I put the power inputs of the card on an o-scope today with the relays clicking away, and I don't see anything that looks odd. Seems to me this has to be a voltage issue, so now I'm a bit confused. If that is the case, then it doesn't make sense that the card locked while running on USB power? Is the PIC power isolated from the power used for the output drivers?
Does you entire system work (without locking up) when not in a really cold environment?
If so, then it will be a temperature issue.
I would make sure that everything works first at "normal" temperature before subjecting it to a harsh environment.
In all the testing I have done here at the house (on a regular 120vAC to 12vDC power brick), I have never once had it lock up or do anything odd. In the limited testing I've done with it on the tractor while parked in the heated shop (running on battery 12vDC), I've never had it lock up. The only two times I've had it lock up are when I'm outside and the tractor is running, which have both been days with temps below freezing.
If it is a temp issue, that would explain why it locked up when powered by USB. If it is a voltage issue, the PIC should remain powered via the USB.
What is odd about the temp theory is that last time when it locked up, I pulled the cover off the plastic box to unplug/reconnect the DC power jack and take a picture. That would have allowed the board to cool off further, and it should have locked up again. I only ran the card another ~4 minutes outside after that, so it wasn't a good enough test.
I'm thinking I need to take the rig outside on a cold day and let it set for a while with the FB active to see if I can get it to lock up. If it does, then I know it is a temperature issue, not a voltage issue. Forecast for tomorrow looks to be about as cold as before, so that should be a good test.
Greenman, when the PIC is powered by USB, is the crystal also being powered by USB?
If its a temperature issue it doesn't matter if the unit is powered from USB or 12v.
"USB only" power is only useful for running the FB when 12v is not required. Though you would need an external power source if you want to use 12v relays! So you might as well run it off 12v in this case. If you entire system runs of 5v only then a "USB only" power is handy (set by jumper).
P.S. Im sure Greenman will say the same - If the FB has power, the crystal has power! No matter if its powered from USB or 12v.
The only thing I would suggest now is have a little heater running in your box just to keep the FB warm as an experiment. If it doesn't "lock up" after a long test, then I would upgrade the crystal then do the test again.
To be frank, I don't think the FB was intended to be operation at really low temperatures. I'm sure Greenman can confirm this.
*Note: If you want to reliably use the FB in a really cold environment for extended periods of time there are more things you have to consider.
Well you had V3 running no problems with an inverter setup right?
Can you put it back on the inverter and see if it works fine?
Originally Posted by 2k1Toaster
My gut feel is that it's an electrical noise issue, not temperature.
Even thought the power is isolated, the grounds are not... and noise can propagate across a PCB even if its electrically isolated.
also, i havent forgotten your watchdog issue
Yes, I had a V3 running on an inverter and it ran great that way. It had issues with the voltage when running on DC right from the battery. Part of the reason behind going to the V4 was to eliminate this.
The V3 was also located in the cab, and I was only tested it in late spring when the temps were much warmer.
There are enough things different now that I'm not sure which one it is. It could be the same old voltage issue, it could be the cold temperature, or maybe something else. If it is voltage, I probably need to have the machine running to cause it to lock. If it is a temperature thing, then it should lock after sitting outside for a while, regardless of if the tractor is running. I intend to test the temperature today.
Today I put the tractor outside for 5 hours and then did a little testing with the engine off. Temperature was under 20 degrees. I connected to the FB and sent/received data for 15 minutes. I then had it clicking a relay for another 15 minutes. No lockups. If this is temperature related, it didn't show up today.
Conventional wisdom says to use the process of elimination, but this is so intermittent that I'm having trouble narrowing it down to anything.
Here's what I'm down to so far. I have experienced it locking 2 times. In both cases, it was while the tractor was running, outside, moving, and with the auto-steer system engaged. In both cases, the system was trying to turn the steer axle (which it doesn't do all the time), and in both cases, it locked while trying to turn left. The left and right solenoids are identical in form and connection, so I doubt that is anything other than statistical odds. If I experience more locks while turning left, I'll start to wonder. If it locks while turning right, then I'll start to wonder if it has something to do with a brown out while one of the steering coils energizes. The brown out theory doesn't make sense because the solenoid circuit comes right off the battery. Both the FB and the solenoids do share a common ground - the engine block.
Temperature tomorrow is going to be warmer, probably above freezing. Maybe I should take it out for some more road testing to see if it will lock on a warmer day.
Anyone got an idea on how to eliminate some factors?
Change the ground of the FB to directly to the battery if you can.