I am doing a project in which I need to control a servo motor. I use a drive/amplifier with only RS232 port to communicate with PC via a special software. But I have to control it via an axis card (+/-10V analogue output), so can I send this signal (+/-10V analogue output) to RS232 port as reference signals of amplifer and then drive the motor?

Thanks.

 

Pay attention that you may control at drive/amp in velocity or current loop only with (+/-10V) analogue output.

 

The amp probably has an analog input for reference
voltage - it probably then translates that voltage
into a velocity or torque command. If it does not and if you do have to go through the RS-232 port then control of the motor will be difficult. If that is the case and the amp offers the option of velocity mode OR torque mode then place the amp in velocity mode. Controlling torque at serial comm speeds will be very very difficult. You can get by with control loops <1 kHz for velocity mode.

Davis Gentry

 

Mr. Gentry,

Would you say it is better with a motion controller, like the UMAC, to control digital servo drives that are setup for...velocity or torque control? It has been my line of thought that if the motion controller can handle it then just put the drive in torque control and let the motion controller do the position and velocity loops. Is this correct? I personally have found that from experience it makes for an easier system to setup and troubleshoot. But, I would assume this puts additional burden on the controller, so in some applications with alot of axes, it might would be wise on non-critical axes to put them in velocity mode to ease up on the motion controller? Correct? Just wanted to pick your brain and get your thoughts.

B

 

Many of the motion controllers currently on the market can easily handle the torque loop as long as they can update that loop more quickly than 1 kHz - in my experience 1 kHz is a bare minimum for torque control with any precision, and I greatly prefer to run the loops at 2 or more kHz. Our default is 2.2 kHz, and I typically double that for any of our Turbo processors (such as the UMAC you mentioned) in almost all applications. I have run them as high as 80 kHz for very high bandwith motors such as voice coils. While I do not have enough knowledge to comment on the performance of all of the controllers out there I do know that I can comfortably control 32 torque mode amplifiers with the UMAC, while the processor also handles trajectory generation, kinematics calculations, lookahead, communications with an HMI (if used) and the machine logic. Our older non-turbo boards can do the same with up to 8 axes, minus lookahead and kinematics (though you can run your own kinematics in machine logic and I have done so).

Davis Gentry
Delta Tau Data Systems

 

To answer the original question, no, you do not send the analog signal using RS-232. The analog out from the axis card gets wired to the analog input on the amplifier. The communications port is typically there for configuring drive parameters. I have seen drives that allow you to set a velociy or torque over commms, but it is not anything you would want to use for position control.

 

So if I have a torque update output of the controller at 5kHz and the torque update input at the drive at 16kHz I should have very good control... setting aside other system variables, of course. Is there any value in having the drive loop update faster than the controller output?