IR Receiver Interferance
tma
Posts: 27
Hi All
As an extension to the Understanding Signals - Chapter 7 - Pulse Width Modulation with Infrared activities I·tried observing the IR signatures from various IR remote control devices using the Parallax IR detector and the OptaScope. I had some problems with spurious responces.
I connected the detector in the standard fashion - Vdd to pin 3, Gnd to pin 2, output on pin 1.
I monitored pin one with OptaScope set: Ch1 2V/div, Hor 5mS/div, trigger normal falling, single aquisition.
The detector would indicate presence of 38kHz at odd intervals by triggering when no IR signal was being sent to it and I had to to try several times to capture the remote signal correctly.
I had read (in the Weekend IR Project from Parallax - I think) that some of the new flourescent lamps might cause trouble. My lighting is standard (old) flourescent but they proved to be the source.
I tried, unsuccesfully, sheilding the detector from the lights with black cardbord. Turning them off worked.
My question: Is the detector susceptable to electromagnetic interferance at other than IR frequencies?
I know that flourescent fixtures are famous for RF interferance. Would shielding the detector in a metal enclosure help?
If anyone is interested I have some info on the protocols used by other IR remotes:
- 38kHz - Hitachi TV/VCR, Electrohome VCR, Sanyo Stereo
- 20kHx - Panasonic
I posted three Optoscope pictures of the Sanyo IR remote signal (button 1) on the Test Forum as "Test OptaScope Attach" to make sure I knew how to attach them.
tma
As an extension to the Understanding Signals - Chapter 7 - Pulse Width Modulation with Infrared activities I·tried observing the IR signatures from various IR remote control devices using the Parallax IR detector and the OptaScope. I had some problems with spurious responces.
I connected the detector in the standard fashion - Vdd to pin 3, Gnd to pin 2, output on pin 1.
I monitored pin one with OptaScope set: Ch1 2V/div, Hor 5mS/div, trigger normal falling, single aquisition.
The detector would indicate presence of 38kHz at odd intervals by triggering when no IR signal was being sent to it and I had to to try several times to capture the remote signal correctly.
I had read (in the Weekend IR Project from Parallax - I think) that some of the new flourescent lamps might cause trouble. My lighting is standard (old) flourescent but they proved to be the source.
I tried, unsuccesfully, sheilding the detector from the lights with black cardbord. Turning them off worked.
My question: Is the detector susceptable to electromagnetic interferance at other than IR frequencies?
I know that flourescent fixtures are famous for RF interferance. Would shielding the detector in a metal enclosure help?
If anyone is interested I have some info on the protocols used by other IR remotes:
- 38kHz - Hitachi TV/VCR, Electrohome VCR, Sanyo Stereo
- 20kHx - Panasonic
I posted three Optoscope pictures of the Sanyo IR remote signal (button 1) on the Test Forum as "Test OptaScope Attach" to make sure I knew how to attach them.
tma
Comments
We recently switched to IR detectors enclosed in metal, not plastic. They perform much better under the lights in our office that drive our old detectors nuts. That's not proof-positive because it could also be a function of a better filter design inside the detector. This will take a bit more testing.
Here's a an IR detector interference sniffing program. All it takes is an IR detector with its output connected to P9 and a piezospeaker with its input connected to P4. It should be useful for comparing RF vs optical shielding with the detector.
·
[color=green]' {$STAMP BS2}[/color] [color=green]' {$PBASIC 2.5}[/color] [color=blue]DO[/color] [color=blue]IF[/color][color=#000000] IN9 = 0 [/color][color=blue]THEN FREQOUT[/color][color=#000000] 4, 100, 4000[/color] [color=blue]LOOP[/color]Boe-Bot book? -bugg
I just checked, and the one in Robotics with the Boe-Bot has a cooler alarm tone that's nested in a FOR...NEXT loop, with flashing LEDs too. Either one will get the job done though.
the Robotics kit. -bugg
It is also possible to modify your program so that it filters out the interference.· Your remote will send specific pulse durations.· Most of the programs just decide whether the pulse is less than or greater than a value.· The program can be modified so that it sets a less than and a greater than value for Start, zero, and one pulses with tighter tolerances.· Then, if one or more of the pulse measurement is not within the boundaries, the packet is discarded, and the program tries again.
As a side note, IR Remote for the Boe-Bot is a text with some more recent program examples.· While the example programs don't delve into filtering for interference, they do save quite a bit of memory.· IR Remote For the Boe-Bot is a free download from www.parallax.com -> Downloads -> Stamps in Class Tutorials.
There are similar techniques you can use for filtering IR interference from the IR detectors.· First, check if the detector is sending a low signal before broadcasting the IR.· Then send IR and check save the signal in a bit variable (standard IR object detection).· After that, check again.· If there's a low signal before and/or after the standard IR objec detection, then a low reading has to be discarded.
RF is nice.· We have a keychain controller and receiver pair.· Even though they cost more, I like to have a couple of transceivers on-hand because there are so many different projects you can do with them.
Post Edited (Andy Lindsay) : 1/7/2005 9:20:29 PM GMT
As you and others have pointed out the sniffer will let me know if I have IR interferance.
I can also use it to test some IR detector modules that I salvaged from old TV's and VCR's.
One of these units is likly 20kHz (one of the remotes in my bin transmitts at that freq)
I will read more on the subject in the Boe Bot manual when I complete the Understanding Signal Unit
tma
I recently was playing around with some of the non-metal encased ones ( http://www.parallax.com/detail.asp?product_id=350-00014 ) on a Javalin stamp, and getting *lots* of spurious signals
I found that when certain VP code was running on the stamp, the problem would be worse than other times.
I have a couple of the metal encased ones, these seemed much more stable, but I wasn't satisfied that the problem was with the component.
Thinking that it might be electrical interference (I don't have a scope, so I couldn't tell for sure), I tried putting an .1uf capacitor across the power leads of the detector. Problem solved, no more spurious readings.
This proved to be the case with several that I tried. I'm thinking they are just sensitive to noise on the incoming power. The VP code was generating a high frequency signal on another pin, which was apparently affecting the power coming into the ir detector.
Might want to give it a try.
Post Edited (JamesC) : 1/11/2005 3:30:02 PM GMT
I also tried decoupling the detector PSU lead - no help.
I hav'nt noticed any interferance from the Stamp or BOE yet.
My interferance is definitly RF radiation from the flourecents, I have eliminated IR by mounting the whole BOE + detector in a black cardboard box with a very small window thru which I sent signals from my remote.
I also modified Andy Lindsay's "DecodeSonyIRRemote.BS2" program so that it essentially mimics the Optascope. It will record the spurious trigger events like the scope does in capture mode. Probably not as well but it's a useful tool.
tma
I might throw out there that long leads can act as excellent antennas, cause cross-talk, etc. If you're doing this on a breadboard, it could be acting as an antenna. If you did try the capacitor, make sure it is as close to the IR detector as possible. Even shortening the detector leads..... Guess what I'm driving at is that the RF *might* not be affecting the sensor directly.
Suppose you could try wrapping the detector in metal (tin foil? being careful not to short the leads), and see if that has an effect. might be a good way to rule out direct vs. indirect interference. I found it strange that the metal wrapped sensors I have (came with BOE bot), did not have the problem, even though the filter cap seemed to fix it for the non-metal wrapped ones. Since the problem (for me) was coming from the other electronics on the board, thru the power, I would have expected the same results, regardless of shielding. Guess there must be some other difference in the sensors. Worst case you might try different make sensor.
Another thing to consider (I'm grasping at straws here), is that the interference from the lights might be coming in thru the AC power line. I've heard that old ballasts can sometimes generate a lot of noise that feeds back into the AC line. I wouldn't think that is the problem (not sure what you are using as a test bed), as most power supplies/ regulator circuits have a variety of things that should filter this stuff out. Isolating the entire project using a battery would be a quick way to rule that out.
I was hoping my experience would apply to your problem. Worked wonders for me.
Good luck, and please post the solution when you find it, as I'm real curious now....
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Name: Bruce Clemens
Work:· Clemensb@otc.edu
Bolg: http://theDeadBug.journalspace.com
However, being just a test, the leads are very sloppy.
I suspect the antenna effect is my problem.
When I get some time I will mount my BOE into one of the bays in an old computer frame thats sitting under my bench.
I will let you know.
tma