Skip to main content
Submitted by Dave Gleason on Mon, 06/21/2004 - 16:15

Does anyone have any experience tuning Aerotech linear stages with Aerotech BA sinedrive amplifiers. The values that I using for KD,KP and KI are really low (2,3,2). Regardless of what I try the system never really demonstrates a good square wave. Any thoughts?
Thanks for the help
dg

Comments 3

EricK on 07/07/2004 - 09:59

Dave,

With such low gains (i.e. KD, KP, and KI), I suspect that the resolution of your system is very high. Linear stages with <0.1 um resolution will often require that the Galil is configured for a higher servo loop rate. The "TM" command will configure the controller loop rates. I suggest TM250. The minimum TM setting depends on your controller model, consult the user manual for details.

In addition, since you specified KD and KP gains of nearly equal magnitude, I suspect that the amp is in a velocity control mode. For best results, we suggest the amplifier is set in 'torque" mode. This way the Galil servo loop does not conflict with the amplifier's velocity loop, which often causes instability.

-EricK

dorundo on 01/05/2006 - 13:53

I have just encountered a similar issue trying to tune an Aerotech linear air-bearing stage with linear motor and Aerotech BL Linear amplifier.

I am replacing an existing Anorad linear stage on rails with the Aerotech air-bearing stage. The previous encoder/interpolator had a resolution of 10000 counts/mm. Previously, we were using the MBA-6-60 Galil amplifier, which accepted the +/- 10V signal from the Galil directly.

The new linear encoder and interpolator combination provide 100000 counts/mm.

I am using a DM-1520, and have already set the timebase to it's minimum of 375 usec.

I used KP,KI,KD values of 30,150,1 before (with timebase of 1000 msec). The 10x multiplier on the encoder value should scale these to 3, 15, .1.

Right now I'm running with values of 3,2.5,.33333 (which when scaled using new timebase are 3, 6.625, .125). I should probably be using a smaller value for KI, but since the minimum resolution for the DMC-1520 is .125, I can't use a smaller value.

I haven't gotten any values from WSDK that work any better than the ones I'm using above.

Is there any other way to scale the PID parameters so that I use smaller KI and/or have a better resolution than .125 on the parameters?

EricK on 01/06/2006 - 09:00

Due to the fact that the DMC-1520 is an older Legacy series product, you have probably already gained the maximized performance from this controller. While increasing the TM rate does allow for better headroom and resolution for the KD term, there is no method to increase the resolution of the KP term on this product. And at TM375, you have maximized the servo rate already.

For these types of applications (high performance and resolution), Galil has introduced the Accelera series DMC-18x6 controller, which is our �next� generation controller. The Accelera series features an all new RISC based processor that allows servo rates as low as TM31.25 (32 kHz) as well as increased PID resolutions by a factor of 4. Currently the Accelera series comes as a PCI card (18x6), however we will likely introduce a standalone version in the near future.

For more information on the Accelera series, go to:

http://www.galilmc.com/products/accelera/dmc18x6.html

-EricK