vorpalrobotics
commented
6 years ago Hello,
You are better off posting hardware questions over on our forums here: https://groups.google.com/forum/#!forum/vorpal-robotics-forum
However, a few notes:
1) SG90 servos will not work for this project. The plastic gears cannot withstand the stress of a walking robot. They'll fail within minutes. This project requires metal gear servos. Also, the SG90 is a slightly different size than the MG90 and it won't quite fit in the servo holders on the robot, you'd need to modify the parts.
2) Your symptoms do sound like either you have a bad servo driver board, or something is drawing a huge amount of current. The thing drawing the current might be a bad servo, I've seen that happen. you could unplug half the servos and see if it still happens, then plug in the other half, and effectively do a "binary search" for the bad servo (if there is one). Of course any kind of short circuit in the wiring could also cause this.
3) If you could, put a current meter in the circuit and see exactly how much current is being drawn, that might be a good idea. There could be other esoteric causes, such as a bad flash on the nano. With the servos idling under no load there should be no more than a few hundred milliamps draw.
Hope this helps!
-Steve P.
Sometimes when starting the bot all the servos instantly go to the max position and make a buzzing noise. The light on the nano is very dim when this happens and it never recovers. It has worked before, but not with all servos connected, only with half or two thirds (worked no problem, all modes responded etc.) What I suspect is happening is that the PWM board is drawing a lot of power, so the voltage drops and the Nano dies because of it. I am currently powering everything from my bench supply, I haven't hooked it up to my BEC and battery because I wanted to know it's safe first. I only tested up to 3.5A limit since the recommended BEC is below that. I am using the recommended parts, but SG90 servos (I know they are cheap and probably crap). Does anyone know what could be causing this? Is the PWM board booting faster than the nano and drawing a lot of power before the SPI connection has a chance to be established?