Re: CNC hydraulic test

Hello,
My very first machine conversion, in nineteen seventy-mumble, was of
a nice old tracer mill. I copped out and installed geared-down (big
timing belts) big steppers, ball screws, and a spiffy new controller
called a Bandit. In retrospect, I think I goofed and wish I had that
old mill back.

Of course, what is easier now, although still difficult, is to solve
the control stability versus following accuracy problems using
digital controls. Remember that with the proportioning valve the
computer commands velocity rather than position. Short answer: since
the real issue is one of tuning nested-loop controls, all a quick
test of the valves will tell you is the flow/carriage speed at "full
throttle", which you could get from the specs. More informative
performance tests (step response) would take considerable equipment
plus setup and testing time. It almost certainly can be made to work
well; it just might take rocket science caliber controls.

I still envy your opportunity.
Fred Renner

--- In CAD_CAM_EDM_DRO@y..., "Sven Peter, TAD S.A." <peteryco@r...>
wrote:

> Hello Friends,
> For years I dreamed of converting the hydraulic control of
> my copy lathe into CNC.
> Right now I have 2 electronically driven proportional valves
> sitting on my table together with the driver boards.
> The client permits me to make a test for my project before I shall
> install them in his machine.
> The input is -10V...0V...+10VDC but I have no computer with

interface

> cards.
> Has somebody an idea how I could make me a simple test set-up to

find

> out
> if these proportional valves would do an acceptable job in an CNC
> application?
>
> In THOSE days when I studied it was said it would go OK to make a

CNC

> machine with hydraulic pistons. Now everybody goes with ball screws

and

> server or stepper motors. Might be I am stubborn but at least I want
> want to have made
> a test before I go to ball screws.
>
> Good luck to everybody and enjoy northern summer we have the rain

again

> in Costa Rica.
> Sven Peter
>
>
> [Non-text portions of this message have been removed]