in various frequency bands.
Several years ago I designed a system that transmits audio and data over
IR. The audio is transmitted as frequency modulated RF over IR. The data
is transmitted is frequency shift keyed RF over IR. I've been thinking
about direct digital modulation of the IR. The IR LEDs would be pulsed.
The bit rate to carry the audio and digital would need to be in the area
of 5 Mbps. Does anyone have experience, thoughts, or references on this?
The IR is transmitted in a large room (movie theater auditorium). I'm
concerned about reflections causing intersymbol interference. What do
people think would be a good method of encoding the data (async, biphase,
etc.)? Are such systems existing? Are there any chip sets?
Thanks!
Harold
--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
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

[Jason White]

Sorry for the poor formatting of my mail! I sent that from my mobile device.

On Wed, Dec 27, 2017 at 6:49 AM, Jason White

Post by Jason White
How about spread spectrum?
Quote from [1] "Applying spread spectrum technology to RF transmissions is
an extremely effective method of securing both wireless voice and data at
the hardware level. It is a very robust TX/RX technology that is inherently
immune to inter-symbol interference (ISI), reflections, noise and other
environmental factors, as well as jamming."
Spread spectrum transceiver chips are readily available however they
typically will operate either at a higher frequency or lower bandwidth than
you want. A down converter may be necessary to shift things back down to the
frequency range that you want.
[1] http://semiengineering.com/kc/knowledge_center/Spread-Spectrum/225

[Harold Hallikainen]

Thanks to all for the ideas! Spread spectrum, ODFM, etc. would, as I
understand it, pretty much be RF over IR, similar to what we're doing now
(multiple RF carriers on IR). For the digital data, I'm currently running
FSK with +/- 100 kHz shift at 1.8 MHz. I am running 10 kbps. As I tried
increasing the bit rate, the coverage area dropped substantially. I'm
using an SA639 for the receiver. It's designed for the Digital European
Cordless Telephone system which runs at about 1.152 Mbps when using FSK. I
use an op amp based preamp after the photodiodes with bias to the
photodiodes applied through a slug tuned inductor which becomes a parallel
resonant circuit when combined with the photodiode capacitance. The preamp
drives the SA539 which upconverts the 1.8 MHz to 10.7 MHz. Ceramic IF
filters and a ceramic resonator in the quadrature detector complete the
receiver.

The max2990 is interesting. It only supports 100 kbps in a 10 kHz to 490
kHz spectrum. The LEDs we're using can be modulated up to about 30 MHz, so
more spectrum is available. The MAX2990 appears to require two way
communications for ack, which would not be possible in our application.

It would be nice if I could just on/off key the IR LEDs instead of having
to amplitude modulate them, but that may not work out.

For the analog FM (carrying audio) receiver, we have two designs. One is a
TRF receiver using ceramic filters. A second design upconverts the RF over
IR signal to the FM broadcast band where it is demodulated in a standard
single chip FM receiver (DSP-based).

Thanks for the comments! I'll do some more research.

Harold

Post by Forrest Christian (List Account)
If I did my calculations right, you're at about 30 meters per half bit.
Not sure about your environment, but it would seem like 30m isn't far
enough especially considering possible strong ir reflective surfaces. My
guess is that you'd end up with at least some dead areas, possibly a lot of
dead areas, where a simple on off bitstream couldn't be recovered. This is
even more true if you are relying on reflections to get the ir signal
through. If you decide to try this, please report back with your findings
as I'm kinda curious.
It would be interesting to investigate whether one of the rf digital
modulation methods (ofdm/qam, psk, etc) could be applied. Some of these
are multipath sensitive, some are not so much.
One possible ic resource is the power line modems such as the max2990.
These are generally ofdm modems in various frequency bands.
On Dec 26, 2017 12:26 PM, "Harold Hallikainen"
Several years ago I designed a system that transmits audio and data over
IR. The audio is transmitted as frequency modulated RF over IR. The data
is transmitted is frequency shift keyed RF over IR. I've been thinking
about direct digital modulation of the IR. The IR LEDs would be pulsed.
The bit rate to carry the audio and digital would need to be in the area
of 5 Mbps. Does anyone have experience, thoughts, or references on this?
The IR is transmitted in a large room (movie theater auditorium). I'm
concerned about reflections causing intersymbol interference. What do
people think would be a good method of encoding the data (async, biphase,
etc.)? Are such systems existing? Are there any chip sets?
Thanks!
Harold
--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
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

[Forrest Christian (List Account)]

On the powerline communications technology there are various standards and
chipsets out there, for instance there are ones for Ethernet over power
line. You may also want to look for a dsl chipset.

My experience is that often there is a test or other mode in some of these
chipsets which allow to turn off the MAC layer and just use the PHY in raw
mode.

I might be tempted to try a higher order QAM modulator to get the bits per
symbol up a ways, which would increase your time per bit, and maybe
subsequently increase the range due to lower intersymbol inference. I'm
not an rf engineer but I know that a mid-90's DSP could handle a fairly
high bitrate when modulating over a copper channel. It wouldn't surprise
me to find that with the march of Moore's law that a medium sized dsp could
handle the bitrate you need. I also know that there are i-q modulation ICs
available which could offload some of this but I'm not sure how they work
for sure.

One final thought would be to add an error correcting code, such as some
sort of parity scheme. I'd suggest forward error correction but that adds
latency, depending on how you set it up. Both are things you probably
should consider.

Post by Harold Hallikainen
Thanks to all for the ideas! Spread spectrum, ODFM, etc. would, as I
understand it, pretty much be RF over IR, similar to what we're doing now
(multiple RF carriers on IR). For the digital data, I'm currently running
FSK with +/- 100 kHz shift at 1.8 MHz. I am running 10 kbps. As I tried
increasing the bit rate, the coverage area dropped substantially. I'm
using an SA639 for the receiver. It's designed for the Digital European
Cordless Telephone system which runs at about 1.152 Mbps when using FSK. I
use an op amp based preamp after the photodiodes with bias to the
photodiodes applied through a slug tuned inductor which becomes a parallel
resonant circuit when combined with the photodiode capacitance. The preamp
drives the SA539 which upconverts the 1.8 MHz to 10.7 MHz. Ceramic IF
filters and a ceramic resonator in the quadrature detector complete the
receiver.
The max2990 is interesting. It only supports 100 kbps in a 10 kHz to 490
kHz spectrum. The LEDs we're using can be modulated up to about 30 MHz, so
more spectrum is available. The MAX2990 appears to require two way
communications for ack, which would not be possible in our application.
It would be nice if I could just on/off key the IR LEDs instead of having
to amplitude modulate them, but that may not work out.
For the analog FM (carrying audio) receiver, we have two designs. One is a
TRF receiver using ceramic filters. A second design upconverts the RF over
IR signal to the FM broadcast band where it is demodulated in a standard
single chip FM receiver (DSP-based).
Thanks for the comments! I'll do some more research.
Harold

Post by Forrest Christian (List Account)
If I did my calculations right, you're at about 30 meters per half bit.
Not sure about your environment, but it would seem like 30m isn't far
enough especially considering possible strong ir reflective surfaces.

My

Post by Forrest Christian (List Account)
guess is that you'd end up with at least some dead areas, possibly a lot of
dead areas, where a simple on off bitstream couldn't be recovered. This is
even more true if you are relying on reflections to get the ir signal
through. If you decide to try this, please report back with your

findings

Post by Forrest Christian (List Account)
as I'm kinda curious.
It would be interesting to investigate whether one of the rf digital
modulation methods (ofdm/qam, psk, etc) could be applied. Some of these
are multipath sensitive, some are not so much.
One possible ic resource is the power line modems such as the max2990.
These are generally ofdm modems in various frequency bands.
On Dec 26, 2017 12:26 PM, "Harold Hallikainen"
Several years ago I designed a system that transmits audio and data over
IR. The audio is transmitted as frequency modulated RF over IR. The data
is transmitted is frequency shift keyed RF over IR. I've been thinking
about direct digital modulation of the IR. The IR LEDs would be pulsed.
The bit rate to carry the audio and digital would need to be in the area
of 5 Mbps. Does anyone have experience, thoughts, or references on this?
The IR is transmitted in a large room (movie theater auditorium). I'm
concerned about reflections causing intersymbol interference. What do
people think would be a good method of encoding the data (async, biphase,
etc.)? Are such systems existing? Are there any chip sets?
Thanks!
Harold
--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
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

--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
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]

Would it help to have a high / central location reflector (or just place
the tx/rx there) that gave a good line of sight 1st path at all times.

Would burst transmission help with fast data rate bursts followed by longer
quiet periods to allow echos and longer paths to fade.

Russell

[Harold Hallikainen]

Both interesting ideas. Our current receivers have photodiodes pointing in
all directions. It would be interesting to have them pointed in only one
direction to minimize reflections. Also, short high speed transmissions
with a gap to allow for reflections to die out is interesting.

Harold

Post by RussellMc
Would it help to have a high / central location reflector (or just place
the tx/rx there) that gave a good line of sight 1st path at all times.
Would burst transmission help with fast data rate bursts followed by longer
quiet periods to allow echos and longer paths to fade.
Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Forrest Christian (List Account)]

In the RF world, often you switch between receive antennas based on some
quality of signal metric. That's the way many of the early AP's worked
with two antennas. - there was some logic to determine which antenna was
'best' and that is which antenna was used.

On Wed, Dec 27, 2017 at 7:40 PM, Harold Hallikainen <

Post by Harold Hallikainen
Both interesting ideas. Our current receivers have photodiodes pointing in
all directions. It would be interesting to have them pointed in only one
direction to minimize reflections. Also, short high speed transmissions
with a gap to allow for reflections to die out is interesting.
Harold

Post by RussellMc
Would it help to have a high / central location reflector (or just place
the tx/rx there) that gave a good line of sight 1st path at all times.
Would burst transmission help with fast data rate bursts followed by longer
quiet periods to allow echos and longer paths to fade.
Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

--
*Forrest Christian* *CEO**, PacketFlux Technologies, Inc.*
Tel: [phone] | Address: 3577 Countryside Road, Helena, MT 59602
***@imach.com | http://www.packetflux.com
<http://www.linkedin.com/in/fwchristian> <http://facebook.com/packetflux>
<http://twitter.com/@packetflux>

[Manu Abraham]

DVB-T with Diversity switching deploys a single demodulator with multiple
antennae and LNA's. depending upon the feedback from the individual
AGC's the LNA/Antennae are switched correspondingly to avoid reflected
paths and or to look at higher signal strengths.

The higher the modulation order, the greater the pain you have with
reflections/doppler effect. Diversity was introduced with mobile DVB-T to
address the multipath issue to a certain extend.

Cheers,

Manu


On Thu, Dec 28, 2017 at 8:10 AM, Harold Hallikainen

Post by Harold Hallikainen
Both interesting ideas. Our current receivers have photodiodes pointing in
all directions. It would be interesting to have them pointed in only one
direction to minimize reflections. Also, short high speed transmissions
with a gap to allow for reflections to die out is interesting.
Harold

Post by RussellMc
Would it help to have a high / central location reflector (or just place
the tx/rx there) that gave a good line of sight 1st path at all times.
Would burst transmission help with fast data rate bursts followed by longer
quiet periods to allow echos and longer paths to fade.
Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

[Forrest Christian (List Account)]

The primary benefit of ofdm is to permit the correlation of signals
traversing multiple paths, in may cases turning multipath into a positive
thing. I'm pretty sure that the issues you described with multipath and
higher order modulation is why pretty much every modern spread spectrum
radio runs some form of ofdm modulation.

Out of curiosity I looked to see what code is out there for ofdm. Seems
like a few options are out there. Of particular interest might be gnu
radio and the various matlab or matlab like sources. Seems to be a few
others as well.




On Dec 27, 2017 11:59 PM, "Manu Abraham" <***@gmail.com> wrote:

DVB-T with Diversity switching deploys a single demodulator with multiple
antennae and LNA's. depending upon the feedback from the individual
AGC's the LNA/Antennae are switched correspondingly to avoid reflected
paths and or to look at higher signal strengths.

The higher the modulation order, the greater the pain you have with
reflections/doppler effect. Diversity was introduced with mobile DVB-T to
address the multipath issue to a certain extend.

Cheers,

Manu


On Thu, Dec 28, 2017 at 8:10 AM, Harold Hallikainen

Post by Harold Hallikainen
Both interesting ideas. Our current receivers have photodiodes pointing in
all directions. It would be interesting to have them pointed in only one
direction to minimize reflections. Also, short high speed transmissions
with a gap to allow for reflections to die out is interesting.
Harold

Post by RussellMc
Would it help to have a high / central location reflector (or just place
the tx/rx there) that gave a good line of sight 1st path at all times.
Would burst transmission help with fast data rate bursts followed by longer
quiet periods to allow echos and longer paths to fade.
Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
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]

Basically ofdm does the fundamental trick, yes. But that alone doesnt
address many issues.

When ofdm went mobile, with rayleigh fading paths, such as moving
trains/metros, ofdm alone doesn't keep up. Hence vendors had to look
for alternatives such as diversity and many others, while facing the
never ending issue with the hunger for more bandwidth.

You can see the issue described clearly in this article
https://cdn.rohde-schwarz.com/pws/dl_downloads/dl_application/application_notes/7bm05/7BM05_1E.pdf

Cheers,

Manu

On Thu, Dec 28, 2017 at 1:49 PM, Forrest Christian (List Account)

Post by Forrest Christian (List Account)
The primary benefit of ofdm is to permit the correlation of signals
traversing multiple paths, in may cases turning multipath into a positive
thing. I'm pretty sure that the issues you described with multipath and
higher order modulation is why pretty much every modern spread spectrum
radio runs some form of ofdm modulation.
Out of curiosity I looked to see what code is out there for ofdm. Seems
like a few options are out there. Of particular interest might be gnu
radio and the various matlab or matlab like sources. Seems to be a few
others as well.
DVB-T with Diversity switching deploys a single demodulator with multiple
antennae and LNA's. depending upon the feedback from the individual
AGC's the LNA/Antennae are switched correspondingly to avoid reflected
paths and or to look at higher signal strengths.
The higher the modulation order, the greater the pain you have with
reflections/doppler effect. Diversity was introduced with mobile DVB-T to
address the multipath issue to a certain extend.
Cheers,
Manu
On Thu, Dec 28, 2017 at 8:10 AM, Harold Hallikainen

Post by Harold Hallikainen
Both interesting ideas. Our current receivers have photodiodes pointing in
all directions. It would be interesting to have them pointed in only one
direction to minimize reflections. Also, short high speed transmissions
with a gap to allow for reflections to die out is interesting.
Harold

Post by RussellMc
Would it help to have a high / central location reflector (or just place
the tx/rx there) that gave a good line of sight 1st path at all times.
Would burst transmission help with fast data rate bursts followed by longer
quiet periods to allow echos and longer paths to fade.
Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

--
FCC Rules Updated Daily at http://www.hallikainen.com
Not sent from an iPhone.
--
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

[John Ferrell]

Amateur Radio repeaters with multiple inputs use a "voting" system to
select the best usable input.

One way to do that is to watch just outside the passband for a zero
noise condition.  If that technique has a patent it should have aged out
by now.

Post by Harold Hallikainen
Both interesting ideas. Our current receivers have photodiodes pointing in
all directions. It would be interesting to have them pointed in only one
direction to minimize reflections. Also, short high speed transmissions
with a gap to allow for reflections to die out is interesting.
Harold

Post by RussellMc
Would it help to have a high / central location reflector (or just place
the tx/rx there) that gave a good line of sight 1st path at all times.
Would burst transmission help with fast data rate bursts followed by longer
quiet periods to allow echos and longer paths to fade.
Russell
--
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

--
John Ferrell W8CCW
Julian NC 27283
It is better to walk alone,
than with a crowd going the wrong direction.
--Diane Grant