Brussless AC Servo Motor ?

[trungsy]

I have two linear actuators type ISPA from IAI Corporation Japan. Please refer to : http://www.intelligentactuator.com/p...SP/Default.asp.
These actuators intergrated with Brussless AC servo motor. I can not see specification of the motor on the internet.
Please tell me how to design driver for the motor?
Thanks!

[DennisCNC]

Thats simple, take a microchip dsPIC microcontroller and IRF IRAMS, write some code for the pic and wala you got a brushless controller.

Or just buy one.

[unterhaus]

what identification is there on the servo motors?

[trungsy]

there is no identification on the servo motors. Because the motor is mounted with linear actuator by factory. Please refer to website which I Listed above.

[unterhaus]

I looked at the web site, and the picture of the actuator doesn't show a servomotor. I assume if you take the cover off, you can see the servomotor. Looking through the catalog for the parts, it seems there is more than one servomotor you can have.

[trungsy]

I saw the servomotor when I took the cover off. But I don't find the information of the motor on company website. This is photo of my linear actuator with the servo motor:

[andy55]

Originally Posted by trungsy I saw the servomotor when I took the cover off. But I don't find the information of the motor on company website. This is photo of my linear actuator with the servo motor:

Take an oscilloscope and look at the signals from all the wires from the motor.

If the motor winding signals(back EMF) look trapezoidal or sinusoidal that will tell you what kind of drive you need.

AC servos usually have hall signals (three) for commutation, each individual hall signal will be on for 180 electrical degrees and off for 180 electrical degrees.
The number of electrical cycles per one mechanical rotation of the shaft will tell you how many poles your motor has.

Your picture shows an encoder which probably has standard quadrature A and B outputs and maybe also an Index output.

I don't know how you would determine the voltage and current for your motor. Possibly by putting it in a dyno and seeing at what rpm/voltage the torque starts to drop off.

I've been working on a drive for my own AC servos, see
http://www.cnczone.com/forums/showthread.php?t=20993
version 2 is ready and much nicer looking than the first prototype

[unterhaus]

I looked at the picture for a while and did not see any halls. I downloaded the catalogs from the company and it didn't answer my questions about this. There are two heads on the encoder, which would suggest an index. If the pink wire at the top has is one of 4, that would be halls. Nowadays, it's quite possible that they are using sensorless drives and got rid of the halls altogether.

Freescale semiconductor has some of the best appnotes about motor drive. The easiest way to get there is through their DSP section.

[andy55]

Originally Posted by unterhaus I looked at the picture for a while and did not see any halls. I downloaded the catalogs from the company and it didn't answer my questions about this. There are two heads on the encoder, which would suggest an index. If the pink wire at the top has is one of 4, that would be halls. Nowadays, it's quite possible that they are using sensorless drives and got rid of the halls altogether. Freescale semiconductor has some of the best appnotes about motor drive. The easiest way to get there is through their DSP section.

Once you know the position of the rotor/stator relative to the index pulse I see how commutation could work only based on encoder counts.

But how is this relative position determined initially ?? (maybe I need to take a look at the freescale docs...)

[unterhaus]

the way I understand it, you give each leg a pulse and then use any number of methods to figure out which one is pulling hardest on the rotor. We have a very nice impedance meter at work, and we used it on an induction motor. The impedance changes significantly as you rotate the motor. The sensorless drives take advantage of this.

My understanding of "sensorless" is that the sensors are replaced by a low value resistance in the return path from the H-bridges. This allows you to measure the current in the three windings with only one measurement.

I know Freescale has app notes about sensorless drive, TI probably does as well. Another place to look is Analog Devices, but I haven't been there in quite a while. Not sure if International Rectifier has app notes about sensorless drive, but they might.