gbalke
commented
3 years ago This was caused by the 8x motor controllers starting up at 3.3V and the MCUs + peripherals pulling a large amount of current. This was handled in #8 as it is not the responsibility of the interface board to handle this spike (spike is now shifted to 12-16V ish which pulls the same power at a much lower current).
When testing the interface board for slow-start, I observed that inrush current at 3.3V was very high for the motor controller boards. This led to numerous I-max bumps and over-current timeout increases to account for 8 boards in series being used by the Blue project. The goal for this issue is to shift this low-voltage current draw into a higher voltage range such that the supply-side sees a much smaller spike. As of PR #8, this has been shifted up to 12V. This issue will track the success of that change and the specifics of the problem.
Note that these scope images are taken with a really cheap car current probe, hence the noise.
These first two images are the only ones with the proper current settings (was learning how to use this setting on my scope). Both are done with 8 motor controllers.
Note the smooth current draw after start-up, this is the charging current draw expected. The second bump is the large current draw of the 8 boards starting up around 3.3V. This is one of the better captures to see this issue. In this case, the current spike caused the controller to shut down the supply.
10x these current readings (five boards)
(100mV = 1A) Three boards:
Two boards:
One Board:
