[chainds]

I have been running an X2 mill using mach 3 and a G540 drive. I started to notice positioning error being introduced while while working on some engraving. It seems the slower I run the feed speed, the more error is introduced. I created a test using the g-code from the engraving. I would start the run with a dial indicator zeroed on the X axis, the code would run and should return to the same position and hopefully read zero on the indicator. At a very slow 0.5 ipm the error was several hundredths. At 1.0 ipm the error was reduced to several thousandths. At 10 ipm the error was sub thousandths if at all.

There is a lot of discussion of errors caused by running motors too fast, but I have the opposite problem. I seems so counter intuitive. I have not been able to introduce the error using MPG input. Does anyone have experience with a similar problem? Thanks.

[Toinvd]

first check your mechanics. is there a bad connection between motor and ballscrew, or a bearing not well tightenned.
Check if afther running your test several times the error increases or stays at the same. If the error is the same a mechanical problem is most expected.

[chainds]

First, thank you for the reply.

I have been running the test looking specifically at the x axis. I wanted to make sure something wasn't binding, so I completely went through the axis - cleaning, greasing, adjusting, etc. I'm convinced the mechanicals are good.

The correlation between the error and a slower feed speed is really strong and repeatable. The slower the feed, the greater the error. I am going to re-tune the motors and see what happens. I feel like it is related to the motor drive some way, but I am grasping at straws.

I am curious why you are thinking mechanical?


Sent from my iPhone using Tapatalk

[awerby]

So if it works when you go fast, and not when you go slow, why go slow?

[chainds]

By the way Tionvd, I used to visit ASMLithography in Eindhoven about 15 years ago on business. I enjoyed my time there.


Sent from my iPhone using Tapatalk

[chainds]

Awerby, I can't spin the spindle fast enough to make the engraving look good for the feed rate required to minimize the error.


Sent from my iPhone using Tapatalk

[James Newton]

I would strongly recommend contacting Gecko drive customer support.

Swap drivers between X and Y axis.

Swap signals on the break out or whatever is doing the motion control. (providing the step/direction signals)

Swap motors.

If you can't or they can't help you, beg, borrow, steal, or :shudder: buy an o-scope and look at the signals going into and out of the driver. E.g. the step and direction and the motor phases out. See if anything looks wrong... not getting a full swing from logic 0 to 1? Noise? Drive cutting out? It can be really hard to see an intermittent issue (which it sounds like this is)

[chainds]

Both of those things are good advice and I was planning both. I happen to have on o-scope, just need some time in the shop. The lack of responses are confirming my fear that this is a more difficult problem and something not many have experienced. Thanks for taking time to respond. I'll update as I gather information.


Sent from my iPhone using Tapatalk

[awerby]

For engraving, the 2500 RPM top speed of an X2 mill is just not sufficient. If you really want to do this, I'd suggest getting an auxilliary high-speed spindle and mounting it to the head of your mill. If you're not in heavy production, something like the Proxxon IB/E would probably work fine, and give you spindle speeds up to 20k rpm. And it wouldn't take all day to engrave one small piece.

[chainds]

Awerby, I understand that the spindle rpm capability on the X2 mill is lacking for serious engraving. Fortunately I don't need to engrave very often. I will entertain an high rpm spindle if I ever have to do more than I am now. Thanks for taking time to respond.


Sent from my iPhone using Tapatalk

[chainds]

It seems I solved my problem just by going back to basics. I re-tuned all the motors and low and behold, the test I have been running started to be repeatable within about a thousandth (within my ability to measure it). I feel like it was a "is there gas in the tank" or "is it plugged in" exercise . I want to thank everyone who took time to respond to my call for help.