Speed of steppers. Anyone know about this?
Hi. I took a look around the site for an answer and now my wife is wanting to go out so I have 5min to ask!
I am converting my CNC router to ballscrews instead of the Festo belt linear actuators it has already. The Festo belts need maintenance and are a little loose for cutting round holes and alloy.The screws are 1605 and are so much more stable (twin on the Y) and accurate. But they are slow. If I speed up my steppers, they make a knock every once and a while which I can only account for losing steps. Also the couplers slip. So I have a machine that is extremely accurate now but very very slow to achieve the results I want.
I am using proper SFU rails and blocks that have a little bit of friction due to the seals.
What do I do here can anyone give some advice? It is easy to stop the couplers from slipping I guess by making some or keying the shafts but steppers losing steps is an issue. I am using Nema 23 approx 465" lbs torque.
Do I go for bigger Nema steppers? I do understand they lost torque with speed. Gearing up the ballscrews is possible but a lot of work.
I love my machine. It is 1800mm x 1400mm. The Y axis is about the maximum length for a ball screw but they do not whip so all is ok. It is looking amazing and is robust and tight. Just slow now if I want to keep my alloy cuts within tolerance. Timber it is not so much an issue.
A thousand apologies if I have run over an old story but my wife is waiting............. :-)
Regards
Boyd
Re: Speed of steppers. Anyone know about this?
If you want faster speeds you should get 10-25mm per turn screws instead of the 5 you have now.
Re: Speed of steppers. Anyone know about this?
Aw man..................... Hindsight is a ***** huh? Although what accruracy would I get with such a higher pitch screw? I would need bigger steppers too to drive them huh? Anyway the expense would make it not worth while. Bhurts is there any other option?
Re: Speed of steppers. Anyone know about this?
I have never used steppers but I think you would be fine with the ones you have and 10mm lead screws. As for accuracy it should be half the resolution but the same overall accuracy. You can try gearing up your motors 1to2 but no way to tell if it will cause issues. It is the cheapest option so maybe worth a try. If you decide to change screws you might be better of waiting till it's more feasible and go rack and pinion. Sorry I can't offer any low cost options. Someone may chime in with a option I haven't thought of. Good luck.
Ben
Re: Speed of steppers. Anyone know about this?
Hi Boyd,
I'm very interested in your experience with the Festo belt linear actuators. I'm putting together a gantry with DGE-ZR-25 belt actuators with 900mm lengths topped with a 960mm ballscrew
and a shorter ballscrew to raise the spindle. Does aluminum carving put too much stress on the belts causing the need for maintenance? Any info could help at this point. Thanks. Kid.
Quote:
Originally Posted by
boydage
I am converting my CNC router to ballscrews instead of the Festo belt linear actuators it has already. The Festo belts need maintenance and are a little loose for cutting round holes and alloy.
Regards
Boyd
Re: Speed of steppers. Anyone know about this?
Stepper RPM is largely related to the applied voltage. The higher the voltage the faster they will spin. Those sppers yo uhave should do about 600 RPM at 48 volts but your drivers may not be able to take that. At 600 you have scarce little torque left so that would be for rapids. with the 5 TPI ballscrews you can expect about 120 IPM tops. For cutting speeds you probably cannot exceed 60 IPM. At 50% of RPM yoou have about 50% of the holding torque. It starts to go non-linear after that on most motors because of inductance effects.
Bigger motors may give you higher cutting speeds but they tend to roll off quicker on torque so you won't get more RPM and quite likely less overall. If your current voltage is lower so will be your speeds. Only other options is courser ballscrews, belt increase between motor and ballscrew (which also ;pwers torque and resolution) or looking at a different kind of motor/controller.
If your speeds are a lto less than the 120 I listed you might look at replacing your controller with a G540Gecko and aa 48V 8 to 10A supply. Its the age old problem of speed versus torque and resolution being on opposite ends of the equation. With ballscrews you have way more resolution than you need. at quite a bit of torque multiplication. . You have to start worrying about lead whip the longer the ballscrew and the faster you spin it. To get 200 ipm you have to spin the 5 TPI at 1000 RPM. That might present probelms depening on the ballscrew mountings and diameters.
Re: Speed of steppers. Anyone know about this?
Hi Kid,
My first and second machine had the festo actuators. It was originally a rollerbearing/angle setup which I found to be jerky as I only used one actuator in the middle. So I changed to some proper linear rails thinking that would tighten things up. The festo actuators were still jerky when they moved. Almost like there was stiction inside them. They were find once they got moving but the little hesitation at the start of a move made all the difference to holes etc. I even tried putting a couple of pullies into the setup to increase the resolution and mechanical advantage of the steppers. No success - they are out. I couldnt be bothered rebuilding them.
Now I have a double 5mm pitch ballscrew setup which is rigid and accurate. In hindsight though, on a table as large as mine I would go for a 10mm pitch next time for a bit more speed.
Hey alloy should not put any stress on your system? Unless you are using large cutters and taking big bites? A 1/4" bit taking 0.5mm deep cuts doesnt really need much to push through.
I hope my little rant was of some help. Regards Boyd
Re: Speed of steppers. Anyone know about this?
Very simple.
Go with the ball screws for sure, at the pitch you have, but replace the steppers with servo drives (DC or AC).
Steppers: limited in size and power, and speed (600 rpm?)
Servos: essentially unlimited in power (how's your budget?), and speed to 6000 rpm if you want. (Can go higher, but how's your budget?)
Cheers
Roger
Re: Speed of steppers. Anyone know about this?
Roger has it.
You have zero other options, as no steppers will be significantly better.
Bigger steppers will be slower in max rpm- so you will not get better results in terms of speed. Low speed torque - yes, but speed, no.
Gearing will decrease your accuracy.
A high-rise screw will decrease your accuracy.
Basically, steppers only work okay to about 600 rpm in contact apps.
(I have run steppers at 10.000 rpm with the right hw. But thats irrelevant, as they then wont have any torque for actual load moving).
If you want to get better results re: speed, use servos.
Re: Speed of steppers. Anyone know about this?
Quote:
I don't think you'll run those Chinese ballscrews at 6krpm. Not many ground ballscrews are rated over 4krpm.
The balls sure would be flying around the recirculate tube!
But the point I was trying to make is that servos will go a LOT faster than steppers, with power.
Cheers
Roger
Re: Speed of steppers. Anyone know about this?
The servo approach has some "cons" as well. While servos have somewhat unlitimited torque (they keep drawing more current until they burn up unless the controller can shut it down. Its not uncommon to see a small 23 frame servo with a continuous run current of 3.5 A and torque of 50iz-in pull over 20A stalled (unless you limit it) yes it will spin 4000 RPM BUT it has the same torque at 1000 RPM where your motion would be around 200 IPM. So A small servo with encoder has 50 oz-in of running torque while the same basic sized stepper has over 300 ox-in of running torque (at 1/2 max RPM) . So one is a high torque, low speed device ant he other is a high speed low torque device. Now you can buy bigger servo motors with MUCH higher running torque but the price goes up too. A 34 frame DC servo can have 300 to 400 oz-in of continuous torque. Why continuous torque versus the peak torque everyone likes to use for ratings? because if you exceed the continuous torque rating for several minutes you start to get over heating. Peak should only be drawn for short bursts.
You are sorta in between, you really need the speed of servos but not the full speed . So if you belt reduced by a ratio of 2 : 1 between the motor and ballscrew you double the torque to give you more "push" and still have speeds up to 400 IPM at that torque. Speed you cannot or do not use is wasted. You also have a limit of how fast you can spin the ballscrews and it goes down the longer the ballscrew is , the smaller in diameter , that they type of mounts on both ends. I have not seen any over 3 ft long that can handle 4000 RPM. The belt reduction trades speed you can't use for torque you can. I have an old table I built as a combo plasma and router with small servos and 5/8" dia ballscrew a lttle over 5 ft long/ To pRPM I can get before the vibration starts to cause binding on the ballnuts si about 180 IPM. They are 5 TPI ballscrews. Made a pretty good router table but was a dud for plasma cutting. torque/acceleration was never a problem but upper speeds were.
While I agree that servos WILL solve your speed problems, just swapping them out directly connected will cause other problems unless you oversize the servos. You can get ones that will twist the cape off Superman but they carry a pretty hefty price tag.
Re: Speed of steppers. Anyone know about this?
Well, all good points.
I think that if you want more power (power = speed x torque) for a frame of that size, trying to stay with NEMA23 size may be a little bit unrealistic.
And as you go up in size, so does the $$. Of course, if all your motor mounts are 23, then you have a problem there as well.
Eh - if you want to go up that far, you could go to linear motors! OK, OK, not really helpful here!
It's all a trade-off.
Cheers
Roger
