Current turning capabilities status

emcPT wrote: The "Unidrive SP SP4202 18.5kW Inverter" is ... expensive. But if it would allow me to use the current hardware and also the C axis, it would be at least something to consider.

The SP4201 is 15kW and disproportionately cheaper:
www.motorcontrolwarehouse.co.uk/unidrive-sp4201/prod_1124.html

Kirk_Wallace wrote: Looking at the schematic, AG215 looks like it shorts X, Y, Z. This would brake? AG214 parallels U, V, W to X, Y, Z, and as mentioned, changes speed? I wonder what XTM means?

I'd sure like to see the rest of the dashed (actually ghost lined) box labeled spindle motor to see what else is considered part of the motor.

Only more a over temperature sensor.

PCW wrote: I think emcPT is correct, the relays select delta or wye connections for different speed ranges.

To determine whether it is a PMSM servo motor or induction motor, you can also short out one winding (probably you would have to short U to X? if the relays are disconnected, in any case use an ohmmeter to find a winding to short) and turn the spindle. It will have a lot of resistance to turning if a winding is shorted and its a servo motor, but shorting will not make any difference if its an induction motor.

Doing a more depth research it is clear that it is for high speed and low speed. The electrical schematic refers that.

Regarding the suggested test (I did this one and not what andypugh sugested as I did not had any power supply at hand).

1st) I shorted U to Z; V to X and W to Y so that the motor is set up as in high speed

2nd) I shorted V to W (so V, W, X and Y were shorten) and no difference was noticed.

May I conclude it is an induction motor? If so would be safe to apply a low voltage to the motor to see if it spins? And if it is really an induction motor how can it be controller to accuratly move 0.001 degrees for milling? Maybe an encoder, but really I cannot find none coming out of the spindle (at the same time it is full of dirt and maybe I missed it).
So many questions, sorry.

andypugh wrote: One way to tell might be to connect a low current supply to two of the wires (say 1A between U and V) and see if you can feel any "notches" as you rotate the motor.
I don't think you will feel any with an induction motor, whereas you will with a permanent magnet motor.

Does the rest of the wiring diagram show any more wires from the motor/spindle to the drive? If the only other wires from the spindle are an incremental encoder then I think it probably is just an induction motor.

This does mean that just about any VFD will work.

I could not found any more wired from the motor to the drive, but I could miss it. The inside of the machine looks like a swamp (oil dirt ...).

If it is an induction motor I might well do the following:
Use a standard VFD for 15Kw, no fancy stuff; make the machine work without Caxis. If the Caxis is important on the future I can add a servo to the spindle and use the servo while the VFD is disconnected. This way I would have a motor for positioning and a motor for fast turning.
This would also lower my initial investment and help me in the initial work.

emcPT wrote: I could not found any more wired from the motor to the drive, but I could miss it. The inside of the machine looks like a swamp (oil dirt ...)

If it had a C-axis then there must be an encoder. However it might not be built in to the motor, and might even be belt-driven.
The controller might be the "missing link" between the encoder and the motor.

Use a standard VFD for 15Kw, no fancy stuff; make the machine work without Caxis. If the Caxis is important on the future I can add a servo to the spindle and use the servo while the VFD is disconnected. This way I would have a motor for positioning and a motor for fast turning.
This would also lower my initial investment and help me in the initial work.

A good VFD ought to be able to work anyway (I keep pushing the Unidrive ones as I know that they can work this way, but any "flux-vector" VFD might work if controlled and set up correctly.
I experimented a while ago with a very old, not flux-vector VFD and _nearly_ ended up with a workable system.

I think any modern VFD would work a lot better, and I could probably have made mine work better with some more tuning (inverse deadband, for example)
Do you have any small VFDs (typical 1kw size) that you could experiment with?

I do not have any VFD for the 220V range as the mains here are 400VAC, but I can buy it.

Now the flux-vector drive seams a good way to go, although I have concerns if it could control accurately like 0.001 radians, but that is something I did not read about it and certainly will. I will try to talk with a supplier that I bought several things from him and see if I can test a unit before I buy it. The main problem will be the voltage. 220V they do not stock.

Filipe