PDA

View Full Version : Anti backlash nut design



nakchak
12-15-2004, 01:39 AM
Right ive had an idea today about how to make a cheap anti backlash nut would really liek some feedback about this idea, sorry if someone else has had the same idea as me, not trying to robcredit where credit is due

ok, the basic idea is that you have 2 plates with a threaded hole in both separated by a compressable shim that offers resistance to being compressed.

The image (antibacklash_idea.gif) pretty much speaks for it self, the material id plan to use for the "shim" would be a plastic pipe of some sort

does this work as an idea?

cheers

Nak.

Ken_Shea
12-15-2004, 08:43 AM
Looks simple, effective, CHEAP :) I like it.
You would probably want to thread both A and B at the same time.

rustyolddo
12-15-2004, 09:02 AM
Remember too, you can work in the opposite direction, use a pair bellville springs and allow them to push outwards.

golash
12-15-2004, 10:11 AM
Hi
I do believe that should work fine. Its surprising, I was going to
place the same idea on the forum last week. I needed something
to take the backlash out of a ball nut.
Please tell us how it works out for you.

Regards barry

nakchak
12-15-2004, 01:09 PM
Had an idea for a cheap source of suitable pipe, Auqarium air pump pipe pretty sure its fairly cheap and sold by the meter, so not to much wastage

nak.

ViperTX
12-15-2004, 01:38 PM
Seems workable on an Acme or threaded lead screw. It might be better to have a threaded collar attached to one of the plates (that plate would not be threaded) in a manner that would allow you to remove the backlash while the whole mechanism is connected. That would allow adjustment without having to remove 4 screws and twirl the plate where space is limited.

PaulH
12-15-2004, 03:04 PM
If you want some cheap silicone tubing that might work, go to any R/C hobby shop and ask for fuel tubing. It's very flexible, can be cut with scissors, and compresses well.

A better idea would be to get some o-rings. They would be harder and would not rip and tear as easily as tubing.

An even better idea would be to use the rubber dampers found in the head of an R/C helicopter. They are rubber and come in a wide variety of durometer ratings. They are designed to absorb some of the vibration in the in the head.

Using springs may actually be easier as they would be self-adjusting. I would put one spring around each screw you use to hold the two halves of the nut together. If you need a good source for small, very hard springs, again consider a hobby shop. You can get springs for a Centax-style clutch that are designed to handle 40k RPM. They should work fine for your applications.

Jerry Dots
12-15-2004, 03:09 PM
Nak,
I built basically the same as you. Mine have a fixed nut in a tube and another nut with a flange on it clamped in the other end of the tube. Turn the flanged nut toward the fixed nut until it binds... then back off to release the friction and tighten the clamp. You have near zero backlash leadscrew. I have Acme threads. Before I made the setup I had .007" backlash. Now I have less than .0005" backlash. With my setup you have to have a tiny bit of backlash because it will bind if you don't.

Jerry

hpghost
05-15-2005, 07:03 AM
Nak,
I built basically the same as you. Mine have a fixed nut in a tube and another nut with a flange on it clamped in the other end of the tube. Turn the flanged nut toward the fixed nut until it binds... then back off to release the friction and tighten the clamp. You have near zero backlash leadscrew. I have Acme threads. Before I made the setup I had .007" backlash. Now I have less than .0005" backlash. With my setup you have to have a tiny bit of backlash because it will bind if you don't.

Jerry
can you show us how it looks like, I am very interested in the system..

lerman
05-15-2005, 11:05 AM
The major issue with this type of design is the strength of the spring. If it is too strong, it doesn't give when it encounters sections of thread with slight variation. (Another way to look at it is that if the preload is too high, it will take too much force to turn the screw.) If it is too weak, the cutting forces will deflect the spring and you will still have backlash.

Ken

spoiledbrat
05-19-2005, 10:58 PM
I have been kicking this around
The belleville springs fit within the shoulders between the two main parts. The can be stacked parallel or series to effect a wide range of preload, and compression; all within a small space. Two steel pins (.25" Dia) are press fitted into the aluminum block. These prevent rotation of the two main parts. The first ballnut is threaded into the aluminum block. The second part is made of bearing-bronze, and the auxilliary holes are drilled slightly over .25". A ballnut is threaded into this part as well. Set screws are installed into both main parts against the threads on the ballnut itself. A small shoulder is "faced" into both parts to hold the belleville springs. The adjustment of preload is effected by rotating the ball nut further into the bronze, and reinstalling the set screw.
Four 3/16th bolt holes are threaded into the aluminum nut to simplify design and mounting
It is important to note that the springs mentioned fully compress at approximately 23 lbs, but can be stacked, and compress at 46, 69, 92 lbs, etc. They can also be stacked such that you get the same preload force with greater amount of travel.

Metals- $15
Belleville springs- $3 for a pack of 10 (MSC item number 87780540)


Does anyone see this as useful?

spoiledbrat
06-06-2005, 02:25 PM
I machined the pieces and the system seems to work. Preload is adjustable.
The main piece (alum) has a 4-bolt pattern for mounting, and both of the blocks have a hole drilled and tapped for a setscrew. The setscrew should probably have a tiny bronze pin underneath, since it will be holding the hardened casing of the ballnut.

The springs used are called bellville springs, and accomplish the task pushing the ballnuts apart in a compact way. And cheap.

java77man
09-05-2008, 11:39 PM
Spoiledbrat:

I like your idea but what specific ballnut/ballscrew are your drawings designed for?
What do you use to thread the alum and bronze blocks with?

Thanks,
Jesse
(Java77man)

LeeWay
09-06-2008, 06:32 AM
Lapping or scraping the ways might help quite a bit. I did lap mine initially with some grinding paste and I could tell the difference after just 30 minutes work. Much less play. Apparently it gives more surface contact by removing some of the high spots. Keeping these well lubed will also be a benefit.



Not a lot of room on an X2 for double nuts. I think Hoss or possibly someone else may have done it though. I remember seeing the threaded end being cut off one of the nuts and then bellvilles used in between.

The method I used on my router is a simple one. Just two nuts, a spring and an extrusion.

The extrusion was 1.5" x 3/4". I milled an inch out of the top side, then installed 4 set screws. The set screws are locked onto the first nut and only just keep the second from spinning.
I had plenty of room here and this is why I went the easy route.
The principle is the same though as the spoiled brats design. His is much more compact though.


http://www.cnczone.com/forums/attachment.php?attachmentid=63747&d=1217069078