Hallo,
I'm using pilight 8.0 on RPi3 with low-pass-filter. I'm sending a signal (INTERTECHNO protocol) genrated by an ATTINY85 / 433MHz-transmitter to the RPi. pilight does'nt detect / decode the signal apart from a few exceptions, while RCSwitch-RFSniffer / wirless sockets do not have any problems. When I do the same with an Arduino Nano all works pretty fine.
Are there minimum requirements for signal strength / waveform different from RCSwitch/wiringpi based in wiringx?
wupperpi
Please do the following:
1. Start pilight-debug with and without filter, and if that doesn't work well enough, use pilight-raw.
2. Send codes with each program you said you tried and post the raw output with and without filter for each of them.
Here are the debugs. I'm sorry that I cannot add debugs without low-pass-filter since the receiver is soldered and unfortunatly I don't have a second one. But as far as I think, it will make no difference, because RFSniffer is getting the RX via the same low-pass-filter.
Some additional informations:
- TX-pulselength is 300 (RCSniffer is detecting ~ 300)
- RCSnffer is calculating pulselength by dividing the long part of the received syncword through the estimated pulselength. There are difficulties with the ATTINY in correct detection too, at half times another protocol is detecteed but curiously with the correct result in system-code, unit-code, state.
- I have tried to create a new pilight-protocol.in 433.92 and traced it. With ATTINY-TX there are only rare calls parse code. May be it stops processing earlier.
(01-03-2018, 06:39 PM)curlymo Wrote: What receiver are you using?
It is a RXB6.
It is a little peculiarly: if I run pilight-daemon and RFSiffer at the same time therre are more detections by pilight compared to running pilight alone. And decoding within pilight is then more precise.
I suppose the interrupt detection within RPi makes the difference. Enlarge receive-tolerance may be a solution. I will try to figure it out this weekend.
