Little trouble with ArduPilot 328 board: NMEA parsing stops, but restarts when touching 3rd Rx channel Input signal pin !!!
A strange thing happens with my ArduPilot board (new 328 version) with shield and LD20031 GPS (uses NMEA parsing). I would describe it as follows:
- Once in a while, everything runs fine: GPS fix locks blue, serial yellow led blinks.
- Sometimes, I had no lock and no blinking yellow led
- When I grabbed the board touching random pins, everything went ok again (blue lock, yellow blinking).
- When I put the board down, blinking may stop after few second or immediately, blue lock too.
- I found out that when I touch the signal pin (or its solder point) of Rx channel Input Number 3, it runs perfectly and doesn't stop working
- I remelted every single solder point to be sure to have good contact, but problem is the same
- For now, I found a way out: I grounded this specific pin to the GND (-) of Rx channel 5 which is unused with a little piece of wire. And it works flawlessly (GPS locks blue, serial stream not stopping).
Now I wonder if the problem would resolve itself when the servo input is connected to the Rx, which is not the case yet.
Would someone have an hypothesis about that peculiar phenomenon and how I could resolve it without the grounding of this signal pin? That would be helpful and I would be grateful.
- Once in a while, everything runs fine: GPS fix locks blue, serial yellow led blinks.
- Sometimes, I had no lock and no blinking yellow led
- When I grabbed the board touching random pins, everything went ok again (blue lock, yellow blinking).
- When I put the board down, blinking may stop after few second or immediately, blue lock too.
- I found out that when I touch the signal pin (or its solder point) of Rx channel Input Number 3, it runs perfectly and doesn't stop working
- I remelted every single solder point to be sure to have good contact, but problem is the same
- For now, I found a way out: I grounded this specific pin to the GND (-) of Rx channel 5 which is unused with a little piece of wire. And it works flawlessly (GPS locks blue, serial stream not stopping).
Now I wonder if the problem would resolve itself when the servo input is connected to the Rx, which is not the case yet.
Would someone have an hypothesis about that peculiar phenomenon and how I could resolve it without the grounding of this signal pin? That would be helpful and I would be grateful.
Replies to This Discussion
-
Permalink Reply by Thomas J Coyle III on -
Reto,
Without a receiver input on Rx channel #3 it is just floating and that is probably causing the Atmega 328 to become confused as to the input signal that it is trying to process. It is expecting a pulse train and it is just getting random noise most likely.
Connecting the servo output from the Receiver to that input will most likely correct the problem.
Regards,
TCIII
-
Permalink Reply by Reto on -
I took off my signal pin grounding. It seemed at first that with the input connected from Rx, the problem was solved. But it seems it is only solved when in "manual" mode. As soon as I switch to "autopilot" or to "RTL", the serial stream gets stuck again!!! This means that there must be something wrong in the board around this signal pin and/or down the line. I'll try to investigate that tomorrow evening.
If Chris or Jordi read this, you may have an idea how I could investigate that further. Thanks much.
© 2009 Created by Chris Anderson
