Jump to content

summers

Members
  • Posts

    1
  • Joined

  • Last visited

Everything posted by summers

  1. Just registered to give my 2 cents and perhaps aid in the development Flirc worked well for me for 2 days, and then i started seeing the repeated keypresses, and significant loss of sensitivity. In that time i built up my media center rack: speakers receiver subwoofer NAS modem router xbox ETC. After several days of tearing everything apart I have isolated the issue with FLIRC TLDR: The issue is an ELECTRICAL NOISE issue, NOT an infrared sensitivity issue. I'm afraid this may be all the way down to the hardware, unless a Firmware fix can filter electrical noise. Long version: I'm an embedded systems engineer, I'm not claiming to be an expert but this isn't exactly voodoo to me. After several days of frustration figuring out why FLIRC was no longer working, i systematically eliminated devices that weren't causing the problem. My final setup is as follows. FLIRC -> HTPC Laptop. (connected to AC adapter) -> HDMI -> Denon Receiver -> RCA subwoofer cable -> subwoofer -> back to the SAME OUTLET.as PC **The failures occur where these are the ONLY devices, and no other devices or speakers are connected to either the receiver or the PC** This is what is required to cause my FLIRC not to work. Ultimately my subwoofer is the root cause of the error. Sure, we can blame the it and say flirc is fine, however EVERY SINGLE other device works just fine, including every other USB device running on the same. If any component in this loop is broken , FLIRC works great, with such great reception i can reflect IR from my Harmony remote off the back wall and it will be recieved. Complete the loop and ~50% of the time FLIRC is lucky to detect IR 6 inches away, and when it does we get repeated keypresses quite frequently. The loop doesn't even have to be fully connected, simply short the grounds on either the HDMI or RCA terminals and FLIRC stops working. (same goes for the data pins too). The difference is dramatic! (Please don't blame lighting conditions, I realize FLIRC is sensitive, but realize that 99.99999% of TVS or other equipment have never received phantom input from lighting causing it to turn on without intervention, and I'd expect the same level of filtering from flirc once it is programmed.) It is interesting to note, that this ONLY occurs when the subwoofer is connected through the same electrical outlet as the other equipment (to prevent ground loops) If i run an extension cord through my house to an outlet on another breaker, FLIRC works fine with everything hooked up, so long as the subwoofer has a different ground. (sure, my APT doesn't have the best wiring, but again thats not the point is it?) This leads me to believe that ultimately electric noise is being read by flirc as an IR signal and being sent back through to the PC via USB as a valid input (or repeated keypress for some reason)??. I don't have an oscilloscope handy here, but my multimeter says the difference in ground voltage between the subwoofer and USB is .07V at its max. I'm not sure if FLIRC can actually be fixed, as i think this may be a hardware problem, Maybe FLIRC does not have the required resistors and capacitors to fully clean up a dirty USB signal, and instead we are leaking analog noise over the digital USB signal to the IR circuit and it is being interpreted as a command? i'm not sure. This is probably why 90% of cases with FLIRC acting up is people with raspberry PI, as its power is usually from a "dirty" wall wart. I've tried various solutions as you would expect USB hub, powered USB hub, changing USB ports, Adding resistors to create a voltage divider to attenuate the nose on the RCA cable, dark rooms, light rooms, other equipment. At this point i feel that any fix (ground loop isolater on RCA subwoofer cable or extension cord running through house to power subwoofer-causing an actual ground loop) would be an unnecessary band-aid on something that should just work. I apologize if i seem "ranty" I really think this is an awesome project but I figured the more info I gave the better! Let me know if you need any more data points! Thanks
×
×
  • Create New...