The Bronze-Loaded Acme Nut Experiment - Part 2

[homebrewed]

Sherline has a slightly different approach to an anti backlash nut design for their mills that might be useful for some other applications. They have a nut in the "usual" place, i.e., underneath the table. On the same axis, they have another nut on the exterior of the table. It is knurled on the OD. A thick washer, knurled on its OD is mounted on the exterior of the table such that the knurls are engaged, pretty much like gears. To remove the backlash, the screw holding the washer is slightly loosened and rotated to tighten the exterior nut, removing the backlash. Then the screw is retightened, locking the adjustment in place. There is no fundamental reason that the two nuts in an anti-backlash scheme have to be immediately adjacent to each other!

I used a similar scheme on the X axis of my mini mill: and now I can perform climb milling in steel, at least on that axis. Backlash is roughly .001 inch and has has been easy to keep that way, since the adjustment is point is easily accessible. The table design makes it relatively easy to mod the X axis in this fashion, not so much on the Y.

 

[AGCB97]

I think using the liquid turtle wax was the problem. Too thin. A hard paste wax like mold release wax would have been a little thicker leaving more clearance.

The problem is the melting point of the wax. The curing epoxy creates heat and if the wax melts below that temperature, the epoxy seeps through it. This is the case with car wax.

I had the same problem with mold release when building a fiberglass cowl for one of my experimental airplanes. I asked an epoxy expert/salesman at Oshkosh flyin about it. He said to use Johnsons Paste Wax. Cured the problem on later molds. By the way, that same wax is excellent for preventing rust on iron surfaces in high humidity conditions. I coat all the surfaces in my wood shop in the spring before those conditions arrive.
Aaron

 

[graham-xrf]

To keep the chuck from unscrewing you need what I call a chuck keeper. Why do I call it a chuck keeper. Because it keeps the chuck from unscrewing when running in reverse. Simple to make. The headstock end is a very close fit to the ID of the chuck. Then there is a long bar that goes through the spindle with a nut and washer that holds it in place. As long as the threads on the nut are different from the 8 tpi on the spindle the two different threads work against each other preventing either from unscrewing. Both ends are machined to fit the spindle to keep it centered in the spindle. I did not come up with this. Another forum member told me about it.

View attachment 359486

This I do very much like, and one of these surely must join the list of essential tools to have.

 

[graham-xrf]

Sherline has a slightly different approach to an anti backlash nut design for their mills that might be useful for some other applications. They have a nut in the "usual" place, i.e., underneath the table. On the same axis, they have another nut on the exterior of the table. It is knurled on the OD. A thick washer, knurled on its OD is mounted on the exterior of the table such that the knurls are engaged, pretty much like gears. To remove the backlash, the screw holding the washer is slightly loosened and rotated to tighten the exterior nut, removing the backlash. Then the screw is retightened, locking the adjustment in place. There is no fundamental reason that the two nuts in an anti-backlash scheme have to be immediately adjacent to each other!

I used a similar scheme on the X axis of my mini mill: and now I can perform climb milling in steel, at least on that axis. Backlash is roughly .001 inch and has has been easy to keep that way, since the adjustment is point is easily accessible. The table design makes it relatively easy to mod the X axis in this fashion, not so much on the Y.

This is nice and simple in concept, and compact. You get me thinking. If you happen to have a Sherline, or can get get a picture of your mill mod knurled part, when you hqave opportunity, it would be much appreciated.

I do hate backlash, though for many, at least a small amount has that "feel", so you know from where to read the graduations. Backlash or not, if ever the tool has to move back, I always back up a tad further, then adjust forward to the place I want. A knurled nut "take-up" arrangement need only be adjusted occasionally.

Of course, for CNC work, especially with servo motors, any backlash at all is big trouble! If the motor moves at at all, without the feedback showing any response (through backlash), it will cause the electronics to drive up the currents to max in microseconds, and then have to drive the other way to stop the overshoot, which is similarly being incorrectly reported. If you want to see stuff doing a blinding loud high speed vibrating instability oscillation jiggle either to overcurrent trip, or snapping stuff, then allow backlash. That is why one of the most popular mods to mill drills for CNC is to replace the Acme nut stuff with ballscrews.

Fortunately, here we are only talking of what is in a little compound, not normally in the works of a CNC drive train, and hardly gets moved much.

 

[MyLilMule]

To keep the chuck from unscrewing you need what I call a chuck keeper. Why do I call it a chuck keeper. Because it keeps the chuck from unscrewing when running in reverse. Simple to make. The headstock end is a very close fit to the ID of the chuck. Then there is a long bar that goes through the spindle with a nut and washer that holds it in place. As long as the threads on the nut are different from the 8 tpi on the spindle the two different threads work against each other preventing either from unscrewing. Both ends are machined to fit the spindle to keep it centered in the spindle. I did not come up with this. Another forum member told me about it.

View attachment 359486

Why are the simplest ideas such genius? LOL.

 

[RJSakowski]

Sherline has a slightly different approach to an anti backlash nut design for their mills that might be useful for some other applications. They have a nut in the "usual" place, i.e., underneath the table. On the same axis, they have another nut on the exterior of the table. It is knurled on the OD. A thick washer, knurled on its OD is mounted on the exterior of the table such that the knurls are engaged, pretty much like gears. To remove the backlash, the screw holding the washer is slightly loosened and rotated to tighten the exterior nut, removing the backlash. Then the screw is retightened, locking the adjustment in place. There is no fundamental reason that the two nuts in an anti-backlash scheme have to be immediately adjacent to each other!

I used a similar scheme on the X axis of my mini mill: and now I can perform climb milling in steel, at least on that axis. Backlash is roughly .001 inch and has has been easy to keep that way, since the adjustment is point is easily accessible. The table design makes it relatively easy to mod the X axis in this fashion, not so much on the Y.

The fundamental reason for having the two nuts in close proximity is that it maximizes the effective travel of the lead screw. If the lead screw is sufficiently long enough, not a problem but if your dealing with a prior design, any additional separation decreases your effective travel. In my particular setup, the nuts are separated by 1 inch and my effective travel is reduced from 5.1" to 4.1".

 

[Larry$]

The backlash compensating nut design or something similar is common on acme screws. I think it is on my lathe cross slide.

 

[homebrewed]

The fundamental reason for having the two nuts in close proximity is that it maximizes the effective travel of the lead screw. If the lead screw is sufficiently long enough, not a problem but if your dealing with a prior design, any additional separation decreases your effective travel. In my particular setup, the nuts are separated by 1 inch and my effective travel is reduced from 5.1" to 4.1".

The added nut reduced my effective travel by about the same amount. As you said, it depends on the particular machine's design.

This is nice and simple in concept, and compact. You get me thinking. If you happen to have a Sherline, or can get get a picture of your mill mod knurled part, when you hqave opportunity, it would be much appreciated.

I do hate backlash, though for many, at least a small amount has that "feel", so you know from where to read the graduations. Backlash or not, if ever the tool has to move back, I always back up a tad further, then adjust forward to the place I want. A knurled nut "take-up" arrangement need only be adjusted occasionally.

Of course, for CNC work, especially with servo motors, any backlash at all is big trouble! If the motor moves at at all, without the feedback showing any response (through backlash), it will cause the electronics to drive up the currents to max in microseconds, and then have to drive the other way to stop the overshoot, which is similarly being incorrectly reported. If you want to see stuff doing a blinding loud high speed vibrating instability oscillation jiggle either to overcurrent trip, or snapping stuff, then allow backlash. That is why one of the most popular mods to mill drills for CNC is to replace the Acme nut stuff with ballscrews.

Fortunately, here we are only talking of what is in a little compound, not normally in the works of a CNC drive train, and hardly gets moved much.

I don't own a Sherline mill (it was in the lab where I worked). Also, the Sherline's table design is substantially different compared to my X2 mini-mill, so my rendition is different, but uses the same principle. Here's a photo of my version:



The table on my mill doesn't have a flat surface for the external nut to press against so I added one, attached with two screws. The nut passes through a hole in the plate, and was drilled with a 1/16" hole to accommodate a piano wire adjusting finger that was bent at 90 degrees. It isn't glued in place, it must be free to rotate in order to allow adjustment. It was inserted in a slotted piece that's used to adjust the backlash.

Before drilling the hole in the nut I threaded it onto the feed screw to mark the location of the hole. I also faced the end of the nut that presses against the plate so it would bear evenly against the plate.

This and the Sherline scheme avoid the potential issue of applying radial force to the nut, which could cause binding or excess wear. That's why I didn't use a grub screw to hold the nut in place.

 

[RJSakowski]

I wouldn't recommend a dynamic backlash compensation system like the one on my lathe for use on a mill. It works on the lathe because for the most part, cutting is done with infeed so the force exerted by the work on the cross slide and the force exerted by the spring are in the same direction.

A mill must be able to exert substantial cutting force in both directions and a spring loaded compensation system would require a fairly stiff spring which would increase wear on the lead screw and nuts.

 

[homebrewed]

Well I messed up in my description of my mill mod a bit. I said that the nut passes through a hole in the plate I added, but it doesn't -- only the feed screw does. The hole is a fairly close fit so the nut has a decent contact area on the plate.

Before I made the plate, I cut a template out of paper to double-check the clearances.