[Varske]

Hi,
I am tired of fighting my cnc router(working, but still in progress).
It is around 0.9x1.5m working area with chinese hiwin rails and 2020 ballscrews and 8.5nm closed loop steppers. Hbs86h drivers, 60v ac.

Problem is with longest dual motor axis. I can only move it up to less than 3m/min while being sure that drivers won't stop and throw an error. Components are more or less aligned, nothing binds and can easily be moved with hand.
While moving gantry at some points motors unpleasantly vibrate and sometimes starts shaking back and forth for 5 to 15 seconds, then shut off and stay with red light flashing. Other axis are moving nice and smooth with a satisfying speed.
At low speeds motor still tend to vibrate intermittently, however keeps operating.

It looks like motors are loosing or gaing some microsteps because of each others movement and trying to compensate them causing unpredictable failutes this way.

Did anyone had this problem before? If so, how did you solve it?
Is it possoble to use closed loop stepper as regalar stepper? To ignore the encoder's signals.

[peteeng]

Hi Varske - You say the long axis has two Motors?

1) are they driven by separate drivers?
2) disconnect one motor and see what happens, then reconnect and disconnect the other motor. This will give you a clue to what is happening

cheers Peter

[Varske]

Hi Varske - You say the long axis has two Motors?

1) are they driven by separate drivers?
2) disconnect one motor and see what happens, then reconnect and disconnect the other motor. This will give you a clue to what is happening

cheers Peter

Yes, sepparate drivers. Tried to run them freely and with ballscrews disconnected from gantry. No problems.
I have noticed that gantry is flexing a bit. For instance if left sida is stationary and i am turning ballscrew coupling with hand, one side can be moved couple of milimeters, therefore I suspect problem is with 'smart' encoders being too smart and not leaving any margin of error.

I was thinking to use one encoder's signal for both drivers so they would be synced(?)

Didn't try to run one side only as this would be quite stressful for the machine

[peteeng]

To run one side only you need to disconnect the drive nut so the screw does nor impeded the motion. This will show up if one driver is behaving badly or not. 2mm differential is not unusual for gantry "walk" . Peter

[joeavaerage]

Hi,
you say that you can only travel at 3m/min? Are the ballscrews direct coupled, I would suspect they are?

If that's the case 3m/min corresponds to a stepper rotation of 150rpm....does that sound correct? And again to confirm
that they perform adequately at 150 rpm and less but behave badly at 150 rpm and above?

Would you post the specs of the stepper motor, particularly the inductance. Its not impossible that you have hit the maximum reliable
stepping speed of those motors.

For them to be 8.5Nm they almost have to be 34size? Larger stepper motors have greater inductance and will struggle to retain useful torque
at high rotational speeds.

Craig

[Varske]

Hi,
you say that you can only travel at 3m/min? Are the ballscrews direct coupled, I would suspect they are?

If that's the case 3m/min corresponds to a stepper rotation of 150rpm....does that sound correct? And again to confirm
that they perform adequately at 150 rpm and less but behave badly at 150 rpm and above?

Would you post the specs of the stepper motor, particularly the inductance. Its not impossible that you have hit the maximum reliable
stepping speed of those motors.

For them to be 8.5Nm they almost have to be 34size? Larger stepper motors have greater inductance and will struggle to retain useful torque
at high rotational speeds.

Craig

Thanks for reply.
150rpm sounds correct, but motors torque is not the problem as Y axis can move 10m/min. Yes, these are nema 34 with 3.8 inductance and around 7A. Connected directly. I am pretty sure mechanics (except gantry 'walk') have nothing to do with this.
And as I mentioned before even at low speeds motors vibrate sometimes, but does not shut off completely.

Today I will try to use jumpers from one driver's encoder pins to another.

[joeavaerage]

Hi,
well 3.8mH is a respectably low inductance for big steppers and I would expect that they should go a damn sight faster than 150rpm before
running out of steam.

The problem is not with the steppers holding torque, it has that in a-plenty, the problem is that the torque diminishes as the speed goes up.
I would guess that the 8.5Nm holding torque would reduce to about 25-30% at 500 rpm, so from 8.5NM to around 3Nm at 500 rpm.
This reduction is a plain fact of physics of stepper motors.

If you believe the line of BS the manufacturers spout about closed loop increasing speed or torque....then I got some bottom land I want to sell you.
Hey...would I put you wrong!!!!

Still it sounds like you have to tune the stepper drive. Do you have the manufacturers tuning software? Do they have have any procedure for it?
Have you asked them for advice?

Craig

[Varske]

There is not much of tuning options. 4 switches for microstepping, 1 for spinning direction and 1 for 'up edge and down edge'. Not sure what the last one does.
Jumper between encoder pins didn't give expected result. Even worse, caused some serious gantry twisting and racking.
At this point I got one more idea that ballscrew coupler is not tightened enough. That could explain symptoms and tendency to fail at higher speeds and acceleration. Don't have a chance to put my hands on today, will check tomorrow.

[joeavaerage]

Hi,
I glimpsed the manual and there are all the PID tuning parameters, K_p, K_v and K_i.

What about those....do you have the tuning software and the programming cable to tune the drive parameters?

Craig

[joeavaerage]

Hi,
this is from page 17 of the manual:
Attachment 438712

Craig

[ger21]

Try changing the microstepping amount. What do you currently have it set at?
Sounds like either resonance, or maybe PID tuning as Craig mentioned.

[Varske]

Thank you, Craig. I see there is a lot I need to learn. Unfortunatelly I have no cable nor software. Need to dig deeper into this.

At the moment i have set 3200 microteps per rotation. Tried lower numbers and situation was even worse.

[Varske]

Tested with 51200 microstepping. Failed with even lower speed. The only difference that it just make a 3-4cm leap and dies instantly instead of shaking and switching off.
Couplers are well tighten.
I guess its time to read manual.

[joeavaerage]

Hi,
I understand that you are testing microstepping hoping to solve a problem but I suspect it will be to no avail.
This principle benefit of microstepping is smooth motion. Microstepping beyond 8 microsteps per full step (1600 step per rev)
is in to the realm of diminishing returns. Most people settle on on the low side, 4 or 5 microsteps, and mid range, 8 or 10 microsteps
and 16 or 20 microsteps on the high side.

You have investigated sufficiently to know that microstepping is not the cause of the fault and has a marginal role to play. May I suggest
set 8 or 10 microsteps per fullstep and leave it there. Changing it all the while just introduces more uncertainty.

Craig

[Hafco]

Hi guys, I’m new here.
Im learning to mill on an rf-30/rong fu type mill. I work in IT and would like to set it up as cnc. From what I’ve seen on line the results are hit and miss. I don’t know anything about the technicalities of cnc setup but clearly you guy do. I’d like to talk to someone who could walk me through the process.

Thanks Dave

[jannos]

Are you sure, that power supply is working correctly? I had similar problem and found "cold" solder in power supply output connector.

[cyclestart]

Hi guys, I’m new here.
Im learning to mill on an rf-30/rong fu type mill. I work in IT and would like to set it up as cnc. From what I’ve seen on line the results are hit and miss.

The 'hit and miss' being due to a round column ? It's not ideal.

edit/ just realized this is well off-topic. It would be better to create you're own thread for this question.

[Songshunsteel]

about the axis, maybe you can pick the 4140 steel round bar, its properties also can satisfy it.