STM32 Motor Control Workbench 6.4.1 and Motor Pilot speed divided by 2 with F.O.C. algorithm

[bfa001]

This is to report an erroneous speed - with a divided by two speed - for the Motor Control Workbench and its associated Motor Pilot version 6.4.1.

The speed information throughout the Motor Control Workbench generates twice the mechanical speed with F.O.C algorithm and STEVAL-SPIN3202 control board.

The  measured tachometer shaft speed which leads to confirmed motor speed and associated computations are the following:

Motor Speed (RPM) = 2 x BEMF(Freq)(Hz) x 60 seconds / Poles pairs

Note: 2 factor meaning:  Each magnet passes over the coil used to sense the Back-EMF twice during each rotation, once over each end.

BEMF(Freq)  is matching HALL(Freq) in our application.

These checks were carried out with two motors:

1,400 RPM with a 10:1 gear ratio delivering 140 RPM at the shaft and 155 RPM with no gear 

In both cases, we have correct mechanical speed for half the speed entered within Motor Control Workbench and Motor Pilot. Motor Pilot reports half the speed of the actual motor rotation measured with a tachometer at the shaft.

The scope capture below shows the 1,400 RPM motor HALL  or BEMF (brown trace)
Motor Speed in RPM = 2 x HALL Freq (or BEMF = same Freq ) * 60 s / pole pairs = 2 x 70.17Hz * 60s / 6 = 1403 RPM - matching the shaft speed with 140 RPM with a certified and confirmed 10:1 gear

[GMA]

Hello @bfa001,

Could you check the Motor parameters (Pole Pairs number instead of Pole number)?

[bfa001]

Hello GMA,

Pole number was never mentioned - only pole pairs.

The pole pairs are correct with 6 pole pairs for the 10:1 gear motor and 14 for the 155 RPM without gear.

These pole pairs are confirmed and valid - the two motors run with Field Oriented Control and 6-step algorithms with these parameters. 

They are well validated.

Kind regards,

bfa001

[GMA]

Hello @bfa001,

In this Shinano motor use case, the targeted speed is 1400 RPM. The motor has two pole pairs, which means there are two H1 signal periods per mechanical rotation as described here. The HALL H1 signal was measured.
Motor speed (RPM) is calculated using the formula:
Motor speed (RPM) = (frequency of Hall sensor) × 60 seconds / pole pairs
For this case: Motor speed (RPM) = (1 / (21.38 ms)) × 60 / 2 = 1403 RPM. This value matches the reading reported by the tachometer.

Does rotating the shaft by hand, there is 3 or 6 HALL H1 period per rotation on your side?

[bfa001]

Hello GMA,

Whatever the computations are the value in Rotation Per Minute does not match the mechanical tachometer speed.

I must enter 700 RPM in STM32 Motor control Workbench 6.4.1 and Motor Pilot for having the 1,400 RPM at the shaft.

The pole pairs are correct and verified - by turning by hand with two coils shorted, we have exactly 60 steps for a 10:1 ratio - i.e. 6 pole pairs and for the 155 RPM motor without gear, the pole pairs are 14 exactly.

The motor parameters entered within Motor Control Workbench 6.4.1. are correct but result in a doubled speed at the shaft.

The error lies with Motor Control Workbench 6.4.1 when using the STEVAL-SPIN3202.

Kind regards,

bfa001

[bfa001]

Hello GMA,

To answer your question

Q1. "Does rotating the shaft by hand, there is 3 or 6 HALL H1 period per rotation on your side?"

A1. There are 3 periods for HALL H1 with 1 manual turn (30 with the 10:1 gear).

The equation Motor Speed in RPM = 2 x HALL Freq * 60 s / pole pairs is verified for 6 pole pairs, we have RPM = HALL Freq * 60 / 3 (for 1 motor turn we have 3 HALL periods) = HALL Freq * 60 * 2 / 6 = 1,400 RPM

Kind regards,

bfa001

[GMA]

Hello @bfa001,

Using MCSDK6.4.1 and the STEVAL-SPIN3202 in FOC mode:
. Close JP5, JP6, and JP7.
. Open JP9, JP10, and JP11 (not mentioned in the Workbench hardware changes tab).
With the Shinano motor, set with 2 pole pairs (2 H1 signal cycle per mechanical rotation as described in motor datasheet) in the Workbench interface and HALL sensor:
. Set a 1500 RPM command on the motor pilot.
. The motor pilot reports 1500 RPM, which is confirmed by the tachometer measurement.

Test confirmed with a 4 pole paire (4 H1 signal cycle per mechanical rotation) LIN motor.

[bfa001]

Hello GMA,

Thank you for your checks and support and I think that the ST motor control environment is suitable for our applications and an excellent choice for F.O.C. algorithm

I confirm that I have the same hardware F.O.C. settings for the STEVAL-SPIN3202 board.

However, I cannot explain the discrepancy in Motor Control Workbench 6.4.1 and Motor Pilot I found.

The pole pairs are verified and I have F.O.C. running smoothly on these two motors - but with the speed settings in Motor Control Workbench entered as half the real motor shaft speed.

I have to enter half the motor speed  in Motor, Speed Sensing Config. and Drive Settings sections for having twice the speed at the shaft.

The Profiler which was finally run successfully on both motors also returned half the speed as the maximum speed detected. The Profiler which runs on a Nucleo set of boards have the same issue with half the speed detected on both motors.

Kind regards,

bfa001