Clearpath SDSK for mini mill?

[Scot808]

I'm on the fence here and am hoping for some advice. I'm very new to CNC and recently finished a Sieg mini lathe conversion using an inexpensive Longs Motor package purchased off ebay. It uses ~390oz/in steppers running on 36v. I haven't experimented with too many setting but, from what I gather, these motors aren't being used to their full potential due to the relatively low power supply voltage. However, for testing, I installed one of them on the mini mill and, while not particularly fast, it seems plenty strong. Very unscientific, but I put a fair amount of my body weight (guessing at least 50 lbs) on the spindle housing and the motor easily raised the z-axis.

Before I started my entry into CNC and while researching stuff, I had my set my mind on a Clearpath system for the mini mill (Sieg X2D with CNC Fusion conversion kit). While the Teknik parts aren't that much more expensive compared to a higher end (i.e. purchased from a reputable US based shop) stepper setup, they are hugely spendy compared to a ~$220 complete ebay stepper package! So, as mentioned at the beginning of this post, I'm now on the fence trying to decide if I want to spend about 5-6 times as much on a servo system. I don't care for the stepper noise and resonance and I like the technology and coolness of the Clearpath servos but what to do?

If I do opt to spend the money, my mind is made up, it will be Clearpath motors. Anyone with (or without) experience with the Clearpath stuff care to offer up any advice or motivation...for or against? Also, I am unsure on what size motor I would want to use. The Teknik rep I've been conversing with, while very helpful, can't really commit to a particular motor as I don't think he knows much about the mini mill. Sizing suggestions would also be extremely helpful.

Thanks!

Similar Threads:

• Need Help!- Super Mini Mill 2 vs Mini Mill Tool Tracking
• Mini tech Machinery 3-axis CNC Mini-Mill 2 CAD/CAM 3d Printing Manufacturing
• Mini tech Machinery Corporation 3-axis CNC Mini-Mill 2 CAD/CAM Manufacturing
• Mini tech Machinery Corporation 3-axis CNC Mini-Mill 2 CAD/CAM Manufacturing

[Jim Dawson]

I have used ClearPath motors for a router retrofit for a customer. I was happy with the result. ClearPath motors are designed as a direct replacement for open loop stepper motors, and close the loop internally. I would say at least an order of magnitude better than stepper motors. Easy to install and with the built in drives, takes a up lot less room in the electrical enclosure.

I'm not able to advise you of the size motors needed. But there are a number of writeups on Sieg X2D conversions, so if you use the same size motor as others have used in steppers, your performance will be better with the ClearPath motors in the same torque rating.

[Scot808]

Hi Jim,
Thanks for the reply. I believe the mini mill conversions usually use a stepper rated around 350-400 oz/in. I think this is the stall or holding torque rating. When you say to use the same size Clearpath motor, do you mean physical size or torque rating? I'm assuming torque rating. And comparing Clearpath peak torque rating to the stepper's holding torque?

[Jim Dawson]

Sorry, I should have been more clear. I would match the Clearpath peak torque to the stepper holding torque or maybe a bit more. Stepper torque drops off rapidly as the RPM increases where the ClearPath continuous torque is pretty flat, and still has plenty of peak torque to accelerate at reasonable rates.

[Scot808]

Thanks, Jim. Would you have any recommendation for a BOB and motion controller? I have an ESS and a cheap Chinese BOB for the mini lathe and that works fine. Wouldn't mind trying something else, if there's anything that's better and also easy to set up.

Has anyone run the Clearpath motors on a mini lathe? I've searched and have yet to run across one.

[Jim Dawson]

Happy to help. Sorry, I have no experience with the lower end BOBs so can't help you there. Most of the stuff I work on is more industrial.

[Scot808]

Hi again Jim,
Well, I went ahead and ordered the Clearpath motors...went with the ones that produce about 600 oz. of peak torque. Anyway, I went with the "regular" version with 800 steps/rev. I'm using 5 turn/inch ballscrews and the resolution seemed like it would be more than enough. Now, I'm second guessing myself. I've seen a couple of Youtube guys (one with a router and one with a mill) and they both are using the high resolution 6400 step/rev motors. Teknik has a liberal return policy and I'm wondering if I should take advantage of that and exchange my motors for the high resolution version. Steppers frequently are run at 16x microstepping and that is quite a bit more theoretical accuracy than 800 steps which equates to 4x microsteps. I do realize the torque loss and lack of closed loop control with steppers. Curious to hear your thoughts re. this.
Thanks, Scot

[Jim Dawson]

Well, 0.200 / 800 = 0.00025'' per step. Unless you plan on doing work for NASA, that's probably close enough. Or use a 2:1 timing belt reduction and get 0.000125 per step.

[SCzEngrgGroup]

And if you believe you're going to get anywhere close to +/-0.00025" accuracy from an X2, you've got a rude surprise coming... Getting even +/-0.0025 true accuracy from an X2 would require a large measure of dumb luck. Putting servos of any kind on an X2 is kinda ridiculous. Decent steppers will give you all the accuracy, resolution and performance the machine itself is capable of, at a much lower cost.

Regards,
Ray L.

[Scot808]

Thanks for the reply, guys. That's about what I figured but then there are those guys with the high res motors, I thought maybe I was overlooking something.

I hear what you're saying, Ray, but due to lack of time and space, I'll likely never have anything bigger/better than the X2. So, being curious, I figured I'd give them a try just to see. I might be foolish and uninformed but I'm not dumb...I'm well aware of the X2's shortcomings.

[ger21]

You probably could have gotten similar performance with AM882 drives, and a 60V toroidal power supply, for about $350.

The best breakout board for the ESS, is probably the MB2, from CNC Room.

[Scot808]

Okay, there has to be some reason why a higher count servo is advantageous. I'm using an Centroid Acorn board for this setup. Centroid seems like a pretty serious industrial type company and they like to recommend high end accessories and such. They recommend a minimum of 4000 count/rev for the motor encoder, saying 8000 count is better. See the first post here:
Encoder Basics - Centroid Community CNC Support Forum

Now I realize they're thinking this stuff is for a high end machine that needs serious reliability, etc. but still it seems unnecessary to specify that much resolution. So...I'm thinking there must be a reason. Come on guys...you guys are smart. Someone must know the reason and is able to explain it to someone like me who's a little slow.

[1Jumper10]

Okay, there has to be some reason why a higher count servo is advantageous. I'm using an Centroid Acorn board for this setup. Centroid seems like a pretty serious industrial type company and they like to recommend high end accessories and such. They recommend a minimum of 4000 count/rev for the motor encoder, saying 8000 count is better. See the first post here:
Encoder Basics - Centroid Community CNC Support Forum

Now I realize they're thinking this stuff is for a high end machine that needs serious reliability, etc. but still it seems unnecessary to specify that much resolution. So...I'm thinking there must be a reason. Come on guys...you guys are smart. Someone must know the reason and is able to explain it to someone like me who's a little slow.

The higher encoder count translates into smoother motion. The ClearPath servos, both the Regular and Enhanced models, use the same exact electronics for the encoder. The resolution difference is only due to firmware limitations in the Regular models. I cant remember what the absolute encoder resolution is, I think its somewhere close to 12500(?), but when the motor is following a command, it uses the full resolution of the encoder to get there. The difference is, in the regular model with 800 CPR encoders, it can only stop at 800 distinct points (.45 degrees) in a rotation. The enhanced model can stop at 6000 (.06 degrees) distinct points.

[SCzEngrgGroup]

If you were building a high-speed, high-precision machine, the extra resolution would be meaningful. On an X2, it's a complete waste. Larger machines are capable of FAR higher speeds (>1000IPM), FAR higher accelerations, and far, far, FAR higher precision, resolution, and accuracy (<0.0002") than an X2 could ever muster, no matter how much money you throw at it. Those big machines NEED higher count encoders to provide proper control at those very high speeds. An X2, on it's very best day, will only be capable of perhaps +/-0.003" average true accuracy, so all those extra counts will be completely lost in the noise of all the numerous other error sources.

Regards,
Ray L.

[Scot808]

Well, for the time being, let's forget this is for an X2. You never know what I may end up doing with it.

So Jumper says smoothness will be improved. That makes sense to me. I would think this would be noticeable more in slow movements where the steps might be a factor.

Now Ray is talking about fast and accurate machines. Okay, the accurate part I get but if you move fast won't lower steps not matter so much?

[SCzEngrgGroup]

For a slower, lower resolution machine, there is NO benefit to high encoder counts. The machine is physically incapable of responding to such small moves. You'll be able to position the motor more accurately, but not the machine. The machine WILL have significant backlash, friction and stiction, so it is inherently incapable of positioning to better than +/-0.003" or so. Attempts to move it 0.0001" will result in NO motion. Think of a D/A converter driving an LED. An 8-bit D/A will give visually perfectly smooth transitions from dim to bright, and back to dim. Will a 24-bit D/A make the LED *look* any smoother or better? (Hint: The answer is "NO", because your eyes simply cannot detect such small differences).

Regards,
Ray L.

[Scot808]

Thanks Ray, understood. But what if I one day modify the mini mill to use linear rails or something like that (very possible)...will it then be advantageous to have the high res motors?

I just spent a few minutes playing around with the motor on the mill. I can see very repeatable motion of 0.001" or so. Backlash on the ballscrews looks to be between 0.001-0.002". I didn't have time to test a single step move and see if it's noticeable or measureable. Of course, this is all with no cutting forces or anything like that so it's all best case.

[SCzEngrgGroup]

Thanks Ray, understood. But what if I one day modify the mini mill to use linear rails or something like that (very possible)...will it then be advantageous to have the high res motors?

I just spent a few minutes playing around with the motor on the mill. I can see very repeatable motion of 0.001" or so. Backlash on the ballscrews looks to be between 0.001-0.002". I didn't have time to test a single step move and see if it's noticeable or measureable. Of course, this is all with no cutting forces or anything like that so it's all best case.

You will never get true 0.001" repeatability in actual use. Thermal factors alone will take you out of that range, even before you get into the MANY mechanical error sources. The whole machine is simply too crude, too sloppy and too flexible, no matter how much money you throw at it. If you're trying to build a competitive race car, you don't start with a 1960's Morgan with a wooden frame. No matter how much engine and suspension work you put into it, it will still be a "Flexible Flyer" on the track, and will never handle anywhere near as well as even the worst modern car with a good, stiff, metal chassis.

Regards,
Ray L.

[Scot808]

Okay, so now that I'm thoroughly discouraged, please tell me what mill in a similar size would be a good choice. I'm assuming there is no such thing, so rephrase the question to what is the smallest mill that's worth considering.

[Kenny Duval]

Don't get discouraged. He is just trying to be helpful and keep you from spending money needlessly. On these hobby machines the better argument would be going with servos because they are quiet compared to steppers. However the minute you turn on the spindle and start it cutting the noise argument goes out the window as well. You won't be making NASA quality parts but you can still turn out some good work on the smaller machines.