[109jb]

Here is a link to a set with 4.7mH inductance motors. Better than what you had spec'd. These are way overkill, and would be better suited to a RF45 or larger mill, but is an example.

3Axis Nema34 CNC Stepper Motor 1230oz in 116mm 5 0A Driver 6A 80V 256 Microstep | eBay

these are a bit smaller and only 3.3mH inductance.

NEMA34 Stepper Motor ? 906 oz in 6.1A Single Shaft (KL34H295-43-8A) | Automation Technology Inc

The e-bay motors you have to dig down to find the inductance, but try to get as low as possible.

I personally have had good luck dealing with John from Automation Tech. His prices are pretty good too. Have not had any problems with the stuff I have received from him.

[martin_05]

Calling those motors cr@p would be an insult to cr@p. At 15 mH inductance, they will be absolute pigs speed-wise. They may have lots of torque at low speed, but that torque will disappear quickly as speed increases, and you'd likey end up having to tune for low speed, and low acceleration, to keep from losing steps. A good motor will have low inductance, certainly under 5mH, and ideally lower than that.

Even for $300 I wouldn't touch that setup.

Regards,
Ray L.

I should add that I already own a set of Oriental Motor Super Vexta drive amplifiers (2ea. UDK5114NA and 1ea. UDX5114N) and motors (PK566BUA):

Attachment 269712

Attachment 269714

However, the driver seems to only be good for 1.4A per phase and the motors only do just north of 100 oz-in. I was hoping to be able to use the drivers and just buy new motors but they don't seem to be adequate. My plan is to sell them on eBay to finance the new motors/drivers.

I am now investigating Dynomotion's card to couple to Mach 3:

Dynomotion Motion Control Boards for CNC Manufacturing and Robotics Applications


-Martin

[kenrc]

109jb you inbox is full, can you pm me?

[Alax7]

Nothing wrong with overkill steppers, that's a lot better than under kill for sure. If you want to put those motors to use, you can always build a 4th axis or an atc and use them there, or maybe you have a lathe you want to cnc as well.

[martin_05]

Nothing wrong with overkill steppers, that's a lot better than under kill for sure. If you want to put those motors to use, you can always build a 4th axis or an atc and use them there, or maybe you have a lathe you want to cnc as well.

That's a good perspective. Thanks.

My problem right now is that while I do understand the physics of stepping motors and can work with the equations, yet I don't have what I often call a sense of proportion about the application. Sometimes you only get that by making a few mistakes. I ran through the math today and I get that a 15mH motor will lose torque faster than a 5mH motor. What I don't have a sense of is what that might mean when running a half inch end-mill at a 0.1in depth of cut on a piece of 6061-T6 or mild steel. In other words, at the end of the day, does it matter? And, if so, what's the difference? Of course, the obvious answer is that there might be a reduction in attainable feed rate. I say "might" because attainable cutting rate is a function of, among other things, spindle RPM's. Since my RF30 is limited to just north of 2,000 RPM there's a practical limit to how quickly I can feed an end-mill into a piece. If a 15mH motor has enough torque at that feed rate to do the job then there's no reason to lose sleep over it and pay more for a 5mH motor. I just don't know. If someone does know I'd appreciate a shove in the right direction.

Today I was looking at a full DC servo setup for $1,100. Doesn't seem like a bad deal. But, you are right, I can always use steppers in another application.

There's also another potential reality here. I was looking at the idea of converting this machine and then selling it. The RF-45 with a square column seems like a really good idea. So, to some extent, if that's the plan, the question then becomes: What's the least I can do to convert this to CNC to then sell it in a few months. I am not talking about making a junk machine. After all, I do want to use it for a while to do projects. Just thinking about options.


Thanks,

-Martin

[pippin88]

Nothing wrong with overkill steppers, that's a lot better than under kill for sure. If you want to put those motors to use, you can always build a 4th axis or an atc and use them there, or maybe you have a lathe you want to cnc as well.

Actually, larger steppers often can not achieve the same speeds as smaller steppers. For most milling this may not be a problem, but it can be in HSM toolpaths.

[Alax7]

Touche, but I was referring to under sized vs over sized, like the 100oz-in steppers he currently has, not under sized vs the right size for fast speeds and no missed steps.

[rowbare]

This post by Mariss (from Geckodrives) does a great job of summarising stepper motor specs: http://www.cnczone.com/forums/gecko-...tml#post530370

bob

[martin_05]

Actually, larger steppers often can not achieve the same speeds as smaller steppers. For most milling this may not be a problem, but it can be in HSM toolpaths.

This is an interesting discussion and I thank everyone for helping me think this through. The important consideration here is to make a selection in the context of the machine. I think it is fair to say that this machine isn't going to be used for HSM. My guess is that an absolute top cutting speed of 25 IPM is reasonable for an RF-30 with a 3HP motor and a round column. Of course, this would further be limited by the material being cut. Rapids maybe at 100 IPM. Reasonable?

The entire moving table mass is probably about 100 lbs. My Kurt vice is probably another 55 lbs. I'll just guestimate that the entire payload --including vices or clamping-- will rarely exceed 100 lbs. Let's call it 150 to have some headroom. So, the Y motor has to move a maximum of 250 lbs and the X motor will see a little less, say, 225 lbs, due to not having to move the Y axis mechanics. Is that reasonable for this machine?

Now comes the question of acceleration. How fast do we get to 25 IPM and how fast do we get back down to 0 IPM? I'll throw 1 second on the table. In other words, 0 to 25 IPM in one second and 25 IPM back down to 0 in one second. Again, I ask, is this reasonable?

Other data needed to do the math are the static and dynamic coefficients of friction of the metal-oil-metal table-on-way structure and the efficiency of the leadscrew or ballscrew. I'll estimate the latter at 30% for the ACME screw and 80% for the ballscrew. No idea on the values that might be reasonable for coefficient of friction. Might have to measure them.

OK, here's where I reveal my ignorance. The other bit of data needed is cutting force. In other words, what is the force required to chew on various materials given a tool diameter, RPM, feed rate and depth of cut. No clue. I'm sure a calculator exists somewhere. All I am after is a reasonable maximum value. And this value can be machine limited. In other words, it is perfectly reasonable to say that on this machine you really can't go past a max value of X.

Armed with all of this data I think one could work it back to calculate motor parameters. Perhaps I am getting too analytical here. As a friend in aerospace likes to say: Sometimes you have to stop the engineers and just fly the damn thing.


-Martin

[109jb]

There are many cutting force calculators available. Here is my preferred:

Engineering Calculators

[martin_05]

There are many cutting force calculators available. Here is my preferred:

Engineering Calculators

Thanks. Well, it looks like I guestimated right. Based on nothing but feel and observation I guessed that a 1/2 inch tool at 2,000 RPM and a cutting depth of 0.125 on 6061 in would require a force of about 25 lbs. Unless I screwed-up the numbers input into the I'm pretty close.

I'd be interesting to play with numbers in the 25 to 100 lbs range for cutting force in the motor calculations.


-Martin

[martin_05]

I wanted to measure the static and kinetic friction coefficients for the table/saddle combo but decided to get lazy. In order to do this I would need to disassemble the mill again and weigh the components accurately. I might do that at a later time but it doesn't seem to be worth the trouble right now. For future reference, the formula is simple:

friction coefficient = Horizontal Force / Weight

To get static you simply have to record the max force before the table starts moving. Kinetic would use the force once it is moving.

I am going to use numbers I got off the web for steel over greased steel. The range being 0.03 to 0.19. Static is always a tiny bit more than kinetic but I'll use the same numbers.


-Martin

[martin_05]

I've been busy with work and haven't been able to work on the mill. I finally caught a break and took apart the spindle. Down side of starting with a used mill is that you really should take the time to clean and rebuild it. The good part is that you get to know every inch and bolt of the machine. I think I am going to take the time to model the important parts in SolidWorks just so the CNC conversion design phase goes smoother.

Attachment 270464

Taking the spindle apart wasn't too bad. WD-40, dead-blow mallet and very light application of a gear puller got the bearing races out. The bearings spin about a second if I spin them by hand (not installed in the spindle). They don't sound bad at all. However, no harm in replacing them with new SKF bearings.

The end of the spindle looks like it has seen a little channel-lock plier action. Surely because of the lack of a spindle lock. Some kind of a spindle lock/brake is definitely in the plans.

Attachment 270466

This made me think about the idea of filing a couple of flats on the very end to fit a 40 mm wrench. You could also drill a couple of diametrically opposed holes for a spanner wrench. Probably no harm in doing either. Any thoughts on that?

I do like the idea of a spindle brake better though. I envision something resembling a toggle clamp mounted such that it contacts the pulley (or something else) so the drawbar can be loosened with one hand. If I were to really engineer such a thing the way I envision it I would create a mechanical interlock of sorts with the spindle power switch for safety. I might just be a little too lazy to take that approach.

I also took the spindle pulley assembly apart and will be replacing those bearings.

Like I said, the downside of buying a used machine it that this work ought to be performed to bring it up to a reasonable standard.


-Martin

[martin_05]

I took apart the Z axis mechanism today. Gobs of ugly black grease in there. Lots of it dried-up and stuck to the gear teeth. I've had let parts soak in degreaser to make cleanup easier. Even with that some of the grease hardened-up to the point that I had to chisel it out with a small flatblade screwdriver.

I am starting to believe that breaking and deburring sharp edges must be a lost art in Taiwan and China.

Attachment 270614

Two more bearings to replace. They sound and feel right but, having gone this far, why not?

I wonder about Z axis backlash with this arrangement. There has to be a fair amount. It's almost like the spring that pulls-up on the spindle cartridge ought to pull in the opposite direction. Both drill bits and end-mills will pull themselves into the work. Just wondering.

It would be nice to get rid of all of this hardware and have the stepper/servo motor transmission drive the spindle up and down directly.


-Martin

[arizonavideo]

I didnt do it but I did see a 31 builds a long time ago that the guy reversed the return spring. This puts pressure on the rack all the time and gives no backlash.

[Bubba]

Martin,
To reduce the backlash in the Z axis, there are several things to do.
1. When you assemble the pinion gear to the rack, make sure you move the worm housing as far forward as you can. There is a bit of movement that you can use to take up the backlash in the pinion.
2. If you look at the back of the chrome bezel for the fine downfeed, you will see a shoulder that supposedly takes up any axial movement of the worm shaft and bearings. When I cnc'd mine, using the fine downfeed, I made a custom spacer to take up all axial movement of the worm and bearing assembly.

Dan Maunch went so far as to make an eccentric bearing for the return spring side of the pinion shaft to allow even more takeup of the shaft.

HTH

[martin_05]

Attachment 270684

Having one heck of a time removing these bearings. I had to cut a couple of notches on the spacer in order to provide clearance for a bearing puller to be able to remove the first bearing. Even with a bearing puller, cleaned and oiled shaft the bearing put up a fight. The second bearing had a retaining ring hidden behind the spacer. I don't call this good design. Well, at least this isn't the way I would have done it.

I don't have a bearing puller long enough to grab the second bearing. I tried blocking it and using a dead-blow mallet but it won't budge. I guess it's off to the store to find a longer bearing puller.

I am replacing the bearings so I don't care about any damage to the existing bearings while pulling them. However, I am concerned about installing the new ones. I'll polish the shaft to make sure there are no burrs or anything to get in the way. My guess is that I'll throw it in the freezer overnight and maybe cook the bearings to 100~150 degrees to potentially make things easier.

Any thoughts, techniques, criticism, advise would certainly be appreciated.


Thanks,

-Martin

[Fastest1]

I use a Makita cordless impact to loosen and tighten the drawbar. You dont need any tools but your hand to hold the spindle.

Do you have an open ended wrench that fits the shaft right below the bearing? A bench vise? Put the shaft in the jaw of the wrench set it on top of the vise with the jaws open enough to suspend the part, either heat the bearing or hit the shaft (with wood on top). Also make sure it doesnt have a retaining ring before doing any of the above.

[martin_05]

Well, I wanted to remove it without causing damage to the bearing but being that I was going to replace it... I just went to my local auto shop and they let me use one of their large bearing pullers to get it out. It put up a bit of a fight but it came out.

Tonight I started re-assembly. Had to use one of my large vices and a closed-end wrench to squeeze the bearing in. The wrench was sized to just make contact with the inner race of the new bearing.

Attachment 271106

Now I have to do the larger bearings for the spindle pulley shaft (the one on the left with the partially inserted bearing) . Again, the suckers are tight going in. What I really need is a hydraulic press and a "calibrated" pipe to push on. I smell a trip to Harbor Freight tomorrow. My bench vice doesn't open wide enough for this job.

Also pictured is the assembled spindle. I got preload down to where I only have half a thou of axial play. Hesitant to tighten further until I run-in the new bearings. And, frankly, on a mill of this class trying to get better numbers is probably utterly pointless.

I am guessing that run-in needs to be conducted in speed steps while keeping an eye on bearing temperature. I'll probably run it for an hour or so at various speeds, from the lowest my belts will allow to full speed.

I also completed cleanup, deburr and re-assembly of the Z axis mechanism. Here it is awaiting grease and installation.

Attachment 271110

Question about motor vibration. I ran the motor without any belts attached (left the pulley on it) and can feel vibration on the head. No, it's not rattling it's brains out but I have no sense of how smooth this motor is supposed to be. It is my guess that this vibration would definitely translate into the finish of a cut. Maybe a thou or two at the tip of an endmill with the quill fully extended. I guess I can put an indicator on the head tomorrow and measure it.

Just how smooth should this motor be? I am thinking of taking the pulley off to compare, just in case the pulley is out of balance.

I guess I am wondering if I should consider a new motor.


Thanks,

-Martin

[martin_05]

Spindle and pulleys with new bearings back in the machine.

I ran it for 2 hours at 250 RPM, then an hour at 1200 RPM and an hour at max RPM (2400?). It got a little warm at max speed but nothing even remotely of concern. I am going to measure runout and axial play tomorrow to see where things stand.

-Martin