"X" axis ballscrew locking up at high speed

[Barefootboy4]

Has anyone had this problem. My X axis ball screw which is 60" long and 20mm dia is locking up when I run my feed rate over 100IPM. It works fine if I slow it down but just won't run faster. It should do it but when doing rapids it will lock up and then the Mach 3 will lose steps and we all know what happens next. The ballscrews are from Mr. Chen in China but the Y and Z axis work fine at high speed. There is no whipping of the X axis and it is not bent.

Does anybody have any experance with this issue or can steer me in the correct direction??

Thanks for the help!

Similar Threads:

• Looking for some "high"er speed spindle ideas for my bf20 clone
• High Speed 12,000 CAT 40 3/4" x 1.75" End Mill Holder
• High Speed 12,000 CAT 40 1" x 4" End Mill Holder
• High Speed 10,000 CAT 40 7/16" x 2.5" End Mill Holder
• My "Crazy" high speed spindle test

[keebler303]

I think you may be confusing the ballscrew locking up with the motor stalling because you are asking too much of it. Now is the ballscrew actually jamming or something or does it just quit turning? Once you exceed a certain speed/torque threshold, the stepper will stop spinning and will just vibrate wildly until the motion is finished and it is allowed to accelarate at an acceptable rate again.

You can try reducing the acceleration rate so that it has a smoother speed up which would reduce stalling. Also make sure your bearing s and everything are free so you have no binding. If everything looks good mechanically then you just have to make do with the speeds which are reliable or buy a bigger motor to prevent exceeding its torque output at speed.

Matt

[Bear5k]

My vote is for the motor, as well. A ball screw that "locked up" would leave a mechanical trace of some sort, up to and including an ugly noise and a bad smell since this would be a friction issue. This would also then preclude "working fine" at lower speeds.

[louieatienza]

You probably have the 5mm pitch ballscrews, which would mean that at 100ipm, your screw has to spin at 500rpm. May not be a problem for your y and z axes which are smaller and lighter, but maybe for your gantry. You could use a higher powered stepper and gear it up. Or you might havev some resonance in the gantry axis, where a resonance damper or "rattler" might help. Another option could be to use a screw with a higher lead.

[Barefootboy4]

Guys,
Thanks for the reply's. I hate to say it but you may all be right. I have NEMA 34's with 640oz of torque which should handle the higher IPM's but sounds like I may be wrong. The same size on both Y & Z and they rip but less weight and forces to deal with. Could swap up to a larger X axis motor which may fix the problem. You are correct and my ballscrew is a 5mm pitch. I'm down to approx 80IPM's on the X axis and it sucks.

What do you feel is the best answer. Higher pitch or stronger stepper motor?? Also what should the max RPM on my motor be. Again keep the ideas coming.
Steve in Maine

[Bear5k]

Guys,
Thanks for the reply's. I hate to say it but you may all be right. I have NEMA 34's with 640oz of torque which should handle the higher IPM's but sounds like I may be wrong. The same size on both Y & Z and they rip but less weight and forces to deal with. Could swap up to a larger X axis motor which may fix the problem. You are correct and my ballscrew is a 5mm pitch. I'm down to approx 80IPM's on the X axis and it sucks.

Look at the torque curve for the motor. While you may start at 640 oz-in of force at 0 rpm, that number will start to fall off pretty rapidly at just a few RPM. What's important when looking at a torque curve is to pay attention to the units on the X-axis. This is often given in PPS (pulses per second) and, worse, can even be shown as a logarithmic scale to smooth-out the drop in higher-rpm torque.

At 500rpm, you are at 100,000pps with a typical 200 step/rotation stepper motor. I don't know if your motors are from Keling, but there is a 640oz-in motor on Keling's website that includes a torque curve -- of a sort (attached picture #1).

http://www.kelinginc.net/KL34H280-45-8AT.pdf

If you look at this graph, and overlook the excessive use of BOLD fonts, you can see that this particular motor starts out at about 4.25 N-m at around 500 PPS and then drops down to about 1.4 N-m at 6000 PPS. Those sound like big numbers until you divide by 200 steps (pulses) per revolution. In terms of RPM, this motor loses 2/3rds of its torque going from 2.5rpm to 30rpm. The response curve, if it is like most of the other published stepper motor curves, is probably going to be sigmoid in shape (elongated "S"), so the drop from 30rpm to 500rpm won't be as steep as this early drop, but it will still be a drop.

What do you feel is the best answer. Higher pitch or stronger stepper motor?? Also what should the max RPM on my motor be. Again keep the ideas coming.

Setting aside the resonance issue, which should be ruled out, the "best" answer is going to be a servo motor, given available technology. The best answer given typical constraints on budgets is "it depends". One answer would be to change your chip load and to take a lighter (less deep) cut using the existing equipment. If you are lucky, you can gain more speed (proportionally) than you take away in depth (e.g., halving your DOC gives more than a 2x increase in speed). The other options may not be as good.

If you double the size of your stepper motor, and hold the shape of the torque curve constant, you will probably get up to 200ipm before stalling assuming no change in cutting parameters.

If you changed your ball screw to a 10mm lead, you would also get out to 200ipm before stalling, but you have to go through the hassle of re-mounting the screw and aligning it properly (a slight mis-alignment can increase the load on the nut, so you may want to spend some time checking the screw alignment just as a precaution). Changing to a 10mm lead will also halve the resolution of your machine in that axis, so that may not be desirable.

For me, I pretty much committed to servo motors from the outset, and my current choice for my own build are the DMM-Tech units. Specifically, the big NEMA34 unit for my own X-axis. It is limited to ~1000rpm (that pesky 200ipm limitation for your machine -- again!), but the torque curve is much more favorable to cutting at higher speeds (attached picture #2).

Motor

If your gantry weighs significantly less than my own does/will (I'm hoping to sneak in under 80kg when fully loaded; I'm over 65kg before adding the Z-axis or motors), then you can probably happily use the 400W motor and get beyond the 200ipm limitation (critical speed will factor into this before you hit 3000rpm). The price differences are not much between the options (~$275 - $325/axis plus power supplies and cables), so suitability is the bigger criteria.

So, the net effect is that you should check your set-up to ensure that you aren't experiencing a resonance issue or a misalignment issue that is causing the motor to stall. If it isn't either (or both) of those, more speed means that it is time to spend more money. You can double your speed by doubling your motor size or by doubling your screw lead. The first can only go so far given electrical limitations, and the second has the drawback of losing resolution. Changing motor technology can also help, potentially a lot, but at a higher cost. No matter which route you go, though, at some point you will hit your screw's critical speed and then you will go no faster.

[ger21]

What do you feel is the best answer. Higher pitch or stronger stepper motor?? Also what should the max RPM on my motor be. Again keep the ideas coming.

Depends on the machine. Any pics? What voltage are you using, and what drives? What's the screw diameter?

You may find, that a smaller motor will be faster. Increasing the motor size, may give you more holding torque (at zero rpm), but may have even less torque at the speed you're running at.

Speed is generally proportional to voltage. If you double the voltage to your motors, you should get roughly double the speed.

[Barefootboy4]

I don't have any pics due being offshore on my ship I work on. It's basicly a shopbot copy. The ballscrew is 20mm dia 5mm pitch 60" long. The motors and drives are the standard KIT from Keling. I have no complaints about the kit other than the NEMA 34 stepper is not cutting it. Running the standard Keling 50VDC power source.

If I replace the X stepper with a Keling NEMA 34 servo @ 1200oz would that fix my problem. Do I need a seperate power source for the Servo motor. Can I still run my Y & Z on my current 50VDC and supply higher DC for the X motor? I just can't live with 80 IPM for my X axis. Will the Mach 3 program run a mix of stepper and servo motors? Any help here is great. This is good solid data from everyone.

Also what about a Rattler. How do you make it and mount it. How much weight?
Steve in Maine

[louieatienza]

I don't have any pics due being offshore on my ship I work on. It's basicly a shopbot copy. The ballscrew is 20mm dia 5mm pitch 60" long. The motors and drives are the standard KIT from Keling. I have no complaints about the kit other than the NEMA 34 stepper is not cutting it. Running the standard Keling 50VDC power source.

If I replace the X stepper with a Keling NEMA 34 servo @ 1200oz would that fix my problem. Do I need a seperate power source for the Servo motor. Can I still run my Y & Z on my current 50VDC and supply higher DC for the X motor? I just can't live with 80 IPM for my X axis. Will the Mach 3 program run a mix of stepper and servo motors? Any help here is great. This is good solid data from everyone.

Also what about a Rattler. How do you make it and mount it. How much weight?
Steve in Maine

I think the torque curve on the larger steppers drops more than the smaller ones. You can mix servo/stepepr with mach3 but I don't know if that would be ideal. You COULD use the larger stepper and gear it UP. But I would be inclined to try a higher-lead screw. Even with a servo, you'll be spinning that ballscrew clsoe to it's max to get a decent rapids speed. I don't know the diameter of your ballscrew, but you can check Nook's online calculator to see what the max rpm you can expect, given you screw length and diameter.

Luckily, since most people think they need fine-pitched ballscrews to achieve accuracies and tolerances that are unrealistic to hold for most, finding high-lead ballscrews, even at your length, aren't the toughest to find at a good deal on eBay.

An obvious thing to do would be to disengage your ballnut from your gantry and move it back and forth, and check to make sure that it [the gantry] moves freely and consistantly at all points and that it is not racking when moved.

As for the "rattle" I cannot remember where the thread is, bu tyou should be able to find it by doing a serach here on the Forum....

[Bear5k]

I think the torque curve on the larger steppers drops more than the smaller ones.

In general, yes, but for good and very specific reasons. Stepper motors are, effectively, constant power devices - though you have to calculate both the power delivered by the shaft AND the back EMF from the inductor to get there. Bigger torque requires a greater magnetic field which requires larger induction. Induction and voltage are what control the shape of the torque curve.

You can have a low torque motor with big inductance and high voltage, but don't get anything flammable near it since it will be dissipating a ton of heat (high resistance) in the process. It is definitely possible to have a low torque motor that loses a lower percentage of its power as RPM climbs, but it will start from a smaller base (lower inductance). It's a "testable hypothesis" as to which motor has more power at high RPM between a bigger motor and a smaller motor, but if there is a big disparity in holding torque, chances are that the bigger motor will still have appreciably more torque at high RPM assuming that both are well-designed and manufactured (the smaller chassis will have trouble with the higher currents used by the larger motor). It's pretty easy to find NEMA34 motors requiring more than 80+ VDC, but a small NEMA17 motor would most likely fry at those kinds of voltages.

Stepper motor - Wikipedia, the free encyclopedia

Conversely, a servo motor is a constant torque device, within reason. It's power is a direct scaler with RPM (simplistically: Power = Torque * rpm). At high RPM, back EMF kills the amount of torque a stepper motor can apply to a load while a servo motor still has the same torque at 10 RPM as it has an order of magnitude or three later.

Stepper Dampers: http://www.cnczone.com/forums/steppe...er_damper.html

[ger21]

If I replace the X stepper with a Keling NEMA 34 servo @ 1200oz would that fix my problem.

As was mentioned, it might get worse. The bigger the motor, the slower they usually spin.

Do I need a seperate power source for the Servo motor.

Depends on the servo. A servos rpm is directly proportional to voltage. You'd need to know the ratings, and how fast you want it to spin.

Can I still run my Y & Z on my current 50VDC and supply higher DC for the X motor?

Yes.

Will the Mach 3 program run a mix of stepper and servo motors? Any help here is great. This is good solid data from everyone.

Yes, but depending on the encoders, you may need a higher kernel speed.

Also what about a Rattler. How do you make it and mount it. How much weight?

http://www.cnczone.com/forums/steppe...tml#post256639

[michaelthomas]

It sounds cheesy, but simply drilling a hole in the center of a hockey puck and pressing it on the rear shaft of the stepper can do wonders for getting past the mid band resonance that you may be experiencing.

There is a thread about that here somewhere , too.

If it helps......you could spend some time making a better damper.

[louieatienza]

In general, yes, but for good and very specific reasons. Stepper motors are, effectively, constant power devices - though you have to calculate both the power delivered by the shaft AND the back EMF from the inductor to get there. Bigger torque requires a greater magnetic field which requires larger induction. Induction and voltage are what control the shape of the torque curve.

You can have a low torque motor with big inductance and high voltage, but don't get anything flammable near it since it will be dissipating a ton of heat (high resistance) in the process. It is definitely possible to have a low torque motor that loses a lower percentage of its power as RPM climbs, but it will start from a smaller base (lower inductance). It's a "testable hypothesis" as to which motor has more power at high RPM between a bigger motor and a smaller motor, but if there is a big disparity in holding torque, chances are that the bigger motor will still have appreciably more torque at high RPM assuming that both are well-designed and manufactured (the smaller chassis will have trouble with the higher currents used by the larger motor). It's pretty easy to find NEMA34 motors requiring more than 80+ VDC, but a small NEMA17 motor would most likely fry at those kinds of voltages.

That's all fine and good, but that doesn't answer the question of whether a larger stepper would help or not, which is the question Barefootboy4 asked. We're all shooting blind here in a way, since we do not have pics of the machine , or know what electronics are used, or even if the computer used is fast enough... As for a stepper's torque at a given RPM, all one needs to do is simply look at the manufacturer's charts/

A good reference on stepper basics is Marriss' on the Geckodrive website. Basically there is a difference between a stepper's torque and its maximum power output; and that you want to utilize the smallest stepper that generates the power you need at around it's "corner speed." Otherwise you're just wasting electricity (and money on higher torque steppers and bigger power supplies), unless you have a cold shop in the winter and need supplimental "heaters..."

His site also has all teh formulas yo'll need to calculate approximately what size stepper you may need...

[Bear5k]

That's all fine and good, but that doesn't answer the question of whether a larger stepper would help or not, which is the question Barefootboy4 asked. We're all shooting blind here in a way, since we do not have pics of the machine , or know what electronics are used, or even if the computer used is fast enough... As for a stepper's torque at a given RPM, all one needs to do is simply look at the manufacturer's charts/

Agree on the charts, but they may die off quickly or need to be converted into something more useful (e.g., the one I linked, above, stopped at 6,000pps or 1,800rpm) or have transforms on the axes that require additional interpretation (e.g., log scales).

A good reference on stepper basics is Marriss' on the Geckodrive website. Basically there is a difference between a stepper's torque and its maximum power output; and that you want to utilize the smallest stepper that generates the power you need at around it's "corner speed." Otherwise you're just wasting electricity (and money on higher torque steppers and bigger power supplies), unless you have a cold shop in the winter and need supplimental "heaters..."

Agree on the Gecko Drive support page being a good starting point. I'll link it here:
Support

I'd prefer the discussion be in SI units and then translated back to English/Imperial, but it still works.

One thing of note since it is important, within practical limits, a stepper motor is a constant power motor (power in watts = torque in Newton-meters * 2pi * rev / sec). Torque is what will vary, and this is one of the fundamental questions for the OP: can his current set-up be tweaked/tuned so that the motor produces sufficient torque at the requisite RPM so that his screw doesn't stall?

Things that can cause a loss of power in the current set-up:
- Resonance
- Mis-wired motors
- Defective drive
- Bad software set-up
- Etc.

In addition to the electrical/electronic issues for loss of power, I would add checking the parallelism of the rails to the screw since a slight misalignment might cause incremental binding (increased load) that isn't being detected at lower speeds where more torque is available to push through (when comparing available torque vs. theoretical load).

His site also has all teh formulas yo'll need to calculate approximately what size stepper you may need...

That part I'm not sure I agree with, especially since this is diagnosing a stepper that is already in service. The OP can go back and "see how well he did", though.

Other background that might help:
Jones on Stepping Motors (the discussion on resonance is a little deep, but this math should be crucial for the OP's current issue -- and not covered by Mariss, above)

So that I am not being taken out of context, when I talked about "holding the shape of the torque curve constant", above, I meant just that. Practical motors can have big swings in the shape of the torque curve depending upon the inductance and the voltage/current. Here is a prime example from http://sanyo-denki-online.com/nema34.htm. Compare the intercept of two NEMA34 motors with the 1 N-m curve, since that's easiest (note the log scales on the X-axis, and the stray graphed line with no legend).


388oz-in, 4A, 0.6Ohm, 1.65mH
1 N-m @ ~8500pps or 42.5rps or 2,550rpm
Zero(?) torque at ~15,000pps


720oz-in, 4A, 2.8V 0.7Ohm, 5.7mH
1 N-m @ ~7500pps or 37.5rps or 2,250rpm
Zero(?) torque also looks to be about 15,000pps

The increased inductance in the larger motor is responsible for the steeper slope of decline in torque through the mid-band. However, both motors reach practical limits at about the same point (15,000pps).

Having looked up the Keling driver I believe the OP is using (http://www.kelinginc.net/KL-5056.pdf), checking the set-up (wiring, microstep settings/DIP switches, etc.) would be a critical first step in troubleshooting this. I had thought the Keling Chinese drivers did not include microstepping, but the spec sheet says otherwise.

[louieatienza]

"That part I'm not sure I agree with, especially since this is diagnosing a stepper that is already in service. The OP can go back and "see how well he did", though."

The think is, we already know the stepper isn't performing. Barring mis-wiring, inadequate wiring, or malfunctioning driver, we know that:

A NEMA34, 640 in*oz motor is used
The highest rapids acieved without stalling is about 80 ipm
A ballscrew with 5mm pitch is used (about 5tpi roughly)

So the stepper has to spin at approximately 400 rpm to achieve 80 ipm. Maybe the charts you posted are a bit deceiving, since the Sanyo-Denki steppers are capable of reaching 3000rpm, where the Keling steppers probably do not. The solution might be to use the Sanyo-Denki steppers and gear them down, to use the motor's higher rpms to get more power, speed and resolution.

Since we tend to use what we got, and the stepper is working fine otherwise, I think the easiest solution would be to get the same diameter leadscrew with a higher pitch, maybe 10mm or 12mm if available. Then all the other components could be used.

[ger21]

The increased inductance in the larger motor is responsible for the steeper slope of decline in torque through the mid-band. However, both motors reach practical limits at about the same point (15,000pps).

For all we know, the higher inductance motor may have been driven at double the voltage. The charts are meaningless without all the info.

And I wouldn't consider the "practical limit" the point where the motor has zero torque.

Compare the intercept of two NEMA34 motors with the 1 N-m curve, since that's easiest (note the log scales on the X-axis, and the stray graphed line with no legend

Are you aware that each graph is comparing two same size motors, one double the current rating of the other? In the first graph, the 2a motor has 1 Nm at 5000pps, while the 4a motor still has 1NM at about 8500pps. I'm pretty sure that comparing the smaller motor with the larger one is apples to oranges, and are most likely at different voltages.

Maybe the charts you posted are a bit deceiving, since the Sanyo-Denki steppers are capable of reaching 3000rpm, where the Keling steppers probably do not

A lot of people consider the maximum usable rpm of a stepper to be 800-1000 rpm.

We don't really know what those charts mean, but consider this. Bear5K is saying the practical limit is 15000pps, but the reality is that a 5Nm stepper has lost half it's torque at 1200pps, or 8% of it's practical limit.

If I replace the X stepper with a Keling NEMA 34 servo @ 1200oz would that fix my problem

I gave you the wrong answer before, as I was thinking 1200 oz stepper, not servo. Most likely, the servo would be faster. But I don't know how much.

[Bear5k]

For all we know, the higher inductance motor may have been driven at double the voltage. The charts are meaningless without all the info.

Specs are specs. If a manufacturer wants to lie/cheat/steal, then so be it. However, most people who work in manufacturing rely on specs at some point, even if the engineering samples from selected vendors go through extensive in-house testing to verify the manufacturer's or vendor's specs.

And I wouldn't consider the "practical limit" the point where the motor has zero torque.

My point was more that a motor's torque should be asymptotic to zero. I find the "zero torque" intercept somewhat entertaining since that's not really where you want to operate a stepper. In the real world, the practical limit is when the motor ceases having enough torque to do what you want it to do. The hope for most of us is that there is sufficient headroom that this is in our idealized performance envelope, and not within the work-a-day envelope. Unless one is buying from a tier 1 supplier like Parker or Thomson (Kollmorgen), one is unlikely to get truly useful information about what a motor's limits are, so one has to sacrifice a few barnyard animals to determine where truth lies before one spends one's money...

Are you aware that each graph is comparing two same size motors, one double the current rating of the other? In the first graph, the 2a motor has 1 Nm at 5000pps, while the 4a motor still has 1NM at about 8500pps. I'm pretty sure that comparing the smaller motor with the larger one is apples to oranges, and are most likely at different voltages.

They are both 4A motors that I'm comparing (I said they were terrible charts -- there are actually TWO products on each with a bad legend to differentiate them and no color coding). The bigger difference is in the inductance, hence the highlighting. The larger inductance presumably has a larger stack than the smaller motor. What matters, though, is the performance. One might appreciate a practical example of the conventional wisdom running around where the smaller motor actually does have more torque at higher RPM than the larger motor.

I think I've offered what I can to the conversation, so best of luck to all!

[ger21]

One might appreciate a practical example of the conventional wisdom running around where the smaller motor actually does have more torque at higher RPM than the larger motor.

Explain how two motors, with nearly identical voltage and current ratings, can reach the same max speed when one has 3.5x the inductance of the other?

I still say the voltages used in the two charts are vastly different, making them meaningless.

The only place I know of where you can find "good" torque charts, is Oriental Motor. Try to find similar charts there.

[Bear5k]

Explain how two motors, with nearly identical voltage and current ratings, can reach the same max speed when one has 3.5x the inductance of the other?

They don't do so with the same amount of available (shaft) torque (more back emf in the higher inductance motor). However, it seems like we are talking in circles. It seems like you are coming back to my original point that one of the key ways of making the OP's machine faster, assuming his current set-up is properly, well, set-up is to go with a bigger motor.

I still say the voltages used in the two charts are vastly different, making them meaningless.

That's fine, but since I just grabbed them from The Google, I can't help you with this since the vendor is labeling the two I was comparing as 4 amp and 2.4V vs. 2.8V. It's a decent discrepancy, but not worth too much more bother for anyone involved.

The only place I know of where you can find "good" torque charts, is Oriental Motor. Try to find similar charts there.

I'm not really sure what the point or question would be since we seem to be agreeing in all but tone. Let me try to sum up again: bigger motors provide more torque, but there are examples where smaller motors might have more available torque at higher RPM, so one needs to check a spec sheet (or an industrial lab) for the specifics of what a specific motor will do in a given situation.

[louieatienza]

They don't do so with the same amount of available (shaft) torque (more back emf in the higher inductance motor). However, it seems like we are talking in circles. It seems like you are coming back to my original point that one of the key ways of making the OP's machine faster, assuming his current set-up is properly, well, set-up is to go with a bigger motor.

According to the charts you supplied, the 390 in*oz motor has slightly more torque than the 720 in*oz motor at around 2500 rpm. In fact the 390in*oz has 1 N*m torque at 2550 rpm, and the 720 has the same at 2250 rpm.

So let's say you have a driver/computer capable of spinning the Sanyo Denki motors at full potential. With the 5mm leadscrews, just off the top of my head, you'd be pushing the gantry with about 700 in*oz of force, at about 500ipm, if your ballscrews can handle it. For comparison, from my experience, I can cut hardwoods on my machine at over 100ipm, with 1/2" 8, 8 start ACME screws, NEMA 23, 425 in*oz steppers, at 3/8" depth with a 1/2" bit! And my gantry weighs ovver 60 pounds! I keep my rapids at about 250 ipm. Heck with my wood machine I was rapiding over 300 ipm because of the lighter gantry. If I had ballscrews with the same pitch I could probably achieve higher speeds. The r&p guys can rapid their heavier gantries at 1000 ipm, using smaller motors than mine.

Point is, you can't just slap another higher torque motor in, and expect better rapids (though you might have better performance for cutting speeds.) Evevrything else must be taken ito cosideration.