Problem with IR detection.
in [Design]
|
Prev: Lights with a mind of their own.
Next: Mosquito killer
From: pimpom on 8 Apr 2010 00:18 I'm trying to devise a presence detector using an IR beam and a standard remote control IR receiver module. I couldn't find datasheets for the receiver modules I can find locally and proceeded on the assumption that they'd be similar to the ones for which I have datasheets, like the Vishay, Sharp and Panasonic types. To minimize interference from unwanted radiation sources, the known types all set limitations on burst length, gap time and maximum number of bursts per second, and I made a note of those. But I cannot get the system working properly. I get detection briefly when the beam is unblocked and I get the expected waveforms at the receiver output, but the signal then dies down quickly. I've tried varying the burst length from 300-800usec with corresponding gap times. I'm not fully conversant with RC codes, but since I don't need to send data, I assumed that I wouldn't need those codes. Obviously I'm missing something, probably to do with using regular bursts. Can anyone point me in the right direction towards solving the problem?
From: mkaras on 8 Apr 2010 02:23 On Apr 7, 9:18 pm, "pimpom" <pim...(a)invalid.invalid> wrote: > I'm not fully conversant with RC codes, but since I don't need to > send data, I assumed that I wouldn't need those codes. Obviously > I'm missing something, probably to do with using regular bursts. > Can anyone point me in the right direction towards solving the > problem. Typical IR receiver modules require the signal being received to be on a carrier of 38KHz or 36KHz. In other words, at the transmit end take your pulse train waveform and use it to gate a 38KHz square wave on and off. The receiver will take in the modulated signal and will, if things are working correctly, output a good approximation of the original pulse train waveform without the carrier frequency present. If you have an o-scope it can be very instructive to open up an old TV remote control and take a look at the signal that drives the transmitter IR LED. mkaras
From: TTman on 8 Apr 2010 02:30 "pimpom" <pimpom(a)invalid.invalid> wrote in message news:hpjlfd$8dn$1(a)news.albasani.net... > I'm trying to devise a presence detector using an IR beam and a standard > remote control IR receiver module. I couldn't find datasheets for the > receiver modules I can find locally and proceeded on the assumption that > they'd be similar to the ones for which I have datasheets, like the > Vishay, Sharp and Panasonic types. > > To minimize interference from unwanted radiation sources, the known types > all set limitations on burst length, gap time and maximum number of bursts > per second, and I made a note of those. But I cannot get the system > working properly. I get detection briefly when the beam is unblocked and I > get the expected waveforms at the receiver output, but the signal then > dies down quickly. I've tried varying the burst length from 300-800usec > with corresponding gap times. > > I'm not fully conversant with RC codes, but since I don't need to send > data, I assumed that I wouldn't need those codes. Obviously I'm missing > something, probably to do with using regular bursts. Can anyone point me > in the right direction towards solving the problem? > Probably your IR transmitter is RC5 type encoding. It won't work, except as you describe. It needs to be as the poster above has described, or a microcontroller sending out UART data at a baud of 100- 500 Hz, modulated with 38Khz. You could go up to 1600 baud ish., which is what we use in our toy laser guns...
From: pimpom on 8 Apr 2010 03:25 mkaras wrote: > On Apr 7, 9:18 pm, "pimpom" <pim...(a)invalid.invalid> wrote: >> I'm not fully conversant with RC codes, but since I don't need >> to >> send data, I assumed that I wouldn't need those codes. >> Obviously >> I'm missing something, probably to do with using regular >> bursts. >> Can anyone point me in the right direction towards solving the >> problem. > > Typical IR receiver modules require the signal being received > to be on > a carrier of 38KHz or 36KHz. In other words, at the transmit > end take > your pulse train waveform and use it to gate a 38KHz square > wave on > and off. The receiver will take in the modulated signal and > will, if > things are working correctly, output a good approximation of > the > original pulse train waveform without the carrier frequency > present. > > If you have an o-scope it can be very instructive to open up an > old TV > remote control and take a look at the signal that drives the > transmitter IR LED. > > mkaras I know about the carrier frequencies. My receiver module is a 38kHz type and I have the Tx carrier at that frequency with pulses of approximately 40% duty cycle. To comply with the noise suppression requirement from the datasheets, the pulses are switched on and off in bursts. I started with a burst length of about 500 usec (about 18 cycles) and about the same gap time. When that didn't work, I varied the burst and gap lengths from 300 to 800 usecs. No difference. The pulses appear at the receiver output for a brief moment when the beam is unblocked and then disappear. The receiver works with my TV remote control.
From: pimpom on 8 Apr 2010 04:44
TTman wrote: > "pimpom" <pimpom(a)invalid.invalid> wrote in message > news:hpjlfd$8dn$1(a)news.albasani.net... >> I'm trying to devise a presence detector using an IR beam and >> a >> standard remote control IR receiver module. I couldn't find >> datasheets for the receiver modules I can find locally and >> proceeded >> on the assumption that they'd be similar to the ones for which >> I >> have datasheets, like the Vishay, Sharp and Panasonic types. >> >> To minimize interference from unwanted radiation sources, the >> known >> types all set limitations on burst length, gap time and >> maximum >> number of bursts per second, and I made a note of those. But I >> cannot get the system working properly. I get detection >> briefly when >> the beam is unblocked and I get the expected waveforms at the >> receiver output, but the signal then dies down quickly. I've >> tried >> varying the burst length from 300-800usec with corresponding >> gap >> times. I'm not fully conversant with RC codes, but since I >> don't need to >> send data, I assumed that I wouldn't need those codes. >> Obviously I'm >> missing something, probably to do with using regular bursts. >> Can >> anyone point me in the right direction towards solving the >> problem? >> > Probably your IR transmitter is RC5 type encoding. It won't > work, > except as you describe. > It needs to be as the poster above has described, or a > microcontroller > sending out UART data at a baud of 100- 500 Hz, modulated with > 38Khz. > You could go up to 1600 baud ish., which is what we use in our > toy > laser guns... I'm not using a readymade transmitter. I made my own using a 38kHz carrier, switched on and off in bursts. I started out with about 500usec on, 450usec off. When that didn't work (the received waveform appears briefly and then disappears), I varied the on-off periods between 300 and 800 usec. No difference. The receiver works with my TV remote control. |