[SwampDonkey]

Hey guys,

I have a Kollmorgen 3.23kw 3ph brushless motor. It came from a Doosan CNC machining center and is equipped with a resolver I believe (I have no experience with resolvers). When the motor was damaged and had to be replaced, I was shocked at how small it was. It has an incredible amount of power for its size. The body is only about 6" long with the resolver removed and it's frame size is a bit bigger than a nema 34.

I'd like to see what all would be involved in using this as a spindle motor. It puts out more power, both speed and torque, than my current stock spindle. Finding a driver may be a bit of a challenge, but the Argon servo drive from Granite Devices can interface directly with the resolver.

Another way to do this would be to ditch the resolver and go with a standard brushless encoder. This may expand my options to other drives that may be cheaper.

What would you guys do?

[Al_The_Man]

I have converted similar motors to encoder with simulated hall effect tracks, you need to know the pole count and also carry out an alignment using a double beam 'scope.
If you short the three stator wires and turn it one rev and count the 'bumps' will tell you the number of poles.
There are quite a few analogue inpujt drives out there, AMC etc.
Al.

[SwampDonkey]

I have converted similar motors to encoder with simulated hall effect tracks, you need to know the pole count and also carry out an alignment using a double beam 'scope.
If you short the three stator wires and turn it one rev and count the 'bumps' will tell you the number of poles.
There are quite a few analogue inpujt drives out there, AMC etc.
Al.

I have converted similar motors to encoder with simulated hall effect tracks, you need to know the pole count and also carry out an alignment using a double beam 'scope.
If you short the three stator wires and turn it one rev and count the 'bumps' will tell you the number of poles.
There are quite a few analogue inpujt drives out there, AMC etc.
Al.

Thanks AL,

And yeah, I kinda wanted to go with AMC for this project. I can see why you're a big fan of them. They seem to have the most bang for your buck on ebay and most will accept the +/-10V thats feeding my VFD right now.
Ive seen and worked with AC servos in this power range. This one is 4500rpm and an unknown number of amps. Maybe you can decipher this:


Attachment 320738

Learning how to convert these units is on my to-do list. Im looking into drives to power this baby, but even among AMC drives they seem inda sparce on ebay at this power level.
Can you decipher all this?

And as always, thanks Al
SD

[kstrauss]

Google for akm53k kollmorgen and you'll find data sheets, pricing, etc.

[SwampDonkey]

Google for akm53k kollmorgen and you'll find data sheets, pricing, etc.

I havent come across a price for a new unit yet. I know it was prohibitively expensive vs shoving the whole machine into a corner.

[Al_The_Man]

Those motors are AC sinusoidal, using the commutation tracks converts it to a BLDC, which does work in order to use a AMC etc drive.
BLDC is usually a bit more torque, just not so smooth at low rpm when used open loop as in a spindle etc.
Al.

[kstrauss]

A quick look finds:
Used AKM53K for $350 (Kollmorgen 3 23KW Brushless Servo Motor AKM53K CCCNR 00 Working But Damaged | eBay didn't sell) and relisted for $25 (Compact Kollmorgen 3 23KW Brushless Servo | eBay with no bids yet)

New AKM53K on ebay for $950 (Kollmorgen PM Servomotor AKM53K ANCNEM00 New | eBay 6 available but none sold)

[DLawrence]

I have converted similar motors to encoder with simulated hall effect tracks, you need to know the pole count and also carry out an alignment using a double beam 'scope.
If you short the three stator wires and turn it one rev and count the 'bumps' will tell you the number of poles.

Al.

Hummmmm ....thinking ...... ? ? ........ short the three stator wires to what ? Each other ? the case ? Then where would the scope leads go ? My guess would be short all three stator leads together , then the scope probe on that bundle ........and scope ground to case ?

Thanks
Dave

[kstrauss]

I think that he was describing two procedures:
1) Short the wires together, rotate the shaft and count the bumps where it gets hard to turn and then easier to determine the number of poles.
2) Use the 'scope to align the encoder to the bumps.