[LazyMan]

Just completed my Grizzly "X3" conversion with cnc Fusion deluxe ball screws and the helical couplers he sells and I am getting .005" backlash in X and y. With the gibs completely loose I get .0015" back lash and with the gibs slightly tightens down I get .005. I checked everything and made sure all screws are tightened. It seems like the springiness of the couplers is causing the problem but I am really not sure.This is pretty disappointing to say the least. Any ideas?

[AiR_GuNNeR]

I'd say your assumption is correct. I'd replace them with the Oldham style couplers. My X2 conversion for CNC fusion used Lovejoy coupling with rubber spiders. I had around .005" backlash as well. Luckily Lovejoy has the .dxf files of their spiders online so I downloaded them and CNC'd out spiders made of an old cutting board. My backlash went down to .002".
The Oldhams are easy to make. The insert is cheap enough it's probably not worth making that part however.

[WOTDesigns]

I have identical results (well- .0043 on x and .0031 on Y)

it was causing error on long runs when I had the backlash adjusted to the thou.... then I started auditing then by moving positive to zero, more positive part zero, make 15-20 moves then go back to zero. adjust accordingly.

I can run a 90 minute program with 6 tool changes and zero back on my origin within a few tenths of zero.

it really is frustrating that tightening the gibs INCREASES backlash. I've torn everything apart twice and can't get it any better, but it will have to do.

not as annoying as the small head nod I get every z+ move. lovely little circles in an otherwise beautiful finish :-/

[Winnfield]

I don't think the couplers are your problem. They should be on the opposite side of the bearing block, and the bearings preloaded before it. There shouldn't be any axial movement of the screw at the coupler.

Naturally, the tighter the gibs are, the more backlash you will see if any of the sloppiness is hidden in bearings that seem tight and that sort of thing. When it can move freely, without resistance, the friction in the problem areas keep them from shifting, but when you tighten the gibs to the point of overcoming that friction, you will see the movement if it's there.

That may or may not be your problem, but I would make sure all of the bearings have clean mating surfaces, etc. and increase the preload to see if that changes it at all.

[WOTDesigns]

this is my first experience with ball screws and my knowledge is limited to what I learned from web research (which just isn't as good as hands on). I am not sure HOW to correctly increase the preload I think. I tried tightening the nut on the outside of the bearing on the X+ end but couldn't. even with the mortis holding it would just turn the screw.

what's the"right" way? is it like adjusting the gibs "just by feel"? I think I would have to clamp the un threaded section with vice grip and shop rag to even move it.

[cornbinder23]

Have you tried lapping the ways? It allowed me to run the gibs a little tighter, I did it after I machined the ways for oil passages, made a huge difference
I have that same kit and I can only measure .001-.002 of BL on all axis
Spend alot of time refitting the Z gib and your nod problems will be minimized drastically.:cheers:

JTCUSTOMS

[WOTDesigns]

I hand lapped and straightened the z but I an sure it would be best to start from scratch

I also used the aluminum polish lapping method for HOURS just to get to 60% smooth. I'm sure it improved it from where it was. I'm thinking back and don't remember breaking the ball nut loose when I took everything back apart last time. maybe loosening and tightening it with the screw re adjusted will help ::shrugs::

[AiR_GuNNeR]

I don't think the couplers are your problem. They should be on the opposite side of the bearing block, and the bearings preloaded before it. There shouldn't be any axial movement of the screw at the coupler.

Naturally, the tighter the gibs are, the more backlash you will see if any of the sloppiness is hidden in bearings that seem tight and that sort of thing. When it can move freely, without resistance, the friction in the problem areas keep them from shifting, but when you tighten the gibs to the point of overcoming that friction, you will see the movement if it's there.

That may or may not be your problem, but I would make sure all of the bearings have clean mating surfaces, etc. and increase the preload to see if that changes it at all.

The problem with the Lovejoy couplers is that they allow the stepper to rotate, compressing the spider, but the screw isn't rotating yet. So I guess it's not backlash in the traditional sense, but more like virtual backlash...

[Winnfield]

In the original post, he said he is using helical couplers, so I think he can probably rule out that as an issue.

[michaelthomas]

The helical couplers that I originally used on my router were complete junk. You could see them winding and unwinding slightly with every reverse of the axis'. I replaced them with an oldham style coupler and was very happy with the results.

I suppose there are good and bad helical couplers......just like anything. Mine were turds........well, they could have been turds if they were a little better.

[AiR_GuNNeR]

In the original post, he said he is using helical couplers, so I think he can probably rule out that as an issue.

Helicals may actually be worse. If you think about it, they are nothing more than a spring. If you have large steel helicals, it may not be an issue, but a 1" aluminum one would be springy.

[Winnfield]

I see what you mean - virtual backlash sort of thing. That makes sense. I guess I have never seen any that do that, but can certainly see where it might be an issue if they aren't robust enough.

[LazyMan]

I see what you mean - virtual backlash sort of thing. That makes sense. I guess I have never seen any that do that, but can certainly see where it might be an issue if they aren't robust enough.

Yes, virtual backlash or effective backlash induced by flexing couplers is what I was referring to. When I have time, I am going to try wrapping the couplers with some tape and see if it improves anything. I also get the Z "head bob" and backlash in that axis too. I really did not want to fiddle around with too much stuff when I bought this conversion as I just want to use it to complete a small project and gain some cnc milling experience. I guess I should have done more research. Already thinking of selling it but who the heck is going to want to by a cnc mill with .004" backlash. My cnc router has less backlash than this! When I have time to tinker with stuff I will let you guys know the results.

[michaelthomas]

Buy yourself one Oldham coupler. Mcmaster Carr sells them and they are very cheap. I would bet you can cut your problem in half without much effort.....at least.

The nice thing about mechanical problems, is that solutions can be arrived at in logical manners......when you find an offending part it can be immediately discarded and replaced without worry of damaging the pride of the unit in question......unlike marriages, unfortunately.

There is a good deal of satisfaction that comes with tracking down an issue and making it right......don't deprive yourself of that!

The suggestion of lapping the ways is a very good one, too. Less pressure required to move the table will cause less distortion elsewhere.

[LazyMan]

I am in the process of lapping the ways and noticed that with the design of the CNC fusion X axis will not allow me to use an oldham coupler because there is no support bearing for the ball screw on the stepper end. Has anyone ever used solid couplers?

[cornbinder23]

I have never used any solid couplings before, but I have seen the aftermath of a misaligned stepper with a solid coupler.
Your best bet is going to be S.S. helicals instead of the anodized aluminum.
One thing you could try is putting a couple of hose clamps around the coupler to keep it from "expanding/unwinding" just to see how much that can be controlled as well


JTCUSTOMS

[LazyMan]

I am using the couplings that come from CNC Fusion. I believe they are steel and never heard anyone say anything bad about them. I was under the impression that they were pretty good.

[cornbinder23]

I have the same kit and IIRC mine are anodized balck aluminum not steel
You might check with a magnet

JTCUSTOMS

[LazyMan]

Yes you are correct, they are anodized aluminum. I ordered an oldham coupler to try on the Y axis to see if there is any improvement. If so, then I will get some more solid helicals for the X

[WOTDesigns]

interested to hear your results with the oldham. I would love to not have such a violent backlash comp