Need help preventing TMC6300 from (maybe) short circuit panicking

I’m moving a project from a breadboard to a perfboard and I’m having a doozy of a time. This portion of it uses an rp2040, TM6300 driver, AS5048A encoder in SPI mode, and GM3506 motor. For some reason, when I set the motor to a position and subsequently push against it, it will softly cog 5 times, the amp draw dipping slightly below 2 amps each time, and then the TMC6300 panics, throws its DIAG line high, and shuts down. If I set the maximum motor voltage to 5v in Arduino Simplefoc, the perfboard circuit works fine, though it doesn’t perform like it does on the breadboard at full voltage. Additionally, the circuit and code work on the breadboard just fine, and In that configuration, the exact same individual ICs and motor can exert a constant 2 amps at 11 volts for however long I care to have it do so. I can push the motor in circles and it will continue to resist as hard as it can. There’s definitely nothing about the perfboard circuit that’s damaging the ICs and they are definitely not damaged, because I continue to be able to swap them back and forth between the breadboard and perfboard circuits.

Do any of you know what sort of situations trigger this panic mode? I’ve read the datasheet for the driver, and it really doesn’t do much to tell me what sort of outboard issues could cause the device to shut down. I know an overtemp, overcurrent, or short circuit event can cause this, but it seems weird that any of these would be happening if the motor can work at half current/voltage. The perfboard does connect the drivers with male/female headers, which the breadboard technically does not, but I can’t imagine that those would pass less current than a breadboard connection and shitloads of jumper wires.

How are you measuring the 2A?
The TMC6300 is rated 2A RMS, which is different from 2A DC, and this is 2A motor current, which looks different from power supply current, too.
I would not be surprised if you are just overdriving it. It sounds like it works just fine with marginally lower current settings, so why not just run it a little bit under spec?