Using surplus servo motors as spindle drive

[Swede]

Hello servo gurus - there has been some speculation within the "Mini-mill" forum about the use of surplus servo motors as spindle drive units, either belted or direct. Servos are designed with a very high energy density, and I think they'd work pretty nice. AC or DC servos in the 800 to 1400 W range would be ideal, with top speeds above 5,000 RPM. For simple use What would be some characteristis in a servo to look for, and how could the speed be varied?

Brushed DC would be easy... use a 10A DC speed control. How about AC? Perhaps a suitable VFD? Would an AC servo respond to such a unit?

The servos would ideally be wound for 90VDC or so. Any thoughts are appreciated.

[Al_The_Man]

Swede, AC servo's or brushless DC have a lower inertia, higher RPM rating than brushed DC, Brushed usually run a max of 3000rpm tops, whereas BLDC and AC run typical 6000 to 8000 rpm.
A VFD will not run a AC servo, I had no luck when I tried it. I would keep an eye out on ebay if possible for a matched AC/BLDC servo motor amp combination with analogue input , and either a potentiometer for control or an analogue output from the CNC controller if it has one, if it does you can use S and M codes for control.
Al

[Swede]

Thanks Al. If a brushed DC had the guts, it could be belted, I suppose, into a higher RPM regime. The "matched pair" concept is a good one, especially one with an analogue input as you mentioned. This may require more investigation.

[ESjaavik]

Hey Swede, it seems our thoughts tends to cross.

I have Ebought a BLDC servo that I want to use on my lathe project. BTW: Do you know that the new KaVo HF spindles are actually BLDC synchronous motor technology?

Some servo drives come as speed controllers rather than position controllers. These are the ones you would look for. Often that is what you can Ebuy, as they often come from conveyor applications that are speed controlled. The difference can often be found by just one digit or letter in the order number of the drive. The real difference lies in the firmware inside the drive. I guess a huge number of buyers that did not bother to check first gets quite disappointed if they want to use it as a position drive.

What you look for in a motor is high RPM capability and low Kv. Power is the product of torque and speed. Usually torque is fairly constant up to max RPM. So you want to exploit this by driving it to max speed. The voltage needed is largely proportional to speed. So as to not get limited by the max voltage you have available, look for a low Kv.

For a high energy density a BLDC motor is to be preferred over a DC. The DC motor have it's windings deeply embedded on the rotor, so cannot easily get rid of heat. The BLDC have only magnets on the rotor, so the only heat generated there are a very low eddy current heating. This is where the BLDC beats the Asynchronous motor that have more current generated in the rotor cage, it's a necessity, as this kind of motor is actually a transformer with one rotating winding set. In the BLDC the windings are in the stator, so it can much more easily be dissipated to the surrounding air. If you add forced cooling you can increase the continous rating by as much as 2, while the max rating will stay the same, usually 4 times the uncooled continous rating. With DC or asynchronous motors the heating in the rotor cannot easily get out to be dissipated.

[Swede]

Thanks Einar, this is outstanding info, lots to think about!

[unterhaus]

Originally Posted by Al_The_Man Swede, AC servo's or brushless DC have a lower inertia, higher RPM rating than brushed DC, Brushed usually run a max of 3000rpm tops, whereas BLDC and AC run typical 6000 to 8000 rpm. A VFD will not run a AC servo, I had no luck when I tried it. I would keep an eye out on ebay if possible for a matched AC/BLDC servo motor amp combination with analogue input , and either a potentiometer for control or an analogue output from the CNC controller if it has one, if it does you can use S and M codes for control. Al

I think a considerable number of vfds probably will drive a brushless servo. I have some AMC drives that will work open loop, and the motors don't care. They work as a three phase sine wave generator in that mode. Actually, we just bought a VFD at work that is made to drive AC servos. And there are a couple of servo drives that will work with AC induction motors. Learning motor control is on my list of things to do. One of the grad students working in my lab is working on detecting impending failures in induction motors. I help him quite a bit, but he's way ahead of me in the motor control department.

What I have thought about doing was just using the servo drive itself. Seems like an ideal application. I'm pretty sure haas we have at work has a yaskawa servo spindle motor.

[arvidb]

Ok, I just couldn't contain my curiosity anymore, so I tried it:

I connected a 3 ph PM servo motor to a V/Hz VFD. I set max frequency (= maximum motor frequency), max voltage frequency (= rated motor frequency), and max voltage (= motor rated voltage), and also something called "electronic thermal relay" which is supposed to turn the motor off in case of an overcurrent situation. I set this to the rated current of the motor (2 A).

I set the frequency to 15 Hz and pressed "Run". Ok, the motor started to turn, but sounded kind of bad (squeaking noise). I checked the current, and saw that the *stupid* VFD pushed some 10-12 A into the poor motor! I turned it off before it burnt the windings, but I guess I now have a motor that's a lot weaker than before (max peak current for this motor is 6 A - poor magnets!).

I have no idea why the overcurrent protection didn't work. Perhaps the VFD performs some kind of integration of current over time to allow for short term overcurrent?

Well, at least now I know how well it works...

Arvid

[Stevie]

I have a 1/2hp VFD and a few 330watt servos; on 220 single phase the VFD will power 1/3hp; so i thought it should be a match; but after reading this I'm not sure
The Toshiba VFD I have looks to be in great shape; but the servos are NIB; hate to melt one

[Swede]

Arvid, I tried it too on a little Omron servo that seems to be a pure AC motor, just three phases with no commutation. At lowest speed, the motor actually turned over, but slowly. As I cranked up the frequency, the servo just grumbled and complained and did not go faster. The servo is a NEMA 23, 200V AC servo. Current seemed constant as I messed with it at 2.5A, which seemed about right. The servo simply didn't work.

I hope you didn't trash that servo!

[ESjaavik]

@Arvidb: My guess is your motor have too little inductance for your VFD. Then the current rise is too fast for the VFD. How did you measure current? The demagnetizing current should not occur for more than a fraction of a second. If you got 10-12A RMS, your current peaks are much higher.

If making a spindle drive out of a BLDC servo, I would go for a PAM drive. It can shortly be described as a filtered output PWM DC-supply followed by commutation transistors. It does not use the motor windings as a component of the drive like the PWM drive.

[arvidb]

Einar, the VFD has got an output current readout. I simply used that.

It ran for maybe three seconds before I could turn it off. Afterwards I could feel the motor getting hot even after I had turned the power off, so the windings must have gotten quite hot.

I connected the motor to its proper drive and put it in dynamic brake mode, and I can feel no difference in the amount of "brake" compared to another (non-abused) motor. I guess that's a good sign?

Swede, any idea why it didn't work? What kind of drive scheme does your VFD use? V/Hz, some kind of vector drive, ...?

Arvid

[basrottier]

Hello,

Has anybody some experience with controlling KAVO motors? They can't be controlled by a standard frequency controller, as they run at very low voltages <50V. the controller would have to supply PWM as a normal frequency convertor but at low voltages. power : rather limited (up to few hundred watt)

Bas