Leadscrew backlash reduction - part 2

[Bob Engelhardt]

[Doozer - you can skip this, I know that you think there's "Nothing wrong with backlash. If you think there is, seek help."]

In Part 1, BCRider had an idea for an adjustable nut, which I interpreted to be something like this:

Threading the 2 gives an accurate way of adjusting the back lash and the adjust-nut is held securely (the 20 tpi is arbitrary):


I was all set to do this on my mill table, but when I removed it, this is what I had:




That's not gonna' work. So I came up with Plan B:


There are other ideas that use a spring between the adjust-nut and the main nut. I don't like that because if the spring is strong enough to not compress under load, it will be putting that force on the lead screw's threads continuously.

In Plan B the adjust-nut is held rigidly with respect to the main nut and the distance between the 2 is adjustable via the differential-thread screws. The 1/4-20 and 10-32 threads give an effective pitch of 0.019 (53 tpi), so that fine tuning the back lash is easy enough.


(I hate it that I happened to pick 2 screws of unequal length - very amateur.)

I have it adjusted so that there is a couple of thou back lash at the ends and 0.015+- in the middle. Yeah, there's still back lash, but just feels so much nicer when I can take it up by swinging the crank 54* instead of 112*.

[CalM]

I'm not sure why method one was rejected? I used that lash adjust scheme on my first Logan lathe with great success. It had a nut similar to the one you have shown.

[BCRider]

I'm not sure why method one was rejected? I used that lash adjust scheme on my first Logan lathe with great success. It had a nut similar to the one you have shown.

It looks like there would have been very little room between the major diameter of the Acme and the minor of the adjustment thread. That flat on the original nut puts a serious limit on the OD of the adjuster that could be used. It would be possible but it looks like it would be pretty tight and fussy.

Bob, your solution is another nice option. Just goes to prove the old saying "There's more than one road to Rome...". Nicely done.

Since some of the adjuster ideas tend to distort the contact between nut and screw I find that after an adjustment the contact points at the ends of the backlash tend to feel soft or spongy through the hand wheel. How did this method work out in that regard?

[J Tiers]

It seems that every single one of the schemes for adjusting out backlash by squeezing lengthwise will increase wear exactly the way Doozer says it will. He's dead right on that. They all reduce the length of thread that is carrying the axial load below what would be available if the nut were not split.

If you did something different, it would not, and it still would keep things tight.

Split lengthwise, as has been suggested before. Let the split feedscrew nut be loose so it can float radially for at least a small distance, but prevent it from turning. or moving lengthwise. Spring load it so that it stays closed on the screw, like a pair of halfnuts.

That should ride right over most variations in the wear of the screw, and still stay adjusted for least backlash. If you are concerned, or annoyed, by backlash, that seems like the best approach. It has the least objections to it, as I count them, and should work well.

The spring does not even need to be that strong, nothing like what a lengthwise spring would need to be, because it only needs to hold the nut halves against the radial component of thrust on an acme screw. There is some, but because the slope of the sides of an acme thread are pretty steep, that radial force is far less than it would be for, say, a standard screw thread.

With a reasonably long nut, it might even help in averaging the wear on the screw, although I'd not make a very strong claim for that.

[Doozer]

All kidding aside, I really like it.
Very creative. Good thinking.
Making it better, I endorse.
Thinking you need to make it
better or bad things happen
is a falsehood that I would try
to discourage. I think we are
both peeing in the same pond.

-D

[Bob Engelhardt]

I'm not sure why method one was rejected?
...

It looks like there would have been very little room between the major diameter of the Acme and the minor of the adjustment thread. That flat on the original nut puts a serious limit on the OD of the adjuster that could be used. It would be possible but it looks like it would be pretty tight and fussy.

Exactly:


And I didn't say, but the stud that holds the nut to the cross slide isn't held by a bolt - it's PEENED over! I didn't want to deal with that!

Bob, your solution is another nice option. Just goes to prove the old saying "There's more than one road to Rome...". Nicely done.

Since some of the adjuster ideas tend to distort the contact between nut and screw I find that after an adjustment the contact points at the ends of the backlash tend to feel soft or spongy through the hand wheel. How did this method work out in that regard?

Thanks! I haven't noticed that effect - I haven't worked with it much at the ends & wasn't looking for it. I will look for it.

[Bob Engelhardt]

It seems that every single one of the schemes for adjusting out backlash by squeezing lengthwise will increase wear exactly the way Doozer says it will. He's dead right on that. They all reduce the length of thread that is carrying the axial load below what would be available if the nut were not split.
...

I have not put any squeeze on the threads - there is still some back lash; meaning that when one flank is in contact, the other is not.

I did not split the nut. In fact I ADDED threads, so the average axial load will be LESS.

[Bob Engelhardt]

All kidding aside, I really like it.
Very creative. Good thinking.
Making it better, I endorse.
Thinking you need to make it
better or bad things happen
is a falsehood that I would try
to discourage. I think we are
both peeing in the same pond.

-D

Thanks. I appreciate your distinction between the doing and the needing.

[BCRider]

It's PEENED in place!

That must have made it all the more interesting to actually drill and tap for the differential screws ! ! ! ! I think you're now officially my Hero Of The Day for your chutzpah.....

[J Tiers]

I have not put any squeeze on the threads - there is still some back lash; meaning that when one flank is in contact, the other is not.

I did not split the nut. In fact I ADDED threads, so the average axial load will be LESS.

In that case, your solution is not involved in my comment.... You can chill.

But, how much did you leave? If you set it for the worn areas, will it not be tight on the unworn? And, if set for the unworn, will it not then be just as loose on the worn areas?

Should be fine for any new screw, but I am not understanding how it is better for one that has wear, which is normally a reason why ther is backlash to begin with. For a worn nut, sure, although a new nut might be good also.

Maybe there is more to it that I am not seeing?

Personally, I generally fix the loose bearings, tighten loose screws, and just deal with the small amount that is left. If it is not a "small" amount, then I will replace the screw, generally making a new one.

[Bob Engelhardt]

It's PEENED in place!

That must have made it all the more interesting to actually drill and tap for the differential screws ! ! ! ! I think you're now officially my Hero Of The Day for your chutzpah.....

I was very careful! It was made much easier by using the adjust-nut as a center punch guide.

[Bob Engelhardt]

...
But, how much did you leave? If you set it for the worn areas, will it not be tight on the unworn? And, if set for the unworn, will it not then be just as loose on the worn areas?
....
Personally, ... If it is not a "small" amount, then I will replace the screw, generally making a new one.

...
I have it adjusted so that there is a couple of thou back lash at the ends and 0.015+- in the middle. Yeah, there's still back lash, but just feels so much nicer when I can take it up by swinging the crank 54* instead of 112*.

Making a new screw - that would be a challenge (for me). I think that I might fit it on my lathe. I'll put it on my I'll-think-about-it list.

[J Tiers]

There is always the idea of bringing the existing one down to match the most worn area, that was mentioned before. Then as a new(ish) screw, you could adjust it closer, and get the 'lash down to where it's only a few degrees, and can be kept track of easily.

I don't like backlash when it gets to be enough in degrees that you kind of lose track of it. The point where it picks up again is not always obvious in "feel". I can visualize smaller amounts, but much over a quarter turn, and I start to lose track of it.

[old mart]

I have made several varieties of antibacklash improvements over the years, a work still in progress.
Firstly, the cross slide of the Smart & Brown model A. The LH ACME 8 tpi leadscrew was very worn in the centre and the nut was not one of their best designs. Actually a very cheap and nasty design, unusual in a toolroom lathe. The nut had a cut and the thin end with about 1 1/2 thread pitches was simply bent to reduce backlash. I chanced upon a new leadscrew with two bronze nuts advertised as 8 ACME. When it arrived it turned out to be 3mm trapezoidal. The seller kindly refunded my money and said keep it. 3mm pitch is close to 0.118" so I modified the leadscrew to replace the original, one nut replaced the old one and the second nut fitted with a spring in the gap between them. The spring was not strong enough to prevent backlash, but gave some preload when the second nuts fixing screws were loosened and retightened. We engraved a new scale with 118 divisions as the lathe was imperial. As most lathe work involves creeping up to a size together with frequent zeroing of the dials, the tiny difference between 3mm and 0.118" of 0.000110236" per turn hardly matters. The 0.001" backlash is largly due to axial play, not nut related. The compound and tailstock both have double adjustable antibacklash nuts as standard.

The Tom senior mill X axis leadscrew was worn in the centre by about 0.010" in a 3/4" X 5 acme, and there was room for a second nut. To retain the travel, I cut 4" extra ACME on the plain end of the leadscrew. These big threads are easy to pickup without much error. As the wear was not too bad, I simply rubbed the threading insert in the most worn part and reduced the less worn and new section to match. I got it close to good, 0.001" in the centre, getting tight at each end. One of the nuts is adjustable with a 40 tpi thread and locking collar held by 2 screws.

The work in progress on the drill mill is going to be even better as both the X and Y leadscrews, 23mm x 2.5mm trapezoidal are in unworn condition. There was plenty of room in both to add long antibacklash nuts without compromising travel. I am finishing off the adjustable axial thrust bearings.

Minimising backlash in a mill makes climb milling a better prospect, as long as the work is secure.