Gerry Chu wrote:

>Can it also do torque control (I think this is the same as velocity 
>control)? Will it output faster than every millisecond?
>
>  
>
Torque control is where the CPU outputs a current value, and the amp 
sees that the
motor gets exaactly that current, which is proportional to torque.  If 
you have a torque-
mode servo amp, EMC will handle that, with appropriate tuning.  Velocity 
control is
where the CPU sends a velocity command to the amp, the amp compares 
actual velocity
by way of a DC tachometer, and sees that the motor moves at that speed.

EMC can go quite a bit faster than one millisecond, 10 KHz is easy, 
25-50 KHz is probably
the limit with current PC systems and interfaces.  (Just to be clear, my 
parallel port
interfaces probably top out around 10 KHz due to the slow port.)

>Let me make sure I understand: there are three cases:
>
>1) DAC: IO card gets a digital number representing voltage from EMC. IO 
>card uses onboard DAC to convert to voltage. Voltage comes out of card, 
>gets amplified by external amp, then goes to motor.
>  
>
Well, in a vleocity servo system, it might be important to know what the 
amp does with
these signals.  Actual velocity from the DC tach is summed with the 
velocity command
from the DAC.  This error signal is summed with motor current and 
develops another
error signal that is then amplified, usually with a PWM system, and 
feeds a voltage to the
motor.

>2) PWM-card: IO card gets a digital number representing voltage from 
>EMC. IO card converts to PWM signal using PWM generator. Then gets 
>amplified somehow...? What's the difference between a PWM generator and 
>a PWM amplifier?
>  
>
A PWM "generator" produces a low-voltage pulse (ie. 0 to 5 V) of varying 
width.
The PWM amplifier amplifies this pulse to higher voltage, like 0 - 100 V 
with the
ability to deliver many amps.  It usually has current limiting also, as 
the computer
usually doesn't get sent a current reading.

>
>
>I heard that linear amplifiers are "smoother" than PWM amplifiers. 
>  
>
If you mean linear, as in no switching circuits, I can tell you this is 
not true, when you
talk about moving motors.  If you need a bandwidth of 0 - 50 KHz, there 
may be some
truth to this, but typical motors have bandwidth of a few hundred Hz, tops.

If you mean "linear" as in analog-input velocity servo amplifiers, this 
may well be true,
especially in the case of low encoder resolution.  Without velocity 
feedback, then the
system breaks down into discrete steps equal to the encoder resolution, 
when the
encoder counts come at some low rate.  The exact point where this occurs 
depends on
proportional gain and other tuning parameters of the servo system.  
Because of the
time-continuous nature of the DC tachometer, the proportional gain of 
the velocity
servo amp can be set much higher, and the loop stays closed even when 
the encoder
produces no counts.

>Thoughts? Since smoothness is important to haptics, I think I should go 
>the DAC route. Is my reasoning reasonable or am I completely misinformed?
>
>  
>
The DAC will only help you with a tachometer and velocity servo amp.  A 
torque-mode
amp with analog input will not do any better than my PWM system, for 
instance, since
it necessarily goes open-loop when there are no encoder pulses coming in 
to define the
position.  It may be much easier and cheaper to go with higher encoder 
resolution.  If the
distance between adjacent encoder counts is so small it can't be felt, 
then the user can't
tell the difference, except by audible whines from the servo motor.

Jon

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to