Re: Dial indicator say machine is accurate in MDI screen. But out when

--- In CAD_CAM_EDM_DRO@yahoogroups.com, "apples0_17" <apalais@...> wrote:

>
> Dial indicator say machine is accurate in MDI screen. But out when
> cutting.
>
> Okay. So I have been mucking around with my plasma/router machine. I
> am trying to get some positional accuracy out of it. It is maily just
> a CNC plasma table. I put a router in it so that I could take more
> accurate measurements, as measuring a plasma cut part has other
> factors to deal with.
>
> So I have been cutting MDF. I tell the machine to move 100mm but it
> will be out. I clamped my dial indicatior to the frame set it all up
> at 0 etc. Went into the "Set axis steps", the automatic thing. I
> tell the machine to move 5mm. After a few goes I get it spot on!
>
> I can then go into the MDI screen and tell it to move 5.00mm and it
> will move the 5.00mm. Then I can go forward another 3.54mm and it
> will do that fine. Then tell it to go back to 0, and the dial
> indicator will be spot on the 0 mark.
>
> But. I go to cut a bit of MDF with the router 100mm long and it will
> be the incorrect length.
>
> What bugs me is that it is working in the MDI screen, ie move 10mm it
> will. Even move 300mm and then back to the 0 the dial indicator will
> be on 0. I was impressed with this.
>
> So, after a bit of reading here ther and everywhere, I decide that it
> could possibly loosing steps when the router is cutting. So I slow
> the acceleration down. This has no effect. I'm using two 1200oz/in
> steppers slaved one either side, then they are belted down. I am only
> cutting at about 500mm/min at a depth of 5mm. So there is plently of
> torque from the stepper motors etc etc. So I think that I can rule
> out the missing stepps theory.
>
> But could it possibly be electrical interference causing my troubles?
>
> Have a look at this picture, this is my power supplie, breakout board,
> stepper drivers etc. It is very messy... etc etc.
>
> http://i90.photobucket.com/albums/k259/peterapalais/electronics1.jpg
>
> http://i90.photobucket.com/albums/k259/peterapalais/electronics2.jpg
>
> What are the chances that this birds nest is causing me trouble? ie
> 20% 40% 60% chance of electrical noise etc.
>
> If so, what is the solution. Make up a new board and space everything
> right out so that things are further apart? Should I use sheilded
> wire etc? If so for what connections.
>
> Cheers
> Peter
> Australia
>

Dry run the program (Router off cutting air) and check. If it's noise
from the stepping/power supply then it should show up there. Your
motor output wiring is wrapped around your low level input (DB25) so
is a place you would need to address if you are getting lost/gained steps.

If the noise source is the router motor itself, then typically it's
conducted noise (something across a shared ground connection) The
radiated noise should be easier to block. A wooden case does nothing
to block EMI and RFI in either direction. Simply enclosing the
electronics in a metal structure can block most of the radiated noise,
but could actually amplify the crosstalk of all the internal components.

Noise suppresion at the source is good but it's better to address both
parts of the equation (source and receivng end).

Chasing and finding noise is a like chasing a phantom. Things that
work for safety grounding and 50/60 hz work poorly or not at all for
high frequency.

The only "real" ground is a solid gold rod driven to the center of the
earth!