Problems with 5-start leadscrew
I went with the 5-start 2TPI leadscrew with the plastic anti-blacklash leadnut from CNC Router Parts.
It may be fine for 2D cutting. But the motion of 3D carving has really stressed it after ~15 hrs or so. It looks shot to me.
Well, it "looks" fine, actually. But the zero-backlash promise is no longer good. I can lift that axis by hand and make it shift maybe like 0.05". Under regular use, the weight of the spindle usually prevents backlash from being an issue, but on plunges and sudden 3D moves and when using downcut bits, that may not be the case.
I do have a heavy spindle, with the Z-axis and stepper it's about 28 lb IIRC, and have had it set at 150ipm/75 in-sec^2. I don't have a counterweight on it but I do want to add one. I realized that leadscrew is getting HOT during an involved 3D job. Like the screw is almost too hot to touch in the range it's going up/down. So... not surprising the nut isn't up to spec anymore.
Should I just switch to a ballscrew? That seems like what I need to do here.
I've heard you don't usually see ballscrews on a Z, in part because they can backdrive when the drives are off and you don't have a brake, allowing the Z to drop. Well it already does that, slowly, but it'll drop on its own in a few minutes, if the dust skirt is left on it'll crush the bristles. The counterweight spring would prevent that on the plastic nut, but it's not gonna be a perfect match for weight... it might still reoccur if I had a high-efficiency ballscrew here.
Hmm the stepper's max power should be about 1200 RPM. With that 2TPI 5-start leadscrew, that would be 600 ipm, but it certainly can't run that fast, the axis stalls much past 150 ipm. So it's giving too much inch/rotation here. A finer pitch would improve it greatly, but ballscrews of an appropriate diameter only go down to 4 cm/rot.
Re: Problems with 5-start leadscrew
Go with a ballscrew. You won't be disappointed. I really don't see why anyone would use acme anymore with the low cost of ballscrews.
Ben
Re: Problems with 5-start leadscrew
Hmmmm... First, the AB leadnuts are rated to 25lbs. That would likely dismiss its use in your application. A counterweight should help. Also if your Z sees this much use some dry lube should help in conjunction with a counterweight.
Second. I do quite a bit of 3D work and cut some abrasive stuff like G10/FR4 and still have the same nuts for 6 years. However my spindle is not as heavy as yours. Still I'd look at possible flex in the coupling and check your fixed end bearings as well.
Third, a ballscrew doesn't guarantee no backlash, unless it has a double nut and or is ground with at least C5 precision. It should however help with performance. You can find 16-02 precision ground surplus. 4mm pitch would be fine but you may need a brake to prevent dropping when the machine Is off. You can make a circuit with a relay that shorts the stepper coils when off, which may be enough hold to keep the Z from dropping.
Last, I would not run AB nuts at 1200 rpm for a long time. Then again I doubt your Z would reach that anyway. Most all commercial machines use ballscrews even on Z. It's more heavy duty, but realize there are accommodations necessary.
Re: Problems with 5-start leadscrew
Yeah I used some moly dry lube. That stuff is remarkable! It stayed stuck to the screw despite the constant contact.
Wait, I got confused. Stepper's max speed is ~1200 rpm. Max power is 600 which is what you tune the drive for. So a 4mm pitch would only be 94 ipm. No, I'm looking for 10mm pitch, max power would give 236 ipm. Or there's the 5mm pitch, putting the max-power at 118 ipm.
The 10mm pitch is close to what I have now in 5-start leadscrew (12.7mm pitch). Again, due to the weight of the spindle and inefficiency of the leadnut... and the stepper may not have been the best choice.... I limited it to 150 ipm. IIRC it would stall on the lift regardless of acceleration somewhere in the 300 ipm range. Counterweight would fix most of that though.
Re: Problems with 5-start leadscrew
Had you used oil on your screw, the nut would probably still be like new. The fact that the screw was hot should tell you that you're lube isn't working.
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
MechanoMan
Yeah I used some moly dry lube. That stuff is remarkable! It stayed stuck to the screw despite the constant contact.
Wait, I got confused. Stepper's max speed is ~1200 rpm. Max power is 600 which is what you tune the drive for. So a 4mm pitch would only be 94 ipm. No, I'm looking for 10mm pitch, max power would give 236 ipm. Or there's the 5mm pitch, putting the max-power at 118 ipm.
The 10mm pitch is close to what I have now in 5-start leadscrew (12.7mm pitch). Again, due to the weight of the spindle and inefficiency of the leadnut... and the stepper may not have been the best choice.... I limited it to 150 ipm. IIRC it would stall on the lift regardless of acceleration somewhere in the 300 ipm range. Counterweight would fix most of that though.
I think the counterweight would help greatly with the stepper and reduce stress on the lead nut.
If you can, look for a Parker HV343 stepper and E-AC drive on eBay. It will give you 300in-oz at 3000rpm (about 950in-oz holding torque) wired bipolar parallel. And the E-AC drive has its own power supply self contained. Should allow you to use a ballscrew with 4-5mm pitch easily with good speed.
Re: Problems with 5-start leadscrew
Well yeah but that's a NEMA34. I would need to use a different mount and drive and power supply.
I have come to appreciate the need for a NEMA34 drive on R&P. Not because NEMA23 isn't powerful enough- NEMA23 is plenty strong enough for torque. But tensioning the XL timing belt in the reduction stage causes too much load on the NEMA23's 1/4" shaft. There's only like one or two oddball NEMA23 motors with 3/8" output shafts. Not a thing on the ballscrew though.
If I was going NEMA34, of course there's a whole page of NEMA34 options on AutomationTechnologiesInc. I'd think the Leadshine drive would be the way to go, too. Well, I see an 80v/7A drive Leadshine DSP drive, and there's 1200 oz-in steppers that pair with that. I know they have much greater rotor inertia that loads down the acceleration profile, but I haven't tried to figure out how significant that is vs the Z-axis weight.
I do like the 3D work and want to optimize the machine for that. Mach3 is turning out to be a limiting factor, though. For reasons no one seems to be able to explain, it's taking substantially longer than it should. The total runtime doesn't scale 1:1 when you double the speed & acceleration- it does take less runtime, but the runtime doesn't halve itself. And comparing it to LinuxCNC runtime, LinuxCNC is about 20% faster and actually does speed up proportionately when you increase the axis capabilities. I don't know why this is, but it's a Mach3 thing that limits 3D work.
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
MechanoMan
Well yeah but that's a NEMA34. I would need to use a different mount and drive and power supply.
I have come to appreciate the need for a NEMA34 drive on R&P. Not because NEMA23 isn't powerful enough- NEMA23 is plenty strong enough for torque. But tensioning the XL timing belt in the reduction stage causes too much load on the NEMA23's 1/4" shaft. There's only like one or two oddball NEMA23 motors with 3/8" output shafts. Not a thing on the ballscrew though.
If I was going NEMA34, of course there's a whole page of NEMA34 options on AutomationTechnologiesInc. I'd think the Leadshine drive would be the way to go, too. Well, I see an 80v/7A drive Leadshine DSP drive, and there's 1200 oz-in steppers that pair with that. I know they have much greater rotor inertia that loads down the acceleration profile, but I haven't tried to figure out how significant that is vs the Z-axis weight.
I do like the 3D work and want to optimize the machine for that. Mach3 is turning out to be a limiting factor, though. For reasons no one seems to be able to explain, it's taking substantially longer than it should. The total runtime doesn't scale 1:1 when you double the speed & acceleration- it does take less runtime, but the runtime doesn't halve itself. And comparing it to LinuxCNC runtime, LinuxCNC is about 20% faster and actually does speed up proportionately when you increase the axis capabilities. I don't know why this is, but it's a Mach3 thing that limits 3D work.
Well yes you'd need a different mount. But you got to reconfigure things anyway with a ballscrew. And the E-AC drive contains its own power supply... plugs into 110VAC, outputs 170VDC.
I like Leadshine stuff, but it doesn't have the performance of the Parker set. You can find on eBay for under $100 each. If you want to stay NEMA33 check out Lin Engineering new high torque steppers. NEMA23 size and 600in-oz torque and only about 3-1/2" long. Plus it's USA made.
As to Mach3, check the LookAhead number in the General Config page. It's set to an abysmally low number by default like 10 or 20. Depending on your CPU speed it can be set way higher. I have mine set yo 200 but the max is 1000. Turn CV on and try tweaking CV distance and CV angle. There's a white paper explaining these settings by Ron(?) Bloomfield of Arcane Innovations, I think it's a must read. Finally a faster motor with more reduction may give you better.performance and you may be able to raise your acceleration on the Z which will help 3D.
Re: Problems with 5-start leadscrew
Really? Parker outperforms Leadshines? I've been using the Geckodrive G540. I figured the Leadshine's DSP stuff would be top-notch performance.
I did try the stuff with Lookahead and making sure CV mode was used. It didn't yield any improvement though.
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
MechanoMan
Really? Parker outperforms Leadshines? I've been using the Geckodrive G540. I figured the Leadshine's DSP stuff would be top-notch performance.
I did try the stuff with Lookahead and making sure CV mode was used. It didn't yield any improvement though.
Leasahine is very good... But the Parker runs at over double the output voltage.... And you don't need separate PSU.
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
louieatienza
Well yes you'd need a different mount. But you got to reconfigure things anyway with a ballscrew. And the E-AC drive contains its own power supply... plugs into 110VAC, outputs 170VDC.
Didn't know it output that high of a voltage. I have one E-AC hooked up to a 4th axis and the motor does get pretty hot. I figured the internals were similar to the E-DC drives, which is only 48volts max input voltage. I have 4 new E-DC drives still in the box waiting for the 3d printer to get built. If it does output 170volts, I guess that is why the motor gets hot. By the way these Parker drives seem to be good or better than the gecko's I have.
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
jfong
Didn't know it output that high of a voltage. I have one E-AC hooked up to a 4th axis and the motor does get pretty hot. I figured the internals were similar to the E-DC drives, which is only 48volts max input voltage. I have 4 new E-DC drives still in the box waiting for the 3d printer to get built. If it does output 170volts, I guess that is why the motor gets hot. By the way these Parker drives seem to be good or better than the gecko's I have.
The insulation on most stepper wire is rated to 340-500VDC... Kollmorgen has a drive that outputs 320VDC....
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
jfong
Didn't know it output that high of a voltage. I have one E-AC hooked up to a 4th axis and the motor does get pretty hot. I figured the internals were similar to the E-DC drives, which is only 48volts max input voltage. I have 4 new E-DC drives still in the box waiting for the 3d printer to get built. If it does output 170volts, I guess that is why the motor gets hot. By the way these Parker drives seem to be good or better than the gecko's I have.
Well the nature of these drives is it selects a constant-current, it can be the current selected or a fraction based on what microstep you're on.
The voltage self-adjusts based on what's needed to create the current. In the case of low speed stepping, it's dominated by V=I*R. As you step faster, the inductance becomes significant and higher voltages are needed to meet the target current. Motor heating at low speed is all I^2*R heat from the winding resistance. At higher speeds, there is some eddy current/iron hysteresis loss heating from the switching. But it's unrelated to the voltage.
We care a lot about voltage because at higher speeds, if the drive can't source sufficient voltage to meet the target winding current, then the current target won't be met at the peaks of the phases and the torque is weaker and more distorted. The likelihood of a stall increases.
The required voltage depends on speed and motor inductance. Most NEMA23 motors do well with 48v DC sources.
Re: Problems with 5-start leadscrew
There shouldn't be a lot of heat at low speeds since the current is limited by the drive, otherwise the stepper would draw so much juice it would destroy itself. This is why the torque curve is more or less flat till it reaches the corner speed as Marriss of GeckoDrive calls it, where the drive transitions from current source to voltage source. So there are separate torque curves for each voltage. And the corner speed is proportional to the applied voltage. Simply put, the more voltage, the more useable torque.
Would I drive a Wantai stepper with 170VDC? Maybe not. Some steppers are just built better.
I am at a loss however as to why you need our advice, if you seem to know the answers to your own questions?! :)
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
louieatienza
The insulation on most stepper wire is rated to 340-500VDC... Kollmorgen has a drive that outputs 320VDC....
Just an observation. The temp is no were near the max rated temp the steppers can handle. I just notice it running hotter than the ones driven by gecko201's. Same 276oz-in automation direct stepper motors. I don't use the rotary axis often so most of the time it is unplugged anyway. I got the E-AC driver for cheap on eBay and having it plugged into the AC made it easier to shut it off when not being used.
Re: Problems with 5-start leadscrew
Quote:
Originally Posted by
louieatienza
There shouldn't be a lot of heat at low speeds since the current is limited by the drive, otherwise the stepper would draw so much juice it would destroy itself. This is why the torque curve is more or less flat till it reaches the corner speed as Marriss of GeckoDrive calls it, where the drive transitions from current source to voltage source. So there are separate torque curves for each voltage. And the corner speed is proportional to the applied voltage. Simply put, the more voltage, the more useable torque.
Would I drive a Wantai stepper with 170VDC? Maybe not. Some steppers are just built better.
Ideally you don't ever want the drive to go into the mode where it's voltage limited- can't get current to rise fast enough to get the winding to the target current within the duration of the step. Once you run out of voltage the drive's torque drops badly. Like I say, 48v on a NEMA23 is usually sufficient. 24v isn't, not when using higher inductance steppers found on CNC routers and trying to run them at high speed.
I was just wondering where it came from that these Parker E-AC drives use 170v. I didn't see it on the spec sheet. I am suspicious of whether their power is through an isolation transformer or not, because that's suspiciously similar to running directly off-line. But I'm pretty sure it's transformer-isolated.
Quote:
Originally Posted by
louieatienza
I am at a loss however as to why you need our advice, if you seem to know the answers to your own questions?! :)
Well TIL Parker drives are a thing. I only knew about GeckoDrive and Leadshine.
I also learned about these Lin Engineering motors... they DO seem to be notably better, specifically the inductance is lower than I'd expect for a motor of that torque and current in that frame size.
