I have an APM2. Right now it is fully functioning. To be clear:
1. It has 4 servos hooked up, 2 aileron (Y split cable to one channel), 1 elevator 1 rudder.
2. It is powered by exactly 5.1V coming from a 35A ESC with a 5V/2A BEC that is connected to channel 3 on the output (servo side rails)
3. I have a 3DR radio for telemetry that works like a champ. I can connect every time and get GPS love
4. Channel 5 is hooked into Ch 8 on the input side, 3 flight modes (default ones suggested) all work just fine.
5. I have never fried the board, nor damaged it.
6. The daughter board is soldered to the APM2 motherboard.
7. The AttoPilot current sensor (50V/90A) tiny circuit board is hooked up exactly as pictured in the wiki to the APM2 rails. White wire from the current hole to the APM2 board, black is ground and red is the voltage. I have double checked this connection a LOT.
8. The jumper pin is connected so the entire APM2 board is powered via the 5.1 BEC hooked into ch 3 on the output rails.
The only thing I did different in the connection is that I made two female-female cables that plug into the APM2 board (in the correct slots without connecting any wires to the 5V middle pin on the rails) and instead of soldering the wires to the AttoPilot current sensor I put in pins and made a removable connection. This removable connection is the same connection servos use. They all tested fine when I checked for continuity via my fluke.
PROBLEM:
When I load up the MP, version 1.1.87 (the latest, I love updates so I know its the latest and greatest), and I connect to my aircraft (after a normal boot of the APM2 board) the flight data HUD displays nothing (as in 0 volts and 0%). I go to the configure tab and select Battery Monitor. I go from Disabled to Battery Voltage (the next in the sequence). This lights up the first input field for BEC voltage as measured minus .3V. I put in 4.8V for the first field as the measured voltage coming out of my ESC is 5.1V. I click on the 3rd option on the pulldown menu and select voltage and current. That allows me to enter the measure battery voltage. I input 11.31. Nothing happens. I click on the flight data button to see if its displaying something. It is not. I click back to battery monitor tab and the 4.8V is still there for first field, but calcd voltage is 0 and my measured voltage input field is also now 0. The bottom two fields sometimes displays 15.7 or 27 whatever, its calcd so I cant mess with that(?). The sensor selected is the 90A version on the pulldown menu. That matches with what is on my AttoPilot circuit as well.
Should I just skip the pin connections on the Attopilot circuit and solder them in like in the wiki? I find it hard to believe that that would cause this problem, but I'm mechanical not electrical engineer.
I'd really like to solve this problem so I can get on with my next problem :)
Any help would be appreciated a great deal!
Replies
So, by accident I figured this out, it is NOT a software issue. Somehow through all of the instructions I missed the part that says to sodder the circuit inbetween your battery and your motor. This pics on the wiki documentation did how this but it was of an older Attopilot circuit and l dismissed the validity of it based on that. The wiki documentation clearly (and wonderfully) shows how the Attopilot current sensor should hook up to the APM2, but not how the battery and motor should hook up to the Attopilot sensor itself. So, I was confused.
Case closed.
It's not the way you wired the Atto, I wired it the same. It is a bug in MP.
Try modifying your process a bit:
Select both dropdowns correctly without entering anything. Then zero out all the fields. Then enter them again.
Some sequence like that eventually got mine to work. Once. Since then it has been too much hassle to get it working that I simply haven't used the Atto. It needs a bug report, if you wouldn't mind please copy your issue above into an issue form here: http://code.google.com/p/ardupilot-mega/issues/entry
Thanks
The Battery Volts Calibration page is/was a bit weird - I had to have several goes to get it to work.
Here is a recent discussion on this.
Sounds like it hasn't changed. It might warrant submitting a bug report.