Re: [CAD_CAM_EDM_DRO] Re: steppers+feedback+emc

awallin25 wrote:

> > With that said if you are having a problem with missed steps
> >during normal
> > use the closed loop stepper is not likely to solve the problem.
> ...
> > I think you need to determine why you are losing steps and fix >
> the problem
> > there.
> >
>
> the 3d table is made by isel, so I think it is mechanically OK.
> I looked at the motors today, and they are isel model
> 473030, with a torque of 110 Ncm, 2,8V / 4.0A Bidir.
> I used a dremel tool with a 3mm mill for milling in MDF.
> as I said earlier, using this setup and a program of my own to
> drive the table it was constantly losing steps so thats why I
> thought feedback would help
>
> experienced guys: can you tell me if milling in mdf and/or
> aluminium should be possible with this kind of table ?
> is the simplest and best solution for this to just drive the table at
> slower speeds? (I did some rough calculations, and with slow
> speed the milling of a 3d part(ca500x150x20mm) could take
> many hours: 5-10h! with a slow speed, is this
> common/acceptable?)

Your real problem is likely to be the acceleration profile and or
smoothness of the step rate. Does your program have linear
acceleration ramps? How smooth are the step pulse trains?
What scheme do you use to produce the steps? Do you have a fixed
interval triggered by the real-time clock (PC timer chip) and
produce or not produce a step pulse at each of these? Do you
just have software timing loops? Are you sure there are no programs
(ie. motherboard memory refresh, mouse interrupts, etc.) that can
interfere with the timing of step pulses?

Can you run your program at extremely slow feedrates? Does that
clear up the problem? Can you run your program with a ballpoint
pen in the spindle and a piece of paper on the table? Does it still
lose steps? In other words, the first thing to establish is whether it is
speed or mechanical load that causes the lost steps.

Feedback will only prevent a program blindly continuing on (now in the
wrong place) when steps are lost. When steppers lose steps, they often stall,
and the encoder will not be of much help when it causes the computer to
tell the motor "go faster".

Jon