Step Loss (again)

I've been trying a few things before replacing the drivers and/or stepper motors. I ran a 2D program, that cut a name plate. The
program was 5200 lines long, and cut 1mm deep in two pecks. I'm looking at the difference in position between the upper and lower level. The lower level is out of position to the top one. I ran the program twice, once with the feed rate set at 1000mmpm and acceleration at 50 and once with feed rate set at 400 and acceleration at 10 (on motor tuning). If this was a problem with losing steps due to over accelerating or going too fast I'd have expected a noticable difference in the test pieces, but there isn't. The movement between the two levels is +0.35 in Y and +0.2 in X. Actual size of the sign (80mm x 40mm) is within 0.1mm -which is reasonable measured with a digital vernier and allowing for possible cuttter wear. Any suggestions where to go next gratefully received, as I'm not too sure right now.
Regards Kevin
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Help is out there Kevin but you're a little too pedantic for me.
What machine?
What software?
What work?
Set your work near the centre of your machines envelope. Try cutting a simple circle, say 50mm in diameter. Do 2 passes, 1 at 0.2mm & 1 at 0.5mm. Returning to zero after the first pass. Do this at a slow speed like 60mmpm. WRITE THE CODE YOURSELF, that way you can rule in/out at least 1 thing.
The only times I've lost steps is (1) My computer running Mach3 couldn't cope & (2) I asked the machine to go to the very limits of its envelope.
In trying to decipher your problem I'm guessing that you're cutting something in 2 passes but the 2 aren't matching exactly?
Have you browsed whatever code you're feeding the machine? The X & Y co-ordinates should be exactly the same, only the Z is differant.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
BooBoo wrote:

Seconded ;) We need to eliminate something from the list of possible problems so we can help.
--
Lester Caine - G8HFL
-----------------------------
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Sorry, just to recap from my earlier post. Machine is a Denford Triac, which has had the original stepper drivers replaced with the 3A ones from Arc Euro and a new PSU but retains the original stepper motors. Control is now via a PC running mach3, through a CNC4PC C11 board. I'm happy the program is correct, and yes the problem is that the two levels of machining are out of position to each other. I'll try the circle trial suggested, but on small programs I haven't been able to find dicrepencies before. I only started to notice the error when I started sending larger programs to the machine.
Regards Kevin
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

I've just run some more test programs, using the wizards in mach3 to generate the code (but I'm positive it's not a code issue as it has done it with code from different sources). I cut two square pockets with a round upstand in the middle. Measuring round the circles they are, as close as I can measure, round and the correct size. Measuring the step between the circles (which is why I left them standing up - easier to measure) in Y the gap is 4.95 on both sides, but in X the gap is 5.00 one side and 4.90 the other. So it seems that I've lost position again. This is the same as all the other tests, a very small step loss. This program had rapid moves in it, between the passes clearing the material away. My next move is to edit the program and make all moves at feed rate.
Regards Kevin
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here. All logos and trade names are the property of their respective owners.