openinverter forum

I bought a Volvo motor to connect to the Volvo inverter.
The motor connector looked the same but is was keyed different. Nothing a file can't fix.
Because this is a motor from a hybrid system it has a clutch to disconnect the driveshafts form the motor.
When the clutch whas active one driveshaft whas able to spin freely but the other one whas still hard too spin.
So i took the motor apart.The needle bearing whas damaged and pushed in to the diff so far blocking it.

Because the motor whas apart anyway I 3D printed a adapter to drive the motor with drill. hooked up a hall sensor as a referents puls.
Took a scope shot of the phase voltage and the reference and a other of the resolver output and the referents.
So for one revolution of the shaft the phase wire gives 5 sine waves. and the resolver outputs 10 "pulses" for one rotation of the motor shaft.

So 5 pole pairs for the motor?
And 2 pole pairs for the resolver?

For the setting in the open inverter software?

Assume you are driving the resolver with a few kHz? If so then you are seeing the waveform envelope on the scope and the 10 pulses are each a half cycle (so 5 full cycles the same as the phase output) so I would guess 5 pole pairs on both.

Edit - what are the ratings of that motor?

I used the resolver exciter signal from the open inverter board.

the motor is spec'd for 50Kw 200 Nm. I am planning to use it in my Saxo electrique it has the same co-axial design as the original dc motor.
So it will fit with the front battery box. I am running the car with out this box now. I'd like some more battery's

I think you are right about the resolver signal. One hump is in phase and the next hump is out of phase with the exciter signal

So 5 pole pairs for both.

Interesting motor! I particularly like the co-axial design and the clutches. Is it liquid cooled too?

The size and co-axial design are the reason I bought it.

Yes it is water cooled.
the clutches is controlled by this motor. it looks like a window motor.
as far as I know it is CAN bus controlled. because why not

I used a screw driver to move it takes just u couple of turns

manny wrote: ↑Wed Jul 12, 2023 9:16 am The size and co-axial design are the reason I bought it.

Yes it is water cooled.

20230712_105730.jpg

20230712_105742.jpg

the clutches is controlled by this motor. it looks like a window motor.
as far as I know it is CAN bus controlled. because why not

I used a screw driver to move it takes just u couple of turns

Hm... how exactly does that work? So you apply 12V to the clutch and it works like ac compressor? Or does it need specific CAN command to put torque through...? Or is this a parking brake?

Very interesting!
What is the weight of the motor?
What are the dimensions, rhoughly?

There's a bit more data here viewtopic.php?t=1433

About 27 cm from mounting flange to mounting flange
About 50 cm from driveshaft seal to driveshaft seal

Weight is around 48 Kg according to the other thread.

The clutch breaks the driveshaft that runs though the motor. And the other one spins free because of the diff.
I do not know how the CAN works with this motor.

The low voltage connector is a:

Hirschmann-automotive
12way 1.2 SealStar F
806-360-521 A bit hard to find, but the pins are available.


The pipe's for the cooling are 5/8" 15.82mm Quick Connect

After reading some more on the forum about tuning motor parameters. So i decided to measure the motor parameters:

Rs = Resistance of the stator winding
flux linkage = BEMF Constant
Ld = d-axis inductance of the stator winding
Lq = q-axis inductance of the stator winding

First Rs this is the simple one of the four. Basically stick a multimeter across two wires of the motor and read the result. But because the motors in car are high current devices the simple approach will not work.

The low tech way of doing this is easy. Use a power supply to push multiple amperes through the motor and measure the voltage as close as possible to the motor windings, to avoid adding wire and contact resistance to the measurement.

With this motor I pushed 10A through the motor from phase A to B. the voltage I got whas 228mV. this is the voltage across two windings.

228mV / 2 = 114mV

114mV / 10A = 11.4mΩ


The flux linkage. In a previous post I made a drill adaptor to spin the motor. With a oscilloscope connected across two phase wires we can measure the period time and the peak to peak voltage. With a formula from the appnote (linked below) we can calculate the flux linkage.

Vp-p * Tel / sqr(3)*4*Pi

From the scope shot in the older post I got 14 cycles in 550ms so this is 39.28 ms per cycle.

28V * 0.03928 / 21.766 = 50.53mWb


Next Ld, this is a bit more involved. We need access to the motor shaft, to align the motor with the phase A windings. The appnote suggest to apply DC power to the motor. phase A to the positive an phase B and C to the negative. The motor will "lock" to one of the stable positions. Counting the number of stable positions will give you the number of pole pairs. With the power applied we need to mechanically stop the motor from spinning. I used two pieces of wood with a single bolt, bolted to the face of the motor and the used a clamp to bite down on the motor shaft.

The appnote suggest to apply a voltage to the motor and measure the current. this is fine with a motor with a Rs of 6.4Ω. If I apply 14V to the motor it will pass 614A so that will not work. By adding in a resistor to limit the current we get a RL network.

I used three 10Ω resistors in parallel to get 3.333Ω. This gave a reasonable current at the voltage I used.

τ = L/R is the time constant. With a step response at one τ the current will be at 63% of the maximum.
the maximum current is about 12.4A 63% of that is 7.8A this is reached in 70us.

Ld = 2/3τ * R

Ld = 0.667 * 70us * 3.333Ω = 155.5 uH or 0,155mH


Last the Lq. To measure this we need to do almost the same setup as for the Ld but this time we need to connect the power supply different than before. Phase B to positive and C to negative, Phase A is not connected.

this aligns Phase A with the d-xis. This time it is more important to lock the rotor in place, because the motor will generate torque when power is applied. the test power is applied the same as before positive to phase A and negative to phase B and C.
This time it takes a bit longer to get to 7.8A, 150us

Lq = 2/3τ * R

Lq = 0.667 * 150us * 3.333Ω = 330.2 uH or 0.330mH

So the values i got are this:

Rs = 11.4mΩ
flux linkage = 50.53mWb
Ld = 155.5 uH
Lq = 330.2 uH

Not sure if this accurate, but is is comparable to other motors. It is similar to the outlander rear motor.
It whas fun to measure it anyway. If you see that I made a mistake please tell me.

Bonus thought is the Ld setup, a way to get syncofs. With my roughly tuned syncofs I got 350 degrees.