View Full Version : Need to figure out lead screws before I begin
Darren 08-25-2005, 12:11 PM Ok, I've been trying to catch up since I put my project on hold for the last year. I have collected most everything I need to begin my build except for the lead screws. I'm using 360oz/in servo's. From what I've read it appears I'll need to do a reduction so I'm assuming I'll need to drive the screws with a belt driven gear reduction.
I'm trying to build this cheaply so I'm considering acme screws. My only concern is machining. Well, that and how to mount them properly. I have no way to do it and really would like to buy the screws machined to my needs. Is there a place that can do that? I really need to figure this part out before I begin building so I can have them in my possession as I design. I'm a bit overwhelmed as I've been reading the forum for the last few days and am a little frustrated.
You learn more about my planned build in this thread here:
http://www.cnczone.com/forums/showthread.php?t=12696
Any suggestions or links to helpful threads would be much appreciated! Thanks.
Darren
Evodyne 08-25-2005, 12:28 PM Darren,
Hi again. You might take a look at the HomeShopCNC (http://www.homeshopcnc.com/index.html) web site. He sells rolled ball screws which are a step up from acme. And he cuts to length and machines the ends for you. Don't know if he can/does provide support bearings and mounts, but you can ask him. I also see he now has complete "dual v-slide assemblies". They look a little more expensive, but they are self contained.
I've bought similar self contained "motion stages" off of eBay-track, linear slides, ball screw, etc. in one unit. Again, the advantage is they are ready to go. Sometimes they come cheap, sometimes they don't
Lance
Darren 08-25-2005, 12:33 PM Ouch, can't afford the V-slide assemblies. I'm interested in the ball screws though but they appear to be fairly spendy as well. I'm really limited on budget so I gotta stay on the cheap. Every corner I turn is another $500 and I can't continue like that unfortunately. I didn't see different prices for different sized diameter ball screws though. I was thinking 1/2" would be good for 48" length but that is just a guess. I'll send HomeShopCNC an email.
rustyolddo 08-25-2005, 06:39 PM Look at Roton, you can get 5/8 ball screw 4 feet for $30 and the ball nuts with fitted balls will run you $20 a piece. Outstanding customer service too.
ger21 08-25-2005, 08:48 PM http://www.nookind.com
They'll machine the ends, but it may not be as cheap as you'd like. You'll also have to fabricate yout own mounts, as their's will cost far more than you could imagine. :)
Darren 08-25-2005, 09:34 PM Thanks guys, I'll look at those options as well!
chuckknigh 08-25-2005, 11:13 PM Maybe this is a silly question, but why do you need the ends machined? Just buy a ball bearing with the ID the same as your screws. Then, use a washer and a nut on the end, to secure it!
There are couplers that can be used to mate the ends of your screws and the ends of your motors, as well, with no machining required.
ACME thread has all the same types of nuts as regular triangular thread...get a lock nut and attach it that way.
It's not rocket science...and it certainly doesn't require precision machining just to mount a screw. Simple solutions...
-- Chuck Knight
ViperTX 08-25-2005, 11:25 PM Chuck, Well on an ACME that might work...but pre-loading the bearing may be a challenge...have you done it that way?
rustyolddo 08-26-2005, 06:11 AM One thing to note on ball screws is that the outer surface is hardened, the inner is reasonably machinable. You could drill & tap the end of the ball screw for a small bolt, 8mm or 1/4-20 use a large washer to secure it in the bearings. If the screw was shorther than the span between the bearings it would allow you adjsut the pre-load. Or You could simply shim the bearings.
Darren 08-26-2005, 08:58 AM Chuck,
I like that solution but wonder about pre-loading. Can it be done that way?
rustyolddo 08-26-2005, 09:28 AM Chuck,
I like that solution but wonder about pre-loading. Can it be done that way?
Sure. It takes very little to preload a bearing. Too much can cause it to fail prematurely.
What I'd probably do is order some bearing springs, they're wavy and designed for the purpose of establishing preload. You can get them from Reid Tool & McMaster Carr. They are sized to the industry bearing numbers.
Machine the pockets for your bearings to the spring compression spec, put the spring in, the bearing then the bolt & washer for the ball screw and tighten it all up. You're good to go then.
The only issue that you have to worry about is extreme force on the ball screw. If you were to push really hard as in a bind, it would be possible to compress the spring a little and introduce some error. You can solve that simply by adding a set screw to the bearing pocket. Assemble everything, allow the springs to tension the bearings, then gently lock the bearing in place with the setscrew, you could get fancy with the mount if you wanted. Chances are that your stepper motors would stall before you'd compress the bearing springs but the setscrew would eliminate the possibility. I wouldn't use the bearing springs in a Bridgeport mill conversion but in a router or plasma table, they are the way to go.
If you are tool poor, you could even skip drilling the ball screw and just buy some 5/8" locking collars with the double setscrews, they're less than $1 each. Assemble everything and compress the bearing springs and tighten the locking collars onto the ballscrew.
Darren 08-26-2005, 09:34 AM Thanks rustyolddo,
This is sounding more reasonable cost wise. It sounds like I could do this with ACME or ball screws just as easily so I might start with ACME first to keep cost low, then upgrade later down the road if I feel it is necessary.
What do you think about this idea for pre-loading using the idea of bearings over the threads...
Fix one end of an ACME thread rod with a bearing sandwitched with appropriate nuts/washers and trap the bearing in it's machined mount on one end of the axis. On the other end of the axis machine a seat for the bearing so that the rod can pass through the mount, then on the "outside" you slide the bearing on and into the seat, follow that with a nut and torque the nut appropriately... this would create pre-tension by pulling on the ACME thread rod but would it be a satisfactory method?
I've got plenty of tools and woodworking machinery, I'm an avid woodworker and most of those tools will work aluminum "ok". I have plenty of carbide :)
I'll have to draw up some plans and post them to be sure I'm getting things right.
rustyolddo 08-26-2005, 10:37 AM Thanks rustyolddo,
This is sounding more reasonable cost wise. It sounds like I could do this with ACME or ball screws just as easily so I might start with ACME first to keep cost low, then upgrade later down the road if I feel it is necessary.
What do you think about this idea for pre-loading using the idea of bearings over the threads...
Fix one end of an ACME thread rod with a bearing sandwitched with appropriate nuts/washers and trap the bearing in it's machined mount on one end of the axis. On the other end of the axis machine a seat for the bearing so that the rod can pass through the mount, then on the "outside" you slide the bearing on and into the seat, follow that with a nut and torque the nut appropriately... this would create pre-tension by pulling on the ACME thread rod but would it be a satisfactory method?
I've got plenty of tools and woodworking machinery, I'm an avid woodworker and most of those tools will work aluminum "ok". I have plenty of carbide :)
I'll have to draw up some plans and post them to be sure I'm getting things right.
Sure that will work, you'll just have to be a little careful torquing the bolt. I'd probably use a dial indicator on the end of the shaft and adjust the bolt until all the backlash is removed and lock it down there.
Using Bearing springs eliminate the need the need for critcal adjustment. On a CNC router, I'd probably go that route.
Darren 08-26-2005, 10:40 AM Ok, one last question on this. If I was to go with the Roton ball screws how would I find nuts appropriate for use on them? I'd get my ballnut of course, but I can only assume ballscrews don't accept nuts of any other sort... right? If so this would eliminate my cheap mounting. Am I correct on this?
rustyolddo 08-26-2005, 11:57 AM Ok, one last question on this. If I was to go with the Roton ball screws how would I find nuts appropriate for use on them? I'd get my ballnut of course, but I can only assume ballscrews don't accept nuts of any other sort... right? If so this would eliminate my cheap mounting. Am I correct on this?
Roton sells the ball nuts. You can use two per piece of ballscrew and make your own anti backlash ballnut or if money is tight, you can ask to have the ballnuts fitted to the ballscrew (larger balls are installed), and use just one ball nut. The ballnuts have a threaded end, 15/16x 20 or something like that, you can buy a tap & make your own mounting or your can buy premade mounting flanges or you can mill a groove in the carrage and clamp the side of the ballnut. Lots of ways to do the mounting. Ballscrew is unique, nothing else will thread on the ballscrew. They aren't any more accurate than ACME, in fact they are a little less accurate but they have very low rolling resistance.
Acme is cheap, especially from Enco, you can use two brass Acme nuts to make a zero backlash nut. Down side is an increase in rolling resistance. You've got good motors so it shouldn't be much of an issue. Acme can be a little flimsier than ballscrew because ballscrew is a hardened alloy. Figure 5/8' in Acme minimum on long lengths.
Darren 08-26-2005, 12:13 PM Ok, I think I'll give the ACME route a try. Cheap, effective, and easy to upgrade later if I feel the need. Now, the task of determining pitch. Info on my servo's and plan can be seen here: http://www.cnczone.com/forums/showthread.php?t=12696. I'm just wanting respectable speed and torque. Not looking to break any records here :) I'll be cutting some aluminum but mostly MDF and hardwood. I'm considering a 24V 10A power supply but you can see more discussion on that in the thread link... I haven't decided totally yet. What TPI would you recommend? There seem to be so many variables I'm having a hard time getting one thing nailed down. Seems the TPI or pitch chosen would be affected by the power supply chosen and the gear reduction ratio chosen etc... I'm needing to figure that out before I get too much further. Here is what I'm thinking though, tell me if I'm on the right track or what you might do in my place with what I have.
• 360oz/in servos (have these see specs in above linked thread)
• 24V, 10A power supply (considering)
• 2:1 reduction (not sure about this)
• Considering Gecko's G320 drivers
• Have no idea about a break-out board
• Lead screws, seem to remember thinking 10 TPI was correct (I may be way off here, same for all axis'?)
Thanks for all the help and any future help guys! If I can get this stuff figured out on paper I'll feel a lot more comfortable ordering the lead screws so I can begin construction.
ViperTX 08-26-2005, 01:16 PM Darren,
Check out buildyouridea.com in the IG88...shows you how he handled the ACME for his system.
Darren 08-26-2005, 01:34 PM ViperTX,
Thanks for that, good visual aid for mounting the ACME screw.
Darren 08-29-2005, 01:36 PM Can anyone help me figure out the ACME TPI I should use. I'm so close to getting everything all lined up.
Below is a suggestion for the servo's I have but I really don't know quite how to judge whether it will be the best route. I'm considering using a 24V 10A power supply instead of the 36V suggested and am not sure how to re-calculate.
Use a 36VDC power supply. <---(I'm considering a 24V 10A power supply though)
No-load RPM at 36VDC = 3,130
(Ke = 11.5, KRPM = 36V / Ke)
80% of no-load RPM = 2,500 RPM
Leadscrew RPM @ 120 IPM, 5TPI = 600RPM
Reduction Ratio = 2,500 RPM / 600 RPM = 4.1666:1, Use 4:1
Kt = 1.351 Ke = 15.5 in-oz / amp
Maximum continuous current = 3 Amps
Maximum continuous torque = 3A * Kt = 46.5 in-oz
Torque on leadscrew = 46.5 in-oz * 4:1 reduction = 186 in-oz
Leadscrew thrust = pi/8 * TPI * in-oz = 365 lbs
Maximum speed at full load = 2500RPM /(ratio * TPI) = 125 IPM
Maximum speed, no-load = 3,130RPM / (ratio * TPI) = 156 IPM
Power delivered to load = RPM * in-oz / 1351 = 86 Watts
ViperTX 08-29-2005, 02:33 PM Stick with the 36 VDC supply, you'll never achieve the continuous torque otherwise.
ger21 08-29-2005, 03:06 PM I thought with servos, the voltage determines the rpm, but the torque should remain constant. So with 24V, you'll only get 66% of the rpm, but the torque should stay the same. Or am I wrong on this?
ViperTX 08-29-2005, 03:50 PM Actually the torque curves are parallel to each other....so, you'll get the same rpm, but the torque will be much lower.
ger21 08-29-2005, 06:25 PM Actually the torque curves are parallel to each other....so, you'll get the same rpm, but the torque will be much lower.
I want a second opinion. :) I think you're wrong, rpm is directly related to voltage. Not sure about torque, though.
fyffe555 08-29-2005, 06:42 PM gerry's right,
rpm is directly related to voltage, torque is directly related to Current, so more voltage means faster speed. More Amps gives higher torque.....
Andew
ViperTX 08-29-2005, 10:41 PM Here's the torque curve for a Baldor DC servo.
fyffe555 08-29-2005, 11:18 PM That's a motor speed-torque curve for a dc motor, and yes I think that says the torque increases with current.
The chart is intended to determine the current and voltage for a motor, the curves show the torque produced at a given speed at a given Current and voltage - the dark area is the continuous running range, the light area the burst or intermittant running range. Out side the 30 or 50v line is more than the motor can produce at that voltage.
So from the baldor chart if you need 2lbs-in at 2000rpm continuous then you need 3.5amps at EITHER the 30v or 50v bus. If you need 5000rpm then only the 50vbus will do. If you need 1000rpm and 10lbs-in intermittant then either the 30v or 50 v will do but you need 18 amps... in all cases the torque is directly related to the amps which is why the scales are together....
fyffe555 08-29-2005, 11:36 PM I meant to reply to darren and forgot, your tpi for the screw should come from your requirements for force and speed. The combination of your motor to screw gearing plus your leadscrew tpi will determine the distance moved and force produced at a given servo speed. In your example of 4:1 and 5tpi that's ~ 120ipm and 365lbs thrust. If you use say a 10tpi or and 8:1 then you will get 60ipm and 730lbs thrust.. If you want more speed less thrust then lower the tpi or reduction ratio...
Note that assumes you still plan on using the full speed range of the motor ~2500rpm which means you should use a 36v supply. If you use a 24v supply then your no load motor rpm will drop to ~2000rpm and working speed of ~1600 so you will need to calculate the reduction, tpi, ipm and thrust accordingly..
Andrew
Cutloose 01-27-2006, 06:36 PM Hey....how did you make out on the Acme screw configuration? I've been following your posts and was considering doing something similar. Did you end up machining the ends..?
Cutloose
Darren 01-27-2006, 07:36 PM Unfortunately I still haven't got my lead screws, life got in the way and I had to put my machine on the back burner for a little while. I just haven't had the time to mess around with it. :(
|
|