[ELane42]

I don’t know exactly if this is where it belongs in the forum, but I’m new to CNCing so bear with me here.

I got a CNC machine (specifically the greenBull) from Build Your CNC, and assembled it all myself. After working out some kinks, the I’ve ran into a series of problems that are seemingly not consistent and therefor I have no idea where to even start looking for a solution. I’ll attach some pictures of the results I’m getting, it seems randomly translated in either y or x. There’s no slipping of the couplings on the axis (I checked). The zero is the same. And the problem almost always occurs when there’s multiple processes involved in the same program.

Any help is appreciated!

[joeavaerage]

Hi,
there are two possibilities, excluding mechanical slippage.
1) Electrical noise inducing extra steps
2) Missing steps.

The later is more likely. If a stepper is marginally overloaded, and that is by far and away most prevalent at speed its common for a stepper to either miss or gain an extra step.
Over a short period of time you get very distorted results such as you have posted. It gets worse as time goes on.

Craig

[ELane42]

Hi,
there are two possibilities, excluding mechanical slippage.
1) Electrical noise inducing extra steps
2) Missing steps.

The later is more likely. If a stepper is marginally overloaded, and that is by far and away most prevalent at speed its common for a stepper to either miss or gain an extra step.
Over a short period of time you get very distorted results such as you have posted. It gets worse as time goes on.

Craig

So what would be the fix to that? Slower feed rate?

[joeavaerage]

Hi,
yes a slower feed rate is a possible, in fact sometimes may be the only solution.

What you need to do is confirm that is what is happening. There is no point in applying what you think maybe a solution until you have conclusively proved a fault.

If you run some code , say one panel, and we'll assume the fault has occurred the at the end hit <Return to Work Zero>. Does the machine return to the start point exactly?
If it does not then we know that somehow the steppers are gaining/losing steps. If that is the case, and seems highly likely then we need to devise an experiment to determine whether
its electrical noise causing random gain/loss of steps or whether its the steppers themselves. This is not that easy.

In the first instance try running the same code but with the tool running above and clear of the material. If the fault goes away we at least know its when the machine is under load, due to
either electrical noise from the spindle or the extra cutting forces. Another experiment is to run the code but at a much reduced, say 1/4 feed rate. Does the fault recur? If not that is an indication that its the cutting
forces that occur at speed that seem to cause the problem.

What steppers and what voltage drivers/supply do you have?

Craig

[CNCMAN172]

Hi,
yes a slower feed rate is a possible, in fact sometimes may be the only solution.

What you need to do is confirm that is what is happening. There is no point in applying what you think maybe a solution until you have conclusively proved a fault.

If you run some code , say one panel, and we'll assume the fault has occurred the at the end hit <Return to Work Zero>. Does the machine return to the start point exactly?
If it does not then we know that somehow the steppers are gaining/losing steps. If that is the case, and seems highly likely then we need to devise an experiment to determine whether
its electrical noise causing random gain/loss of steps or whether its the steppers themselves. This is not that easy.

In the first instance try running the same code but with the tool running above and clear of the material. If the fault goes away we at least know its when the machine is under load, due to
either electrical noise from the spindle or the extra cutting forces. Another experiment is to run the code but at a much reduced, say 1/4 feed rate. Does the fault recur? If not that is an indication that its the cutting
forces that occur at speed that seem to cause the problem.

What steppers and what voltage drivers/supply do you have?

Craig

Based on the photo it appears you might have backlash in your system. Search backlash and you will find how to check

[ELane42]

Hi,
yes a slower feed rate is a possible, in fact sometimes may be the only solution.

What you need to do is confirm that is what is happening. There is no point in applying what you think maybe a solution until you have conclusively proved a fault.

If you run some code , say one panel, and we'll assume the fault has occurred the at the end hit <Return to Work Zero>. Does the machine return to the start point exactly?
If it does not then we know that somehow the steppers are gaining/losing steps. If that is the case, and seems highly likely then we need to devise an experiment to determine whether
its electrical noise causing random gain/loss of steps or whether its the steppers themselves. This is not that easy.

In the first instance try running the same code but with the tool running above and clear of the material. If the fault goes away we at least know its when the machine is under load, due to
either electrical noise from the spindle or the extra cutting forces. Another experiment is to run the code but at a much reduced, say 1/4 feed rate. Does the fault recur? If not that is an indication that its the cutting
forces that occur at speed that seem to cause the problem.

What steppers and what voltage drivers/supply do you have?

Craig

I have Nema 34 motors and DM556 drivers. The voltage from the power supply is 220V and the voltage from the driver is 20-50 each (for three) with a max current output of 5.6 A. Right now I have it set to the max setting for current output and 3200 Pulse/rev for the drivers. The motor tuning in Mach 3 is something around the 1000 mark for X and Y (Z is different). I’ll start running some of those test today and get back with the results.

[ELane42]

Based on the photo it appears you might have backlash in your system. Search backlash and you will find how to check

The only axis on my machine that’s controlled by a screw is the Z-axis. Based on a quick bit of research that I did, it was describing screws and the “play” between the threads. I don’t know if this can happen on chain drive setups.

[awerby]

What voltage is actually being supplied to the drivers? 20-50v is a big range. It's rare that a big NEMA34 stepper can run well on 20v. If the drivers can handle it, go to 50v and see if your results are better.

I just noticed that you mentioned chain drive. There's a reason you don't see many CNC machines driven that way. It's fine for bicycles, but too sloppy for precision cutting.

[CNCMAN172]

The only axis on my machine that’s controlled by a screw is the Z-axis. Based on a quick bit of research that I did, it was describing screws and the “play” between the threads. I don’t know if this can happen on chain drive setups.

Not true, you can have it with ballscrew drives, gear drives, belt drives, cable drives. This just means some slop exits each time the axis changes direction.

[ELane42]

What voltage is actually being supplied to the drivers? 20-50v is a big range. It's rare that a big NEMA34 stepper can run well on 20v. If the drivers can handle it, go to 50v and see if your results are better.

I just noticed that you mentioned chain drive. There's a reason you don't see many CNC machines driven that way. It's fine for bicycles, but too sloppy for precision cutting.

I can set the max current draw supplied to the steppers from the drivers, right now that’s at max. The chain has been fine so far, will say the couples they gave me did not size up to the actual rods which connect to the chain driver sprocket. I’ve been marking for slippage but I haven’t noticed anything yet, still have to do some more tests when I get time.

Today I did some dry runs to see if the bit just couldn’t handle the feed rate, and I noticed a shift in the Y axis on both runs. One was with the spindle off (very small shift) and the other with the spindle on (about a 4 inch shift in one direction). I think I might have to do some mechanical adjustments, but there is a clear pattern between the y axis shifts and the spindle on. Again, more testing needed when I find time, but for now that’s what I’ve got.

[awerby]

There's a difference between current (measured in amperes) and voltage. If your voltage is variable between 20 and 50v, there's probably a dip switch that sets it.

You didn't mention which type of spindle you're using, but 3-phase spindles in particular are known to emanate RFI from their power leads. They need to be shielded and the shielding needs to be grounded to prevent it from interfering with the logic-level signals that control your machine.

[ELane42]

There's a difference between current (measured in amperes) and voltage. If your voltage is variable between 20 and 50v, there's probably a dip switch that sets it.

You didn't mention which type of spindle you're using, but 3-phase spindles in particular are known to emanate RFI from their power leads. They need to be shielded and the shielding needs to be grounded to prevent it from interfering with the logic-level signals that control your machine.

https://www.damencnc.com/userdata/file/1750-1_DM556_datasheet.pdf

This is the data sheet for the drivers, as far as I can’t tell the dip switches set the max current draw and the pulse/rev, the switch four also changes the current from static to dynamic.

The spindle I’m using is a 2.2 kw 220 V Huanyang water cooled spindle.

[joeavaerage]

Hi,
I've already made some suggestions that might help you diagnose the problem....have you tries any of them?

Craig

[ELane42]

Hi,
I've already made some suggestions that might help you diagnose the problem....have you tries any of them?

Craig

Yes I did. I did a dry run without the spindle on, it returned to the zero with a very slight change in Y-axis. did a dry run with the spindle on and the Y-axis had shifted about 4 inches. Today I should be able to run some more tests.

[ELane42]

Yes I did. I did a dry run without the spindle on, it returned to the zero with a very slight change in Y-axis. did a dry run with the spindle on and the Y-axis had shifted about 4 inches. Today I should be able to run some more tests.

And by 4 inches I actually meant about 1/2 an inch, my mistake.

[joeavaerage]

Hi,
that rather suggests that electrical noise from the spindle is causing one axis to gain/lose steps. This is a common problem.

Do you have an AC Line filter on the 230VAC input side of the VFD? Do you have an AC Line filter on the230VAC side of your PC/motion control electronic power supplies?

Craig

[ELane42]

Thanks for y’all’s help. I got it working accurately now.