TS-PC struggles to finish startup calibration.

[EpEpDono]

So I bought two TS-PC servo bases for chips, but they both have startup calibration issues, both start up and spin left to right a couple times, same fast speed both ways, and then tries to center, usually they miss the center, after which its back to spinning left and right. One of the bases has managed to finish calibration twice, after I fiddled with the Hall effect sensor, but seems like the calibration was off somehow, I then turned the wheel all the way clockwise and it rapidly spun all the way back counterclockwise, after which it seems the calibration was perfect, and FFB seems to work from the control panel . But when powering on the next time it struggles to calibrate again. It often times returns to dead center, but then starts spinning again...

I have read that the T300 sometimes has a issue where the rotor loosens from the shaft, but in those cases the spin in one direction is usually slower than the other? I also found that Angle Motor Calibration tool and tried to run it, but it does not seem to even detect the bases, does it only work on the T300? Any idea what the issue could be? Firmware and drivers are fine.

Also, does anyone know the significance of the line on the driving cog? I assume its supposed to be aligned with the steering wheel at center?

 

[EpEpDono]

Ok, I think I have found a way to get this thing to calibrate somewhat reliably, I put this piece of cardboard under the Hall sensor to push it up from the fastener, and during the calibration, which is still very violent and strikes the stops more than twice, I move the sensor around a little, then after three of four strikes it calibrates. Could the magnetic field have changed over time? Maybe a good idea to swap out the magnet and see if that fixes it? Anyone know what strength this magnet is supposed to be?

 

[EpEpDono]

A little bit more progress, If I position the sensor just right using that bit of cardboard, which seems to be rotated slightly from the plastic mounts, then it calibrates pretty much perfectly and does not slam into the stops at mach speed. I guess this means that the sensor magnet or motor core magnet is slightly rotated? The sensor magnet mount looks and feels very solid, so idk how that could have rotated...

 

[EpEpDono]

I played around with it a bit more and it seems like the FFB is much stronger and faster in the counter clockwise direction so I guess that's a core that's rotated on the shaft...