PWM servo amp switching frequency

[ScottI2R]

Hi all,
We have been currently driving a brushless DC servo with an amplifier that was recently obsoleted. This amp had a PWM switching freq. of 40 Khz. We MAY still be able to get a similar model that switches at 20 Khz. What I would like to know is 1) How critical is this switching frequency as far as performance goes? 2) What can I expect to see by using a substitue that operates at a different switching frequency?

This servo is used in an EDM head application. The overall mass of the moving components is quite small and response of the servo is critical to optimal performance.

Thank you in advance for any assistance.

Scott

In a hundred years, it isn't going to matter anyway.

 

[sreid]

20 kHz is a common switching frequency for servo amplifiers using IGBT power transistors. 20 kHz can't be heard and most IGBT are optomized for 20 kHz operation. Lower voltage/lower amp amplifiers use FET transistors and can switch at higher frequencies.

The servo amp current loop bandwidth is limited to about 1/10th the switching frequency but the actual bandwidth depends on the motor inductance and currnt loop compensation values. For high performance, this is more important (usually) than the switching frequency.

 

[blacksea]

The current loop bandwidth is 1...2kHz as a rule, i.e. any problem for even 20kHz switching frequency.
But very important issue that must be checked - current ripple defined by motor inductance. The some PWM amp manufacturers limit it - see Elmo (

http://www.elmomc.com)

for example.

 

[ScottI2R]

Hi and thank you guys,
I have found that the higher switching freq. gives a better efficiency for the amp. However, one bit I read stated that power output ratings were reduced when using the models with 40 or 60Khz switching.Hmmm..have to look into this. Thanks again.

Scott

In a hundred years, it isn't going to matter anyway.

 

[UKpete]

Scott, the losses in the motor are probably higher with the higher switching frequency, this may be the reason for the derating at 40/60kHz.

I have seen this myself on very high speed PM motors, the higher the PWM frequency the greater the eddy current losses in the rotor magnets. Any PWM-induced harmonics in the stator currents will set up a rotating field at that frequency, rotating faster than the rotor. The higher the frequency the greater the induced eddy currents in the rotor as per Faraday's law. In addition, the iron losses in the stator may be higher too.

Rotor eddy-current loss not an issue of course if the rotor can and magnets are non conducting.

 

[blacksea]

The incresing of amp's power rating with higher frequencies relates with increasing of amp power stage losses in switching devices (IGBT or MOSFET).
More details regarding motor power losses I found at:

http://www.drbrushless.com/articles.html#A2.