[tdnp]

My son and I where running our home build router and it did not turn out. I do not know if the machine loast where it was at or what happened. Hopefully the pic will explain. The second letter was an "A" you can see where it started to make the bottom and then it moved up and started the rest.

It is a trim router and it was under a load.

Any advice is appriciated.

[ger21]

Yes, your steppers are stalling or losing steps.

[louieatienza]

I would check to see if you are getting any binding or tightness in your leadnuts and linear bearings... Pic of the machine and specs would help.

[Fastest1]

If running Mach3, try lowering your acceleration rates. Go thru all axis with a dial indicator. Move away and back to the original position a few times. Make sure the indicator returns to the exact position. There is more to it than that, but that is a start. Probably best to run thru the calibration and backlash tests and adjust and or lube all moving parts.

[Fastest1]

Btw do all of the tests at your rapid speeds and above. It must repeat. Also how about a picture of your set up just in case there are known issues and or solutions?

[Plan_B]

Hey.. I think a picture of the mechanicals on the machine could go a long way. what's happened is that, for some reason your X axis (or whatever you call your axis that runs from top to bottom of those letters) has lost steps. I've had that happen to me a few times for a couple different reasons..

1. A small piece of wood (and one time a wood screw) jammed between one of my carriages and the table, causing the gantry to jam, thereby losing steps.

2. I once had a bearing failure that caused my Y axis to jam.

When this happens you see exactly what you have there... a 'jump' in the relatuive postion of the tool path, by the amount that the axis should have traveled had it not jammed..

I doubt that kind of big jump (in isolation) is caused by missing steps due to too high acceleration rates, but I suppose it's possible. you should be able to determine that by a test using G code to rapid back and forth, and after a few rapids, measure it's position. If it's skipping, lower the max accel for that axis and try again. I'll bet you anything its a mechanical problem like a 'bug' (or in this case 'a screw') in your system )

What size steppers are you using, and what are the drive and linear carriage arrangements on that axis?

[tdnp]

Thanks for all the information. I run 200 oz motors and mach3. i was messing with the setting some time ago to try to get it to cut faster and probably messed something up. Does any one have a link on the proper settings in mach 3.

Thanks in advance.

[Fastest1]

Mach3 has to be set up for your machine. Which part isn't right? Does it move the specified amount when commanded? If not use the axis calibration tool to correct that. If it is losing position, is it a mechanical binding or is losing steps due to accel values being too high? There is no easy way. To get the best results here like any other profession takes preparation. Try rushing it and you end up with scrap.

[BAMCNC.COM]

Hmmm, axis calibration was talked about here quite abit... Everyone should read this thread, especially those who believe in axis calibration!

http://www.cnczone.com/forums/showthread.php?t=165154

[Fastest1]

Yes, I read that thread a few times over. I don't know who is right or if both sides can't be right. I do know that I have used the utility and though I can mathematically calculate my steps per unit ( inch in my case) actual movement traveled is never an even number. My machine appears to move the prescribed amount to the best of my ability to measure it. If I can't repeat, I go back and figure out why, usually a loose assembly of some form. For us mere mortals that if it is off .001 it is going to fit guys. It works great. In fact I wonder how I would work around the fact my machine wouldn't go 1" with the mathematical equation?

[BAMCNC.COM]

Yep its tricky, like I said over there, science and reality are sometimes two different things.

[109jb]

The picture from post #1 of this thread is not showing a calibration problem. Whether you agree or disagree about calibration doesn't natter in this case because the picture shows that the machine has lost position, AKA lost steps. There are various reasons for lost steps, but calibration isn't one of them.

[BAMCNC.COM]

Fastest's reply seems like the perfect reply in my book...

Mach3 has to be set up for your machine. Which part isn't right? Does it move the specified amount when commanded? If not use the axis calibration tool to correct that. If it is losing position, is it a mechanical binding or is losing steps due to accel values being too high? There is no easy way. To get the best results here like any other profession takes preparation. Try rushing it and you end up with scrap.

[PaulRowntree]

This thread seems to be revisiting some historical discussions which really isn't helping tdnp out. The issue is clearly not caused by mis-calibrated drives. The acceleration settings and machine tightness are the usual suspects, followed by bad cables/connectors.

[Fastest1]

109 you might be right however my first step would be to verify actual distance commanded against distance traveled. If all was ok there, then tests of acceleration and velocity til I was capable of returning to the original position. I used a few of Bam's recommendations in set up recently and came away quite impressed with the overall improvement of my mill. That being said, His recommendations weren't new or his only, The improvement has made me rethink my adjustment techniques and move away from the tight gib theory I formerly prescribed to. In fact when I get over my lazy tendencies, I am going to use the same techniques on my manual lathe and mill and bet the action improves as do the finishes.

[louieatienza]

Thanks for all the information. I run 200 oz motors and mach3. i was messing with the setting some time ago to try to get it to cut faster and probably messed something up. Does any one have a link on the proper settings in mach 3.

Thanks in advance.

Depending on your stepper drive, power supply, and type of leadscrew, you coulf be stalling just by trying to go too fast for your setup. Or the screw is spinning so fast that it starts towhip, which can also cause the steppers to stall. You can run the same job with a slower rapids speed and see if the job will complete correctly, then bump the speed incrementally until you get the stall.

It would be easier if you gave more specs: drive type, psu voltage... a pic of your machine can shed new light...

[109jb]

109 you might be right however my first step would be to verify actual distance commanded against distance traveled. If all was ok there, then tests of acceleration and velocity til I was capable of returning to the original position. I used a few of Bam's recommendations in set up recently and came away quite impressed with the overall improvement of my mill. That being said, His recommendations weren't new or his only, The improvement has made me rethink my adjustment techniques and move away from the tight gib theory I formerly prescribed to. In fact when I get over my lazy tendencies, I am going to use the same techniques on my manual lathe and mill and bet the action improves as do the finishes.

All I can say is look at the picture in post #1. There is no way that the error shown is a calibration error. A calibration error would only cause the letters to be made the wrong size, it would not account for the gross error shown in the making of the letter "A" in the first post here.

[109jb]

This thread seems to be revisiting some historical discussions which really isn't helping tdnp out. The issue is clearly not caused by mis-calibrated drives. The acceleration settings and machine tightness are the usual suspects, followed by bad cables/connectors.

I totally agree. The problem shown in the picture in the first post clearly shows a loss of position. As you pointed out, this would be from machine binding, or mal-adjusted drive settings causing lost steps.

This is why I made my first post in this thread because the mention of calibration doesn't make sense for the problem posed in this thread. Calibration is useless until you have a system that doesn't lose steps in the first place.

[jneutron]

From the looks of the pattern, you first cut the bottom horiz of the A, lift Z, go to the next location, drop z and continue.

Perhaps during your Z rise you are also stepping the other axis? Sounds more like a code problem to me.

jn