Thread: New Quill Drive Design

Hi Ray,

I think that the 3-3/8 quill distorting is a long shot for sure. The most likely cause of the tiny backlash you are seeing is the ballscrew support bearings. NcCams wrote in serveral threads detailing the issues with ballscrew support bearings here and it is good reading. True ballscrew support bearings cost several hundred a pair but will reduce backlash caused by shift in the ballscrew to as close to zero as can be measured. Plain standard angular contact bearings, non preloaded, with deflect very slightly as the load increases and cause tiny amounts of backlash (2tenths to a thou).

I am betting you tried it, but if not, try pulling the quill down by hand. It amazed me just how much force it takes! All that force is being absorbed by the screw bearings. A half thou deflection wouldn't be bad with any bearings costing under $200 a pair.

I just lucked out and found a couple pairs of those 20tac47 special purpose ballscrew bearings for my mill (usually about $800 a pair!!!). They are a special breed indeed. They are ground for 500 lbs preload to minimize deflection. Ballscrews I have bought on ebay have also had those dedicated ballscrew bearings on them which is a real plus.

The high quality matched pair bearings have a chevron marked on each outer race, those are lined up during install for minimum runout.

In your case, the high buck, ballscrew bearings are not going to be cost effective. Hopefully you already have angular contact bearings at least. A half thou isn't bad backlash!!! My recent bridgeport V2E3 retrofit I did has between 1 and 2 tenths backlash on all axis but it has the good bearings and the factory ballscrews are NSK C3 grade ground units.

Your quill drive is looking fantastic! Very nice work!

Originally Posted by HimyKabibble

Walt,

How does your drive attach to the quill? Thats the source of the little bit of flex I have in mine, and it appears to be due to the quill itself distorting. Certainly the yoke is not flexing, and there is no backlash in any of the mechanical bits.

Regards,
Ray L.

Originally Posted by Sparky_NY

Hi Ray,

I think that the 3-3/8 quill distorting is a long shot for sure. The most likely cause of the tiny backlash you are seeing is the ballscrew support bearings. NcCams wrote in serveral threads detailing the issues with ballscrew support bearings here and it is good reading. True ballscrew support bearings cost several hundred a pair but will reduce backlash caused by shift in the ballscrew to as close to zero as can be measured. Plain standard angular contact bearings, non preloaded, with deflect very slightly as the load increases and cause tiny amounts of backlash (2tenths to a thou).

I am betting you tried it, but if not, try pulling the quill down by hand. It amazed me just how much force it takes! All that force is being absorbed by the screw bearings. A half thou deflection wouldn't be bad with any bearings costing under $200 a pair.

I just lucked out and found a couple pairs of those 20tac47 special purpose ballscrew bearings for my mill (usually about $800 a pair!!!). They are a special breed indeed. They are ground for 500 lbs preload to minimize deflection. Ballscrews I have bought on ebay have also had those dedicated ballscrew bearings on them which is a real plus.

The high quality matched pair bearings have a chevron marked on each outer race, those are lined up during install for minimum runout.

In your case, the high buck, ballscrew bearings are not going to be cost effective. Hopefully you already have angular contact bearings at least. A half thou isn't bad backlash!!! My recent bridgeport V2E3 retrofit I did has between 1 and 2 tenths backlash on all axis but it has the good bearings and the factory ballscrews are NSK C3 grade ground units.

Your quill drive is looking fantastic! Very nice work!

Nope, the bearings seem quite tight, and I measure no movement on the ballscrew. But if I run the quill into the stops, I can *see* the yoke angle slightly, so there is definitely movement at the interface between the yoke and the quill. The yoke is in hard contact with the quill both above and below the bolt, so I don't see another explanation. Still, overall it works as well as I had any right to expect, and actually better than I'd hoped, so I'm quite pleased.

Regards,
Ray L.

Nice job Ray!!

Hi Ray -

I just saw this thread and your quill drive looks great!

Any updates?

Mike

Originally Posted by mikegann

Hi Ray -

I just saw this thread and your quill drive looks great!

Any updates?

Mike

Mike,


I wish! I have been buried, for what seems like months, in Mach3 bugs, and SmoothStepper bugs. It has been incredibly difficult to get anything done!

The quill drive itself is working flawlessly. I was able to do a batch of parts last Sunday that came out perfect - depths were right on. It is soooooooo much nicer than using the knee for the Z axis, and peck drilling is now a pleasure, rather than a chore. But, using the knee and quill pretty much *requires* using tool length offsets, and the tool table, and that's where I've been running into one problem after another. And "problem", in this case, almost always seems to translate into another broken tool. Right now, I'm having probing problems when I probe for tool offsets, to load the tool table. Sometimes Mach3/SS just decides to ignore the probe input, and try to drive the tool right through the table. It's broken several (brand new, expensive) tools, and three touchplates. When it works, it works great. When it doesn't it is frustrating as hell. I'm working with Greg and Brian to get the problems sorted out, so hopefully they'll be sorted soon.

Regards,
Ray L.

Do you plan to sell kits or just for fun? Your quill systems is good.

Originally Posted by asuratman

Do you plan to sell kits or just for fun? Your quill systems is good.

The original plan was to sell them. There may now be a monkey wrench in the works, as, after 19 months of unemployment, I just accepted a new job. So, my time will be considerably more limited after Monday. However, I'm sure I will at least do a limited run of them. Now that it's all fixtured and programmed, it's not too difficult to machine the parts. Well, it won't be once Mach3 settles down and starts cooperating again....

Regards,
Ray L.

Originally Posted by BobWarfield

Ray, is there anything you can share in terms of fixing your Mach3/Smoothstepper problems? Are they chronicled in another thread?

I collect solutions to such problems in case I eventually need them!

Cheers,

BW

Bob,

Many of the problems I've had are unique to my configuration. For instance, the CV problems were a result of a) incomplete CV implementation in Mach3 (still not fixed....), that b) would not affect most people unless their Z axis performance was considerably less than their X/Y performance. One of my SS problems was due to a change in the latest plugin that I didn't notice, and, surprisingly, didn't occur to Greg either. The Spindle PWM, in plug-in versions prior to ogx, was ramping 8X faster than it should have. When I switched to ogx, suddenly my spindle was taking 30 seconds to ramp up to speed. Re-tuning cured that, once Hood cued me in to the change in the plug-in. I had a flaky SS - it was temperature sensitive, and would do all kinds of bizarre, irreproducible things. It's since been replaced. At the moment, I'm having several probing problems with the ogx2 plug-in, yet to be fully resolved. Sometimes it probes at a FAR slower rate than commanded (like 0.0001"/second or less). Sometimes it will completely ignore the PROBE input, and try to drive the probe, or tool, through the table. I have an impressive pile of broken probes, tools, and touchplates from this one.... I'm also having a problem with G-code programs simply stopping at completely random times - no error message, no indication of any problem whatsoever, it just freezes mid-program. Hit CycleStart, and it continues on as if nothing happened.

Regards,
Ray L.

Ray,
did you ever look in the SS Monitor when you got the random stops to see if there were any clues there?
Hood

Originally Posted by Hood

Ray,
did you ever look in the SS Monitor when you got the random stops to see if there were any clues there?
Hood

Hood,

Due to other problems, I haven't been able to run long enough to encounter that problem again.....

Regards,
Ray L.

Ray,

First off congrats on the new job!

There was this reply on another thread regarding probing issues. He was having problems when using keygrabber. http://www.cnczone.com/forums/showpo...9&postcount=13

I have also had some smoothstepper flakiness. My problems would happen when taking a heavy cut. I would get some error message, mach would essentially lock up (though my spindle would keep turning grrr) and I would have to unplug an replug SS to get it working again. I changed some noise settings and that seemed to solve it.

bob

Originally Posted by rowbare

Ray,

First off congrats on the new job!

There was this reply on another thread regarding probing issues. He was having problems when using keygrabber. http://www.cnczone.com/forums/showpo...9&postcount=13

I have also had some smoothstepper flakiness. My problems would happen when taking a heavy cut. I would get some error message, mach would essentially lock up (though my spindle would keep turning grrr) and I would have to unplug an replug SS to get it working again. I changed some noise settings and that seemed to solve it.

bob

Bob,

The problems I'm having now are hard, repeatable bugs, so they will get fixed. Fortunately, the remaining ones are mostly benign, or can be tested for before they do anything destructive, so I can live with them. Greg is hoping to have the new SS plug-in ready shortly, and I'm sure it will be a major improvement in several respects.

Regards,
Ray L.

Originally Posted by BobWarfield

Ray, are you expecting your probing issues to be fixed with the next Smoothstepper plug-in?

Best,

BW

Bob,

I hope so. It seems to be down to a rather benign problem now - the very first probe operation runs at a really slow feed rate, but all subsequent ones are fine. Greg was going to take a look at it shortly.

Regards,
Ray L.

Ray,

Is there a chance that you might share your conversion to Mach3? I didn't see any mention of that anywhere.

Thanks,

Ernie

Just re-reading this thread since Ernie popped in, the thread has been dead since Feb.

Bob, I don't know if the issues with probing and SS were solved in the SS plugin, but Dave Bagby and I had an issue with probing in the MSM screenset at the very beginning of the MSM beta and Brian fixed in in Mach3. Working for me now, at least in MSM with Dave's probing routines, not sure about what's happening with Ray's probing routines.

Ernie, Ray isn't around as much as he used to be, but I'm sure he'll share his Mach mods with you. Perhaps email him directly in case he doesn't see this. I have a quill drive and my knee (A axis) is being used for tool length comp. sort of like Ray's, but I couldn't get Ray's macros for probing to work so I did it myself, but manually. It works, and Ray's M843 and M849 macros DO work for using the knee for tool length offsetting.