I believe I broke a few ground pins. I only get 2.3V measuring to 3.3V pin on two pins. The other ground pins has basically 0V. The remaining ground pins measure 3.3V as one would expect.
My Setup AS5047P encoder is connected as specified in the documentation.
I also switched the resistor and jumper to get 3.3V mode on the encoder. I have shielded the encoder cable and put ferrite on the motor cable close to the Odrive and XT90 to help with the inrush current.
I have the 0Dirve 24V with a 20V battery.
The problem occurred once after applying a lot of force to in closed-loop about 100W → getting: MOTOR_ERROR_UNKNOWN_TORQUE
The other two times It just happened sitting around possibly plugging in or out but I don’t have anything in mind → error ENCODER_ERROR_ABS_SPI_COM_FAIL. (happened with different encoders but same motor and Odrive)
So my question:
What could have caused this?
Not as important: how can I fix this? Is there some chip/resistor I can replace?
I’d like to add, that once the encoder is disconected the one pin goes back to 3.3V - but the other 2.3V is now also at 0V …
There are two 3.3v LDOs. Where are you measuring? And what is your setup? Schematics of some photos are required.
According to the schematics, AGND and GND are tied close to U8 LDO. Check if there’s good connection . And check if all GND pins are connected.
Is something hot on the PCB when motors are OFF and just DC power is applied?
First of all, thank you for your help!
Regarding the connection between the GND pins: the two pins that are at 0V have no connection to the other ground pins on the board or between the two of them. The working GND have a connection between them (meaning quite low Ohms but I can’t give an accurate measurement).
Regarding temperature: I don’t have access to a thermal camera, so I can’t be sure. But feeling around nothing seems out of the ordinary. (I also don’t remember the encoder getting hot at the time)
But if you say U8 could be the reason I can try to replace that.
More important for me now, is how could that have happened?
A short maybe but I would not know how that should have happened
According to changelog v3.6 PCB has separate ground plane layers, so no connection between GND pins seems to look like poor soldering.