Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: edbanks on October 13, 2008, 06:08:35 PM

Title: Gross loss of position
Post by: edbanks on October 13, 2008, 06:08:35 PM
My setup: A older Torchmate 1 machine, converted to use Keling 495 oz/in (X&Y) and 156 oz/in (Z) motors all hooked to Keling 6050 drivers. I also have a THC-300 height controller and a Powermax 1000. Using Sheetcam for G-code generation and Mach3 to run it all.

The problem: Gross loss of position after a M05 command only when actually cutting. It seems like the rapid to the next torch start never actually moves when commanded to do so. This does not occur after every M05 (see pictures below). With a marker taped to the torch, it "cuts" fine every time. The amount deviated can be quite a bit, I measured ~.5" +Y and ~3" -X direction on one part. I thought interference might be causing the problem, so I separated the torch and motor cables, same thing. This problem occurs when using the THC or when not using it.

If anyone could shed some light into this issue, I would really appreciate it. I've been pulling my hair out for the past two days trying to resolve this issue.

(http://www.iamapathetic.com/catmiscut.jpg)
(http://www.iamapathetic.com/cat01.jpg)
Title: Re: Gross loss of position
Post by: Chaoticone on October 13, 2008, 06:50:00 PM
Quote
With a marker taped to the torch, it "cuts" fine every time. The amount deviated can be quite a bit, I measured ~.5" +Y and ~3" -X direction on one part. I thought interference might be causing the problem, so I separated the torch and motor cables, same thing. This problem occurs when using the THC or when not using it.

So it follows the path fine while using the THC as long as there isn't any M5s in your code?

Brett
Title: Re: Gross loss of position
Post by: edbanks on October 13, 2008, 07:03:29 PM
So it follows the path fine while using the THC as long as there isn't any M5s in your code?

I did not edit a program so that it never turns off the torch, I will add that to the list of things to try. I did edit out the G28's and run with the THC off (in mach), also set the post to ref. home at every torch start, both with no effect. I also tried editing the G00 rapid to G01 with F100 after the M05. I've adjusted velocity/acceleration down and increased step/dir pulses, which did nothing as well.
Title: Re: Gross loss of position
Post by: edbanks on October 18, 2008, 01:53:58 PM
Allright, I was out of town this past week for work, so it took me a few days to get back at it. I just tried running a program with torch on/off like normal and another with only one torch on/off at start/finish. With the torch firing only on start/finish it cut without issue, with it firing multiple times in the program it moved the same as before. Any ideas?
Title: Re: Gross loss of position
Post by: jimpinder on October 19, 2008, 02:03:08 AM
What fires up the torch.

1. Has this ever worked correctly - i.e. are we looking for a fault that has developed, or one that is occuring from the start.
2. You say that Mach will describe the pattern correctly every time, unless the torch is firing up during the run.

It would seem to me that there are some spikes, or something in the supply lines that is badly affecting one (or both) of your drives to the X and Y axis, as the torch starts up - you have probably thought of that.

Are the power supplies for the torch and the motors completely seperate, and are they all correctly earthed.

Check the return path for plasma cutter starter circuit, particularly if it is a high voltage starter.

I would also check to see that theĀ  ground connection from the computer or BoB IS NOT connected to the ground connection on the high voltage side of the driver cards.
Title: Re: Gross loss of position
Post by: edbanks on October 22, 2008, 09:59:44 PM
The THC-300 board fires the torch through a relay. Yes, this had worked prior to adding the THC, I was running the torchmate software when I first got the machine a few months ago though. When I added the z-axis and swapped out the motors/controls I started using mach and had it running before messing with the THC although I did not run a lot of programs through, maybe 10-15. I was thinking it working before adding the THC-300 and completely disconnected the unit, wired the torch on directly to my breakout board and remapped the torch on in mach to correspond. I then cut a test program of four 4" circles, about a 8 or 9 times without issue. So, i'm pretty sure there's some issue with my THC-300 unit, i've dispatched a post on the yahoo forum for the THC-300. Thanks.