Re: RESOLVED: EMC is poorly designed and implemented.

["Les Watts" <leswatts-at-rabun.net>]

Very informative post Keith.

I bit the bullet and bought the STG II and am hooking it up
but I have also been looking at the EMC source code.

As I am a commercial operation I chose not to use EMC's
limit/estop system. I have a hardwired one that breaks the
servo/ amp connection via a snubbed contactor that shorts
the servos through a low value resistor for braking. All the
"panic buttons" around the machine and the limits are logic
ored (series).At the same time I send an inhibit to the amps and a fault to
EMC, although it should fault from following error pretty quickly anyway. I
am not using EMC limit over ride.I think others have done this. I have a
schematic but I want to test the system before I really show it to anyone.

I still plan to use EMC's home function, but I wonder if I should
strip out some of the limit polls. Prob in the same byte so I
could not easily.

As far as the PID section of EMC, I have been looking at that
code also. Adaptive tuning would be nice. I find that the simple
difference equations would seem to result in a single pole
differentiator that might be very noisy with much gain. I wonder if people
have trouble when trying to use much D.
I more typically see a FIR digital bandpass filter for
such a function. Filter coefficients can be changed on the fly for adaptive
control. Or... perhaps I just did not see it. There are a lot of files
there.

I am giving EMC a go, but I hope it it can be a high performance controller.
No reason for it not to be other than
just getting it done if needed- after all it is mostly all soft.

The best part of EMC appears to be the rs-274 parser. I might find the need
to use it as a standalone with something
else. I wonder how portable it is.


Les
Leslie Watts
L M Watts Furniture
Tiger, Georgia USA
http://www.rabun.net/~leswatts/wattsfurniturewp.html
----- Original Message -----
From: "Keith R. Bolson" <krbolson-at-visi.com>
To: "Multiple recipients of list" <emc-at-nist.gov>
Sent: Wednesday, January 30, 2002 11:29 PM
Subject: RESOLVED: EMC is poorly designed and implemented.


>
> RESOLVED: EMC is poorly designed and implemented.
>