Re: [CAD_CAM_EDM_DRO] Re: Re: EMC interpreter was Re: BoBCad Question

Ray wrote:

> Randy
>
> From: cnc002@...
>
> <S>In this case the radius compensation doesn't present a big problem as
> >that feature on the old A-B controls of this customer never worked
> >properly anyway and they just do the toolpath in BobCad.
>
> <rant>
> I have real problems with both the generalization about AB and the
> generation of tool comp in a cad program. 1- AB always seemed to do
> adequate tool comp for the operators I know and the machines I own.

I started with an AB 7320, a 1975-vintage control. The cutter radius and length
compensation worked exactly as the programming book said it did. Now, circular
interpolation could be screwed up by not following the book, and it would zing
off or make spirals, but the offset comp worked solidly.

I will note that there is real art in setting up the cutter lead-in and lead-out paths
for radius compensation, and if you do it wrong, you get gouged parts or error
messages. This is true for AB or EMC, and they work much the same.

> (emc
> does good comp if you treat it right). 2- Comp built into a program is okay
> but what happens when your operators have to replace the end mill at
> midnight and the cad operator isn't around. Perhaps what you were seeing
> was a user problem rather than an AB control problem.
>
> >I know that EMC is supposed to support servo control but have
> >heard that is doesn't work very well.

Well, have you actually talked to someone using EMC with servos? I am, and
I am immensely satisfied with the servos in general, as well as how well EMC
performs with them. (I will note I am using true velocity servos, with velocity
command DACs, not pseudo servos with no velocity feedback and sending step
pulses to the "servo amps".)

>
> You must be listening to competitors or idiots!
>
> >I have downloaded a copy of EMC and am working on a prototype control
> >using it to see what features are really there and if a true servo, closed
> >loop system can be developed around it.
>
> Good. Perhpas you will detail the nature of your test bed along with your
> results.
>
> Jon E. has been using such a system with the EMC for more than 18 months.

Now, over 2 years.

> He uses the servo to go board on his system so that the loop is closed in
> the EMC pc itself rather than in some outboard device. His new EPP board
> at names ran a closed loop servo just fine. If there were any problems
> with the EMC and closed loop motion, my guess is that he would have been
> the first to say it.

Since I make parts with the machine, if it was suffering from inaccuracy or
rough motion, I'd sure notice it. I do a LOT of circular interpolation and tapping
with a CNC tapping head (not hard tapping yet, no spindle encoder) and it works
great!

Jon