Author Topic: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d  (Read 1798 times)

0 Members and 1 Guest are viewing this topic.

Offline Mauri

  • Active Member
  • Posts: 292
    • View Profile
DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« on: October 20, 2015, 09:40:10 PM »
Hi,
In the past we have machined a lot of engravings in Hard Brass with 20D .1mm Cutter with no issues.
Recently we have been having issues with breaking points on 20D .1mm tip tools on engraving.
We did not have this problem before and machined at faster F rates.
We checkout the milling machine and could not find an issue (The Head Z Axis is always within .0001").
What we just have found is that the G-Code Z instruction is go -.1mm Deep, and when you look at the DRO it first goes down to -.100mm at F20 then as it moves on the X or Y at F200 it drops further by -.0313 to a total of -.1313 at the faster F Speed causing the tip to break.
What is causing the Mach3 to not do as the G-Code has instructed it to?
Is it the ESS dropping it further by not properly compensating the Backlash Value?
The backlash for Z is .0330mm
It is mentioned that the Latest ESS has issues with Backlash?
Any ideas to what is happening?
I cannot use Mach4 yet at the CV setting are not implemented yet.
Regards,
Mauri

Online ger21

  • Global Moderator
  • *
  • Posts: 6,156
    • View Profile
    • The CNC Woodworker
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #1 on: October 20, 2015, 09:53:28 PM »
The DRO actually shoes it dropping? Can you post some sample g-code that does this?

You might want to try 3.043.062, as .066 is known to be very buggy.
Gerry

2010 Screenset
http://www.thecncwoodworker.com/2010.html

JointCAM Dovetail and Box Joint software
http://www.g-forcecnc.com/jointcam.html

Offline geh7552

  • Active Member
  • Posts: 240
    • View Profile
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #2 on: October 20, 2015, 11:18:27 PM »
Don't use mach V.066 it has known issues. I was running it for awhile with a ESS and would get intermittent Z axis problems... mostly after I would do a tool auto zeroing with a touch plate. It would try to drive the cutter into the work. Dropped back to V.062 and haven't had the Z axis issues again. V.066 should be removed.

I wouldn't be in any rush to Mach 4... it has it's own set of problems and far from a finished product with no light at the end of the tunnel.  

Offline Mauri

  • Active Member
  • Posts: 292
    • View Profile
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #3 on: October 20, 2015, 11:27:59 PM »
Ger21/gen7552,
I tested Mach 3.42.066 with ESS_v10h2d1a and ESS_v10r2d1d same issue with both ESS's.
I tested Mach 3.42.062 with ESS_v10h2d1a and ESS_v10r2d1d same issue with both ESS's.
I tested Mach 3.42.066 with no Driver No Issues all work fine, so the problem is with the ESS Drivers.
I then remove the Backlash from the Z Axis in Mach3 and all works fine, all engraving cuts are at .1mm deep.
Regards,
Mauri.

Offline geh7552

  • Active Member
  • Posts: 240
    • View Profile
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #4 on: October 21, 2015, 06:26:43 PM »
If backlash is not setup properly in Mach 3 it can cause strange actions. I played around with it to see how it reacts and ended up with more questions than answers. I'm not even sure backlash is working properly in Mach 3. It's always better to fix mechanical problems rather than doing it in software.   

Offline Mauri

  • Active Member
  • Posts: 292
    • View Profile
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #5 on: October 21, 2015, 06:54:02 PM »
geh7552,
Appreciate your answer, I was a Metrologist for a number of years and know how to measure accurately, however unless you are using Linear motors and Rails all mechanical systems will have backlash no matter how much money you spend, even in temperature controlled rooms. Metal will expand stretch and contract in mechanical motion and heat/cold. All guides with gibes will have issues. The balance (back and forth motion) and mass of the Z Drive head as it move up and down will have issues with minute angular tilts. The accuracy of the Ballscrew/Nuts and Trust Bearings and tightness all have backlash issues being some.
We have just completed two engravings without putting any backlash in Mach3 with ESS and both produced good results o.1mm deep.
Our precision Dial indicator shows only .00011" backlash at the engraving level when measuring the tip movement.
We had no issues with Parallel Driver with Mach3 also early ESS drivers did not produce these recent problems.
Regards,
Mauri.

Offline geh7552

  • Active Member
  • Posts: 240
    • View Profile
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #6 on: October 22, 2015, 05:37:34 PM »
I would be interested in seeing your backlash numbers.

Offline robertspark

  • Active Member
  • Posts: 585
    • View Profile
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #7 on: October 22, 2015, 07:51:02 PM »
There are numbers posted on the warp9 forum and above.

0.0330mm backlash was noted, with a z steps/mm of 640

Rob
Rob

Albert Einstein ― “If you can't explain it to a six year old, you don't understand it yourself.”

Offline Mauri

  • Active Member
  • Posts: 292
    • View Profile
Re: DRO vs GCode on Z with Mach 3.46.066 and ESS_v10r2d1d
« Reply #8 on: October 22, 2015, 08:16:07 PM »
Rob,
Posted reply to your post in Warp9 forum.
Problem is with both USS and ESS drivers.
Regards,
Mauri.