tom@syssoft.com (Rodentia) (04/04/91)
Greetings: I have been experimenting with an IR receiver I modified from an article. It is a two stage amplifier, meant to get high gain, but I am trying to receive square waves, and it does not appear to be sufficient for my needs. At this time, I am sending a 25KHz 25% duty cycle square wave (at least that's how it looks on the transmitter). On the reciever, it is very smoothed out, and resembles a sine wave of 50KHz. I suppose this is due to the harmonics present in the assymetrical waveform. Adjusting the gain is very touchy. There is little latitude from no signal to "chaos to the rails". In addition, it is very sensitive to distance and alignment (when tuned for poor reception, if I point it directly, it goes wild). My first goal is to clean up the received wave. It can be adjusted such that the beginning of each wave form generates a scopeable spike (at 25KHz, not the 50KHz harmonic) and this differential signal seems that it may be enough to reconstruct a good approximation of the transmitted wave. A similar spike on the falling edge may be there, and it would be a plus to use that information as well, but it is not necessary. I can assume the pulse width. My second goal is to get the bloody 60Hz and 120Hz out of there. This is complicated by the fact that these sources of interference (flourescent lights, televisions) may have fast rise and fall times fooling the circuitry I am considering to reconstruct the wave. One bit of information I don't want to lose is the pulse position (or perhaps frequency). I am attempting to transmit analog signals, but I am pushing the transmitter to the rails, so all of my modulation is in the time domain. This would be interesting to try on digital, but just using two tone detectors won't cut it, I need continuous values. Maybe I could threshhold them after this stage for digital reception. If I can't generically reconstruct received waves (through some sort of automatic gain control or edge detection), I am willing to try FM reception. Unfortunately, I know even less about this, but it does seem that good sensitivity is possible. FM chips I have seen do not go down to KiloHertz carries frequencies so I'd have to roll my own. It seems that this may be the kind of problem HAMs might have (though with a twist), so I'm hoping that there are some suggestions from that sector. References are good. Examples, discussions, tutorials are even better. Please copy me by e-mail on any followups. I DO read these groups, but news has been so constipated out here, your responses may get expired before I receive them. Thank you for your patience. The schematic is included below for your information (heck, you may want to build your own; it can test your remote controls if nothing else). Vcc | P1 +----P2--+ | | | +----P3--+ +------C1--+->I-\ | | | | :OA1 O>-+-R3--+->I-\ | PhT +->I+/ :OA2 O>-+----C2--+ | | +->I+/ | | | | | | Vcc +--------+-----+ P4---JAK | | | | | +----------R1--+-R2--+ | | | | | | C3 | | | | | | GND GND GND GND Legend: Vcc - 9V to 12V GND - Ground PhT - PhotoTransister (IR sensitive) C1 - Capacitor (.001 microFarad) (.01 was also tried) C2 - Capacitor (.01 microFarad) C3 - Capacitor (220 microFarad) R1 - Resistor (10KOhm) R2 - Resistor (10KOhm) R3 - Resistor (1KOhm) P1 - Potentiometer (1MOhm, center tap wired to one end for log scale) P2 - Potentiometer (1MOhm, center tap wired to one end for log scale) P3 - Potentiometer (1MOhm, center tap wired to one end for log scale) P4 - Potentiometer (10KOhm, to ground with center tap to jack) OA1 - Op Amp 1 (1/2 353) (I- is - input, I+ is + input, O is output) OA2 - Op Amp 2 (1/2 353) -- Thomas Roden | tom@syssoft.com Systems and Software, Inc. | Voice: (714) 833-1700 x454 "If the Beagle had sailed here, Darwin would have | FAX: (714) 833-1900 come up with a different theory altogether." - me |
yjj@ctr.columbia.edu (Yuan Jiang) (04/07/91)
In article <1991Apr3.211135.21492@syssoft.com> tom@syssoft.com (Rodentia) writes: >Greetings: > > At this time, I am sending a 25KHz 25% duty cycle square wave >(at least that's how it looks on the transmitter). On the reciever, >it is very smoothed out, and resembles a sine wave of 50KHz. I >suppose this is due to the harmonics present in the assymetrical >waveform. > Remeber sine wave has negative cycles. [sin(wt)]**2 is a 2w sine wave plus dc. > My second goal is to get the bloody 60Hz and 120Hz out of there. >This is complicated by the fact that these sources of interference Will a filter help? > PhT - PhotoTransister (IR sensitive) They are known for being noisy and having low responsivity. A PIN photodiode costs several dollars. Good luck
tom@syssoft.com (Rodentia) (04/10/91)
In article <1991Apr6.203416.14689@ctr.columbia.edu> yjj@ctr.columbia.edu (Yuan Jiang) writes: >In article <1991Apr3.211135.21492@syssoft.com> tom@syssoft.com (Rodentia) writes: >>Greetings: >> >>resembles a sine wave of 50KHz. I >>suppose this is due to the harmonics present in the assymetrical >>waveform. >> >Remeber sine wave has negative cycles. [sin(wt)]**2 is a >2w sine wave plus dc. > I understand that there is biasing involved, but the receiver is AC coupled. Where does the squaring of the wave come into play? The asymmetry to which I was referring was in the duty cycle. >> My second goal is to get the bloody 60Hz and 120Hz out of there. >>This is complicated by the fact that these sources of interference > >Will a filter help? > I imagine so, any suggestions as to what stage would be the filter? Is making one of the op amps an active filter best? > >> PhT - PhotoTransister (IR sensitive) > >They are known for being noisy and having low responsivity. >A PIN photodiode costs several dollars. > I wanted to be cheap, but I will look into decent parts. >Good luck Thanks. Update: I changed to a 1458 op amp (from a 353 BIFET). This is much more stable at high gains. It still seems to be sensitive to the volume control on the Radio Shack audio amp I connect. Any suggestions on how to match this to my circuit? Thanks all. -- Thomas Roden | tom@syssoft.com Systems and Software, Inc. | Voice: (714) 833-1700 x454 "If the Beagle had sailed here, Darwin would have | FAX: (714) 833-1900 come up with a different theory altogether." - me |
yjj@ctr.columbia.edu (Yuan Jiang) (04/14/91)
In article <1991Apr10.040631.15699@syssoft.com> tom@ssi.UUCP (Rodentia) writes: >In article <1991Apr6.203416.14689@ctr.columbia.edu> yjj@ctr.columbia.edu (Yuan Jiang) writes: >>In article <1991Apr3.211135.21492@syssoft.com> tom@syssoft.com (Rodentia) writes: >>>Greetings: >>> >>>resembles a sine wave of 50KHz. I >>>suppose this is due to the harmonics present in the assymetrical >>>waveform. >>> >>Remeber sine wave has negative cycles. [sin(wt)]**2 is a >>2w sine wave plus dc. >> > I understand that there is biasing involved, but the receiver is >AC coupled. Where does the squaring of the wave come into play? The >asymmetry to which I was referring was in the duty cycle. I don't remember whether you specified what kind of light source you used. The squaring applies to lamps. If you are using a LED or laser and operating in the linear region, then I don't have any clue where the 2w frequency is from. > >>> My second goal is to get the bloody 60Hz and 120Hz out of there. >>>This is complicated by the fact that these sources of interference >> >>Will a filter help? >> > I imagine so, any suggestions as to what stage would be the >filter? Is making one of the op amps an active filter best? Depending on where this line frequency comes in. I guess it's from the light source since 2w(120Hz) is there. Filter the DC power supply. If possible, seal everything is a metal box. >>