ACIM motor with resolver testing  [SOLVED]

Post Reply
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

ACIM motor with resolver testing

Post by arber333 »

I have made a dual Huebner controler drive on a single Volt inverter platform.
viewtopic.php?f=2&t=1027#p16839

Idea was to use two power stages independently from one another. Both support resolver interfaces.
So for the first controler i use 4.90R FOC PMSM drive connected to one Outlander motor.
The second motor is ACIM and i have 4.90R sine firmware loaded.
First brain is master and keeps the precharge and DC switch controls. Both drives are shown DC voltage through a single iso opamp.
I have succesfully started inverter on 90Vdc and spun the PMSM motor and ACIM motor at the same time.

However i see a problem with resolver interface. Or does it just seem that way.
My data shows motor is 8pole (4 pole pairs) with 4 pole (2 pole pairs) resolver. However resolver has 10 teeth on its stator! Shouldnt that made it 10pole? Strange... But i report i have quite forcefull start with setting 4P motor 2P resolver and good control of RPM at that setting. Distribution of torque is rough though. Motor wants to shake upon throttle application and amp draw is high comparing to PMSM motor.
I tried to eliminate some factors....
1. I start in manual sine and i give it command. It start smooth and run smooth up to 60Hz on manual.So motor is not the problem.

2. I program resolver mode and start in this with motor 4pole pairs and resolver 2 pole pairs. Motor would turn but very rough. If i change resolver poles it would not turn but jump through poles.

3. When i scope resolver i get good signal result from exciter and 0V to 3V3 signal on Sin Cos return. Is this value too much? Should i reduce my power supply for Sin cos signal? I use 5V split down by 470R/220R to 1,6V.

If i compare Lacrosse resolver with Outlander, there i get below 3Vpp return. So this resolver may be a bit different from others. Maybe resolver return signal is saturating the analog input on Olimex. I will try to reduce voltage on sin/cos input with 510R/220R resistor combination or maybe even 680R/220R. Otherwise i may recommend we use 5V to 1V5 regulator with additional resistor divider to further reduce supply.
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: ACIM motor with resolver testing

Post by arber333 »

I am trying to smooth out the ACIM motor resolver control. I think i have a problem in 2nd brain sensing somewhere.
I can run ACIM motor perfectly fine with master brain (1st controler) if i swap Olimex chips. But when i try to run with 2nd brain i constantly get rough running or OC error. I see PMSM code is more sensitive than ACIM and it seems it senses something before ACIM.

I managed to scope exciter signal on three exciter resistors. And immediately after first resistor signal becomes triangular instead of square.
On my first resolver signal translates nicely to a square weave directly on EX pins. I am thinking three 22n caps are suspect. I will replace them and see what happens.
User avatar
johu
Site Admin
Posts: 5682
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: ACIM motor with resolver testing

Post by johu »

So on the non-working brain you have square wave exciter signal?
On the resistor chain it should be square -> triangular -> almost sinusoidal -> sinusoidal.
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: ACIM motor with resolver testing

Post by arber333 »

johu wrote: Thu Oct 01, 2020 8:44 pm So on the non-working brain you have square wave exciter signal?
On the resistor chain it should be square -> triangular -> almost sinusoidal -> sinusoidal.
No. If i look at 1ms scope range.
On the working brain i get square signal on first resistor and up to input into LM386. Therefore i get resolver sin/cos return in square signal. Of course this square signal is modulated into sin/cos loop "potatos". And FOC and ACIM resolver code works very smooth with this brain circuit.

On the non working brain i get square signal up to first resistor and then triangular signal with a lot of spikes in it to LM386. So my sin/cos return is also spiked. Also sin/cos "potatos" are not expressed very good. At first i thought signal might get over 3V3 into AD, but i see the same level signal also on the first brain so i believe AD is not overwhelmed.

I will take some pictures later today.
User avatar
johu
Site Admin
Posts: 5682
Joined: Thu Nov 08, 2018 10:52 pm
Location: Kassel/Germany
Has thanked: 153 times
Been thanked: 960 times
Contact:

Re: ACIM motor with resolver testing

Post by johu »

That is very surprising, I didn't expect it to work with square wave excitation. So we wouldn't even need that low pass ladder it seems?
Support R/D and forum on Patreon: https://patreon.com/openinverter - Subscribe on odysee: https://odysee.com/@openinverter:9
arber333
Posts: 3241
Joined: Mon Dec 24, 2018 1:37 pm
Location: Slovenia
Has thanked: 74 times
Been thanked: 223 times
Contact:

Re: ACIM motor with resolver testing  [SOLVED]

Post by arber333 »

Well last week i finally managed to trace the weird resolver signal to seemingly good caps. They were from a different series than my usual ceramic caps. I went on and replaced them and i get good resolver signal. Also the pesky Lacrosse motor is now running really good with resolver and 100Vdc battery.
Post Reply