bldc running a Megatorque servo?

Hi, all;

Just wondering if anyone has even thought about trying a Megatorque servo without the amp under the LinuxCNC HAL. I have acquired (cheap) a rebranded NSK Megatorque servo and driver which is unfortunately missing one of its cables. I suspect that it's not too usable because NSK seems very specific that the driver, servo, and cable are tested and tuned together. That may just be them encouraging people not to buy other brands, though. This whole setup seems to date back to about the mid to late 1980s, although it seems very well made.

My particular servo is an RS type, I think an RS0608, and seems to have windings labeled A+, A-, B+, B-, C+, C- and E on the one cable that I do have. Per the docs these are two wires each for phases A, B, C, and a ground(E).

It also has a feedback connector with five pins, which the NSK manual labels as REA, REB, REC, Common, and frame ground. These are supposed to be resolver phases A,B and C plus common and ground. A three phase resolver?

I'm not sure amp I have works, and I can't test it without a good deal of effort because it uses some very old connector styles I haven't seen anywhere else. I can't find the correct replacement cable set anywhere, and I suspect if I did it would be priced rather higher than I'd care to pay.

Is there any chance this servo could be controlled with an H-bridge and resolver feedback? Should I try to find a more modern/working servo drive with a cable set that fits? Should I sell the whole mess and forget about it?

Thanks for any input,
Erik

green751 wrote:

Hi, all;

My particular servo is an RS type, I think an RS0608, and seems to have windings labeled A+, A-, B+, B-, C+, C- and E on the one cable that I do have. Per the docs these are two wires each for phases A, B, C, and a ground(E).

Yes, a standard 3-phase motor winding for bruhless motors.

It also has a feedback connector with five pins, which the NSK manual labels as REA, REB, REC, Common, and frame ground. These are supposed to be resolver phases A,B and C plus common and ground. A three phase resolver?

Yes, these are sometimes called a "syntach". The stator windings of a resolver can have any arrangement,
and the 3-phase version is less common, but some outfits use them.

Is there any chance this servo could be controlled with an H-bridge and resolver feedback? Should I try to find a more modern/working servo drive with a cable set that fits? Should I sell the whole mess and forget about it?

Erik

Not a single H-bridge, it will take 3 of them. And, if the resolver is 3-phase, not a lot of generic
resolver converters will handle that. it may make sense to pull the resolver and install
a "6 track" encoder with built-in commutation outputs.

Jon

jmelson wrote:

Yes, a standard 3-phase motor winding for bruhless motors.

Hmm.. okay. I'm used to the UVW three phase wiring, which I guess is a wye type system? I don't think I've seen three phases with two wires each before.

jmelson wrote:

Not a single H-bridge, it will take 3 of them. And, if the resolver is 3-phase, not a lot of generic
resolver converters will handle that. it may make sense to pull the resolver and install
a "6 track" encoder with built-in commutation outputs.

Jon

Okay. I'll see what's involved in pulling this thing apart. Hopefully the resolver is modular rather than integral. I do have a spare harosyn resolver I could install, or I have some 5000ppr encoders with commutation tracks. Given that this would be direct drive I'd guess I'd want higher resolution than that, though.

3 H bridges isn't a show stopper. However, is it possible to wire the three phases together UVW style and (assuming proper commutation synced with the windings) run this servo using a normal brush less drive?

Thanks again,
Erik

green751 wrote:

Hmm.. okay. I'm used to the UVW three phase wiring, which I guess is a wye type system? I don't think I've seen three phases with two wires each before.

Yes, this gives you the option of Delta or Wye connection. For six-step drives, I'd definitely only
use the Delta connection. For a sinusoidal drive you could use either.

Okay. I'll see what's involved in pulling this thing apart. Hopefully the resolver is modular rather than integral. I do have a spare harosyn resolver I could install, or I have some 5000ppr encoders with commutation tracks. Given that this would be direct drive I'd guess I'd want higher resolution than that, though.

3 H bridges isn't a show stopper. However, is it possible to wire the three phases together UVW style and (assuming proper commutation synced with the windings) run this servo using a normal brush less drive?
Erik

Yes, definitely so.

Jon

If this is the type motor I have seen I would try using the original drive...

Its a 3 phase high pole count (20 or more poles) variable reluctance motor
with integral 3 phase resolver, designed for applications like direct drive robot
joints.

Variable reluctance motors have very different drive circuitry than BLDC type motors
(U/V/W wiring will not work)

This agrees with my initial exploration into the casing.

The motor is about 8 inches across and seven tall, shaped like a squared off torus. It looks about like this: thumbs3.ebaystatic.com/d/l225/m/mNN9MRDySMuHG7eOtgQivWg.jpg

I took off the top dust cover and found what I think is an example stator/rotor. It kind of looks like the inside of a stepper, they both resemble very fine tooth ring gears with the points facing each other. I didn't see any obvious resolver, and I suspect it's built in to the center of the stack of rings somewhere.

Soo.. I guess I'll have to see if I can get the drive working. One odd thing is that it seems to want 100 volts AC, not 110.

Does anyone know what type the blue connector is in the pictures of this drive?
www.ebay.com/itm/NSK-MEGATORQUE-DIRECT-D...DRIVER-/161097441437

Edit: Found a better representative, this is essentially what my drive looks like:
www.ebay.com/itm/NSK-EE0408C05-25-Servo-...&hash=item19edea1c9f

My version of the drive has a small one of those for CN3 as well as CN2... I can probably stick some soldered pins in it for testing purposes, or maybe I should just replace the connector with something easier to find? Odds are pretty good I'll have to open up the drive anyway.

The blue connectors are made by Honda and at least the common sizes are easily available