VESC-Tool 2.04: FW 5 >> A BIG STEP FORWARD

[b264]

@b264

I just did an experiment and it’s not really changing the numbers much at all, having the belt and wheel connected. However, changing the firmware version changes the numbers a significant amount.

these are consistent with or without drive train connected

TB6380_3 new values detected at commits a8ccdba / 12a7106
I: 22.51A
R: 14.0mΩ
L: 6.01μH
λ: 4.335mWb

these were the average of several detections, all with no drivetrain connected

TB6380_3 old values detected with firmware 4.2
I: 23.37A
R: 12.20mΩ
L: 8.98μH
λ: 5.013mWb

Both of these were the same physical motor, ESC, and wiring. The enclosure hasn’t even been opened.

The drivetrain didn’t seem to affect the numbers at all on the experimental version but I hadn’t tried that test before.

I have new information to report about this.

It seems that any combination of 5.x firmware detects the lower L: 5.93μH values. Regardless. Firmware 4 was consistently around 9μH. Firmware 5 is consistently around 6μH.

However, when it comes to resistance, it appears to depend whether I do the detection directly, or use the wizard. I almost never use the wizard. Using the wizard very consistently finds around R: 12.2mΩ. But doing a manual detection very consistently yields around R: 14.0mΩ. This is even on the same firmware/tool version.

So that it matters whether you use the wizard or do the detection directly from the Motor Settings / FOC / General screen is a new discovery for me.

Which values are more reliable?