Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: Ozo126 on November 04, 2013, 02:30:44 PM

Title: Mach3 problem- losing position
Post by: Ozo126 on November 04, 2013, 02:30:44 PM
Hello Guys,

Have an issue with Mach (I think) and I’m hoping someone can help.

I have a connect minor mini mill converted to run on mach, I’ve been running it without issue for approx. 3 yrs.

Last week I was running a short program 6 times in an afternoon, the next morning I re probed my setup to run again.

Everything looked good at the start of the cycle, so I hit the button and walked away, big mistake….
The mill crashed a 10mm spot drill into my stock/table and proceeded to rip an easy change tool holder out of the spindle.

I E-Stopped the machine, had a quick check and seen that my DRO position was reading incorrect to where it should have been, I blamed this on the crash locking up the axis motors.

I restarted the PC, checked the spindle tram and re probed my part- this is where it gets confusing……

I then mdi G00 X50.  & then re-zeroed my X. This being my actual X0 for the program. I proceeded to single block the program, it was soon obvious that my x,y,z axis zeros were all incorrect.

Then I closed mach, re probed and then re loaded the program,  I again started to single block the program, mach then proceeded to freeze whilst still moving the axis for a few seconds.

I did this 3 times with the same results.  After the first time I removed a flash stick as I thought this might be something to do with.

I decided to do some basics tests to work out the problem, firstly I re probed my setup set my X0 Y0 where they should be, using the MDI I told the machine to move the X axis 2.5mm it moved it approx 1mm.
Very annoying though I could take a 50mm slip and using a  dial indicator measure it as exactly 50mm on the DRO on the X axis.

At this point I left it and decided to think about it over the weekend as I was getting nowhere.

This morning I started by testing some more with the dial indicator I noticed that my backlash had went from 0.06 on X and 0 zero backlash comp allowed for on Y, to the point I needed 1.5mm backlash on X and 0.5 on Y.  I adjusted the axis moved the correct distance (problem with the 2.5mm only moving 1mm solved) but it doesn’t make sense how backlash can jump like that so quickly or explain the initial crash.

I decided to run the original program again, having checked thru it to make sure there was nothing wrong really standing out in the code. I probed my setup, then measured my tool offsets, and started the program it was instantly obvious something was wrong- Mach had lost the zero setting for x and y and was cutting approx. 50mm  away from where it should have been (probably best to confirm here that I had zeroed and programmed the code in the same work offset i.e. G54, it’s also worth noting that the second part of the code called for G55 which had already been set, the difference in the 2 planes was 2mm on the Y axis, I regularly run multiple work offsets in the same code without issue).

I then re-probed the setup, reset my tool heights and decided to run the second part of the code which was created in Cambam,  I deleted the first part out(hand written by me) and proceeded to single block the program, again I lost the zero for the part almost instantly on X,Y,Z axis, the X and Y being approx. 50mm out and the Z 40 (too high thankfully!)  the x and y figure derived from using the probe to find the original X & Y  zero and reading the difference in the DRO. Its also probably worth noting that I used the MDI to ensure I was in the correct work offset- G55 before the program was started.

I’m running a full version of Mach ver- R3.043.022. My problem seems something similar to this:


All the different symptoms point to different culprits for the issue or a combination.
I think its safe to say its not a losing a few steps here and there issue.
I need to work out why it crashed initially. Is the backlash really mechanical? Is the problem Mach/PC/Drivers or a combination?
I’ve subbed the rest of the work out I had planned to be running on the machine but I would really like to get it back running reliably.

Any thoughts would be greatly appreciated.

Many Thanks,
Title: Re: Mach3 problem- losing position
Post by: Hood on November 06, 2013, 03:09:51 AM
A lot to read and digest here and afraid I dont have time at the moment to do so but if you attach your xml I will have a look at that and hopefully I will get a chance later on today to read more thoroughly.
Title: Re: Mach3 problem- losing position
Post by: Ozo126 on November 13, 2013, 08:40:28 AM
Hello Hood,

Thanks for the reply & apologies I'm only getting back to you now.

Yes there is a lot going on!

Please find attached my xml file.

Title: Re: Mach3 problem- losing position
Post by: Hood on November 13, 2013, 04:33:04 PM
Your setup seems basic so dont see a problem with your xml. I see you are using unusual pins for your axis, what kind of drives/breakout are you using?
Title: Re: Mach3 problem- losing position
Post by: Ozo126 on December 04, 2013, 05:11:01 PM
Hi Hood,

Thanks again for the reply.
Opto board and drivers bought from so probably cheap chinese but they were handy at the time being in the UK. Yes the pin numbers are quite odd.

I think I may have found my problem, the PC started to randomly give me a blank screen, the only way I could get rid of it was by switching the moniter on and off, it then proceeded to bring up a nice blue windows error message followed by a physical memory dump, so Im guessing the issue is with the PC. Its a dell optiplex, has nothing on it except Mach3 and the driver test shows excellent performance so not too sure whats gone wrong.

Seriously considering a laptop and a smooth stepper at this point, but I still cant work out the sudden leap in backslash...

Title: Re: Mach3 problem- losing position
Post by: geh7552 on December 04, 2013, 05:52:46 PM
Sounds like a RAM failure problem. Try Google  Dell Opitiplex memory dump error messages problem and see what shows up.