The power readings on the 5V lines were remarkably stable. Movements were less than 0.5 volts. Maximum amperage on the motors is difficult to measure, but as I said with both stalled it’s at 25A and the motors are identical so I’d guess that means 12.5A each. Those numbers are peak and our standard circuit breaker is only 20A for the entire system.
Test results today
I pulled the speed controller out of our test vehicle to try it out on a board and eliminate wiring and environmental factors and make different motor testing possible. Unfortunately the speed controller contacted a screw head on the new test board and burned out the microprocessor command output for channel 2. (Note to others the resistors that normally prevent this are on the bottom board and shorts across the connector pins between the two boards can apparently be lethal to parts of the processor).
I’m back to installing another board in our test vehicle and trying to make that one fail, but to tell you the truth at this point it seems that I might be better off returning the remaining speed controllers I have (except the one I accidentally damaged if you like). and converting our design over to some other controller. We simply can’t afford to figure out what is going wrong with this particular batch of controllers and we certainly can’t ship them.
A change of 0.5V on the regulated 5V line definitely does not constitute “remarkably stable”. In general, randomly shorting two nodes together can be lethal to any board; I strongly recommend you do your wiring and disassembling with power disconnected.
I understand if it doesn’t make sense for you to continue troubleshooting the problem; you can contact us directly for an RMA number. However, there are no deliberate changes to “this particular batch of controllers” and we have not replicated the problem here, so converting to a different controller could just be putting a bandaid on a more fundamental problem with your setup. We expect to have a several new motor controllers coming out in the next few days, so we might have more options for you soon.
Fair enough. Power was disconnected to the controller, but I jumped the gun on testing if my wiring was in place on the board and missed a small screw that was in the wrong place.
I’ll get exact fluctuation on the 5v and post it once I have a new controller in place. The under .5v variation I referred to was way under .5v but I didn’t have the exact numbers when I posted the update and didn’t want to exaggerate the stability. I seem to have implied that it was not stable at all - not what I meant to do and I will get an exact number. To clarify, there was no visible change on my scope when power was applied but there was variation of a small but not precisely known amount around the base supply voltage as soon as the device was turned on.
While there certainly could be a problem with our setup we are in a production environment here too. Every speed controller before these for several years in the same product with our same design has worked fine but now we are having replicable but very intermittent problems. It’s possible that we screwed something up, it’s possible that a supplier changed a part without changing the model number, it’s possible that there is a component issue, it’s possible something got stored in a warehouse and the air conditioning failed. I don’t know. I’m just looking at a whole bunch of components that haven’t changed but have suddenly stopped working together. Personally I hope it’s something we’re doing because that I can correct relatively easily.
The only issue we’ve had with speed controllers before this is from insufficient seating of the surface mount components during soldering on several units which caused a high pitched whine. Push on the chip and it went away. Not your problem - it was another company, but it did show me that there can be individual unit quirks even in a highly automated assembly process.
Given we had to work with a different speed controller I went ahead and put it in an different type of vehicle with different wiring lengths, that did not use mixing and has different motors to do a quick test.
Specifications: 12.8v resting, 11.1v at full stall. This unit is running two parallel motors on one channel. The second channel is not used. 8A peak run amperage combined for the motors (wheels spinning, vehicle held) with 11A when we briefly were able to fully stop the motors (this took multiple people holding the wheels themselves as the unit does not have enough traction to stall the motors on its own).
We were finally able to get consistent episodes if we flipped the control stick to full power then immediately back to neutral providing the motors were under at least some load. Full power was ‘up’ with a longer pulse length. Identical behavior was exhibited for two receivers by different manufacturers. We measured pulse outputs via a Y on the receiver and did not get any unusual readings. We believe we were able cause the behavior with a Hitec servo tester, but the requirements plus the tester itself make it difficult to be 100% sure that the temporary uncontrolled movement was caused. We are 90% sure.
So we pulled that speed controller and installed another one. That one worked flawlessly. We installed the first one again so I could get LED readouts and now to my surprise the one that had showed the behavior over and over again on command is now working perfectly. Battery supply voltage is within 0.1 volt of the original level. There have been no wiring changes of any kind. I waited several hours and tried it again. We’re back to intermittent events but not easily on command like before.
Apparently our easy to read scope has a limited Vpp and may not be able to catch very brief high voltage pulses. This will be resolved on the testbed setup.
