Maker Pro
Maker Pro

Infrared Receiver

A

Andrew

Jan 1, 1970
0
Hi all

I've been trying to design an infrared receiver, but have several grey
areas. My transmitter signal is a simple 10us pulse with a period of
2ms (500Hz)- it's just a remote 'on' switch. The system needs to work
in bright ambient conditions over at least 40m from a mobile hand-held
Tx.

No off-the-shelf products are available becuase they all have a BPF
tuned to, say, 36KHz. (Plessey SL486 used to do this exact job). The
output will connect to a microcontroller so the software can detect
the signal.

Using my limited knowledge I've come up with a circuit where the
signal is processed as follows:

1) A photodiode connected to a wideband transimpedance amplifier (this
works well)

2) A voltage controlled gain amplifier (AD602 or similar)

3) A Peak Detect that is connected to the VCG amp to detect the peak
voltage of the incoming signal and adjust the amp's gain to avoid
saturation if the Tx is too close. And increase it if it's far away.

4) A comparator to output to TTL

5) A pulse stretch (555) to interface with microcontroller

Am I barking up the right tree?

I'd love to know what it is I really need in this circuit. Do I need
an AGC to stop my amp saturating? Do you think I should include
filters into my design? Is a peak detector the best way of controlling
the gain of the amp in this case?

Any help at all that anybody can offer will be very much appreciated.

Andrew
 
R

Robert Baer

Jan 1, 1970
0
Andrew said:
Hi all

I've been trying to design an infrared receiver, but have several grey
areas. My transmitter signal is a simple 10us pulse with a period of
2ms (500Hz)- it's just a remote 'on' switch. The system needs to work
in bright ambient conditions over at least 40m from a mobile hand-held
Tx.

No off-the-shelf products are available becuase they all have a BPF
tuned to, say, 36KHz. (Plessey SL486 used to do this exact job). The
output will connect to a microcontroller so the software can detect
the signal.

Using my limited knowledge I've come up with a circuit where the
signal is processed as follows:

1) A photodiode connected to a wideband transimpedance amplifier (this
works well)

2) A voltage controlled gain amplifier (AD602 or similar)

3) A Peak Detect that is connected to the VCG amp to detect the peak
voltage of the incoming signal and adjust the amp's gain to avoid
saturation if the Tx is too close. And increase it if it's far away.

4) A comparator to output to TTL

5) A pulse stretch (555) to interface with microcontroller

Am I barking up the right tree?

I'd love to know what it is I really need in this circuit. Do I need
an AGC to stop my amp saturating? Do you think I should include
filters into my design? Is a peak detector the best way of controlling
the gain of the amp in this case?

Any help at all that anybody can offer will be very much appreciated.

Andrew

It is my guess that the "bright ambient conditions" (= noise) alone
will give S/N problems, and if not present, the distance (= diffusion)
will also give S/N problems.
Instead of a CW signal that you describe, if you would use a carrier
(hmm...try 36KHz for grins), then one can have a much better chance at
success.
 
A

Andrew

Jan 1, 1970
0
Robert

Thank you for your reply.

Unfortunately I am bound by the 10us pulse because the system I'm
designing has to be backwards compatible with the Tx.

The ambient light causes the photodiode to conduct resulting in a DC
off-set. This DC voltage can prevent the Tx signal from being detected
and so needs tracking and cancelling. Fluorescent lights are a problem
because they introduce AC noise. Therefore, I feel that I need active
filters to prevent this noise from propagating, but I don't know how a
square wave (10us pulse, period 2ms) works with regular op-amp
filters.

Cheers
 
R

Robert Baer

Jan 1, 1970
0
Andrew said:
Robert

Thank you for your reply.

Unfortunately I am bound by the 10us pulse because the system I'm
designing has to be backwards compatible with the Tx.

The ambient light causes the photodiode to conduct resulting in a DC
off-set. This DC voltage can prevent the Tx signal from being detected
and so needs tracking and cancelling. Fluorescent lights are a problem
because they introduce AC noise. Therefore, I feel that I need active
filters to prevent this noise from propagating, but I don't know how a
square wave (10us pulse, period 2ms) works with regular op-amp
filters.

Cheers

Two things you could do: 1) use lenses, both ends, to help keep the
beam on the photodetector; 2) use a long tube, flat black inside, as a
light shield around the photodetector.
 
G

Glenn Gundlach

Jan 1, 1970
0
Robert

Thank you for your reply.

Unfortunately I am bound by the 10us pulse because the system I'm
designing has to be backwards compatible with the Tx.

The ambient light causes the photodiode to conduct resulting in a DC
off-set. This DC voltage can prevent the Tx signal from being detected
and so needs tracking and cancelling. Fluorescent lights are a problem
because they introduce AC noise. Therefore, I feel that I need active
filters to prevent this noise from propagating, but I don't know how a
square wave (10us pulse, period 2ms) works with regular op-amp
filters.

Cheers
There is a reason for using 36/38/40 kHz modulation, to avoid the very
problems you're dealing with. So, you have a 10uS pulse every 2 mS ?
That is plenty of time to do some elementary modulation on the TX side
and use a commercial IR receive module. I used a 68JC908JK3 running 32
MHz a divide by 15 prescale and a 22V10 for a further divide by 56 to
make 38095 Hz (for a TV). For what you want,you likely don't even need
a processor, just a little extra gating, maybe even a 555 (I don't
like using an electronic hourglass to measure time) to make your 10 uS
into say, 1mS. Then detect it with a Radio Shack $3 IR module.

GG
 
A

Andrew

Jan 1, 1970
0
Hi Glen

I'd love to use a mudulated signal, but unfortunately I can't change
the transmitter at all.

An interesting point I discovered last week though: I was able to set
up an off-the-shelf 38KHz IR receiver to pick up my obscure signal
even though my signal isn't anywhere near that frequency- in fact, it
doesn't have a carrier at all! I simply powered the device up and
fired my Tx at it and it worked. How can this be when the IR Rx is
supposed to filter out all signals other than 38KHz? I need a device
that will give me more range though, which is why I'm going to plod on
with my own Rx design.

Many thanks

Andrew
 
Top