Cross slide play? How to get rid of it?

[bellyupfish]

Ok guys,

Looking to improve my cheapo HF lathe/mill in the interim until I upgrade.

So far, for my purposes, this thing has sufficed, but I can see where a nicer unit.

How can I adjust this play out of the cross slide?

Thanks

http://youtu.be/7iCO1Qd9AYY

[dp]

Check the nut to see if it is a split nut. If so you can adjust the split gap to reduce play. This also reduces the life of the screw and the nut. Otherwise, replace the nut with an Evanut as described here in the archives.

[justanengineer]

That "slop" is called backlash and personally I'd just work with it. Theres always going to be some amount of backlash bc of the difference in fit between the feedscrew and the feednut, if thats ~0.020" thats not great but not too awfully terrible for any machine. If youre set on minimizing it you can turn yourself either a new feedscrew, nut, or (I usually do) both. One trick to see if one might be wearing more is to measure backlash (ie. flip the dial and feel like you did in the video) at different points in the travel, if there is any significant changes in backlash at any point in the travel then the feedscrew is worn. If the backlash is pretty consistent across the range of travel then the feed nut is likely a bit worn. If you single point thread each on a lathe you can get a VERY close fit to them, Ive had backlash down to 2-3 thou on lathes in the past.

[oldtiffie]

I don't bother about it as its not serious -at all.

I just work with or around it.

The"slack" is not just in the lead-screw and nut but between the lead-screw collar ends and the thrust block/son the end/s of the table.

[bellyupfish]

Well, alright. I won't sweat it.

[darryl]

There's a few areas where this play can be- one is the lead screw threads in the nut. Another is where the nut mounts to the slide, another is whether there's a bend in the lead screw, and another is where the lead screw goes through the casting and the collar and handle are fixed.

First thing I'd do is carefully check to see if the lead screw is bent. You normally can see this without removing it from the assembly. If you can't see a wobble, move on. Next thing is to check the nut where it mounts to the slide. The nut should not move fore and aft, and should not rotate to any degree at all. You should see no movement of the nut at all relative to the slide.

If as I suspect the play is in the bushing where the screw goes through the casting, it sometimes makes sense to machine up a collar that can be press-fit over the shaft and up to the shoulder where the threads start. This collar becomes the thrust bearing, and you would make it larger in diameter than what is there now. Since it will bear directly on the cast part, it should seat against that without wobble. On the handle side, something similar could be done- or perhaps adding the inner collar would help the outer part seat better against the casting. There usually isn't a good fit in this area, and over all you'd be doing good to get the play down to 5 thou. Chances are that after some use you'd find that you'd have to live with about 10 thou of play. If you try to make it tighter than that, you'll find that it binds over part of the lead screw revolution, and it loose at other spots. I would have to say from looking at the various machines in my price range that not much attention is paid by the manufacturer in getting a good fit in this area. If the hole in the bushing where the lead screw shaft passes through is significantly larger than the shaft, you will have some side play, and that will make it difficult to lose play without getting binding. You might be able to shim the bushing, or make a new insert that is a closer fit. Chances are there's no insert now, just a hole through the casting, so a shim would likely be your best bet to get better control here.

What I've done on my mill is add needle thrust bearings to each side of the boss where the lead screw passes through. That way you can snug it up with less chance of it binding at any point. Whatever you do here of course only affects play in this area. There is going to be play that you can't really get rid of in the lead screw threads and nut area. You can adjust this on most machines, right at the nut itself, but it will soon wear into a new 'comfort zone' and you will have to live with that or drive yourself crazy.

You have 20 thou of play- that would be considered acceptable by most I'm sure. You can improve it, and it might make sense to do so as the cross slide is operated a lot. Other areas on your machine seem to be good from what I've seen in your video, so if this is the only area to improve I would do it if it were my machine. Remember, the inside lead screw collar takes most of the forces from machining as the tool is being forced into the workpiece by that collar only. If all you do, or all you can do is improve the contact area here, it would be worth the effort.

[oldtiffie]

The best and short answer to the OP's query is to fit pre-loaded precision "ball screws" (with nuts) to the lead screw as and precision pre-loaded thrust bearings to the mill table thrust blocks and lead-screw ends all as are fitted to many/most (all) CNC machines.

[Richard King]

On your video around 38 seconds the little collar looks like it moves forward when you turn the feed screw. The first thing I would do is to set a grab hold of the turret and pull it toward you and then push it away from you. Observe what is moving. You need to play detective and find the problem on your own and not "just let it alone" . If you got a book with the machine see if it tells you how to adjust the feed nut backlash. Take a look at the middle top of the cross-slide and see if there is an Allen cap screw and make sure it is tight. It might have 2 cap chews and a set screw. The center set screw is like a hedge and if you loosen the out-side cap screws and tighten or push down the wedge with the center set screw and it spreads the 2 outside nut blocks. Loosen them like a 1/4 turn at a time. It could also be the thrust bearings or washers on the ends of the screw Many times it has 2 nuts that works as a double lock.

There is all kinds of info on you tube. https://www.youtube.com/watch?v=A99YLU36AEM Try to fix it. It might save you a world of hurt. Rich

[PixMan]

The best and short answer to the OP's query is to fit pre-loaded precision "ball screws" (with nuts) to the lead screw as and precision pre-loaded thrust bearings to the mill table thrust blocks and lead-screw ends all as are fitted to many/most (all) CNC machines.

For a Horror Fright class machine to be fitted with precision ball screws, the term "as casting pearls before swine" comes to mind.

Use it as it is, as intended, and accept the mediocre (as best) results of it.

[bellyupfish]

For a Horror Fright class machine to be fitted with precision ball screws, the term "as casting pearls before swine" comes to mind.

Use it as it is, as intended, and accept the mediocre (as best) results of it.

I won't be throwing much $$$ at it, I'm upgrading soon..

I have been able to make some descent items for my purposes.

Pistol grip bolt and folding stock hinge pin.. Be kind..

[garyhlucas]

0.020" backlash and you are concerned? My clapped out old Bridgeport is closer to 0.040"! I am working with a new guy that has machining experience. He is really concerned about it. I just laugh because he comes from a large CNC background where you mostly forget about these issues. The hardest part to learn is approaching a location from the same direction all the time. Once you get that concept you can get results as accurate as the pitch accuracy of the screws. If you want better than that you'll need digital scales on each axis. Even then you will be limited by the straightness and squareness of the ways etc. I'd get a long throw (1" or longer) dial indicator with a magnetic back. That gives you an accurate location over a short distance cheaply. Like when cutting a keyway and you need to move over just a couple of thousands on both sides to keep it centered but make it a little wider.

[dp]

This has become one of those odd HSM threads where commenters ultimately ignore the OP's question and begin to question his motives. So - find your post and ask yourself if you answered the OP's question or if you questioned his motives or the worth of his goal, or provided any other information not asked for.

OP: How can I remove slack from my cross slide?
Answer: what ever it takes to remove the slack. Anecdotally, that much slack is rather harmless but if removing it is not difficult it is probably worth doing. (Several valid suggestions have been made in the comments)

Wrong answer: (paraphrasing) Who cares about a little slack? (didn't answer the question).

[J Tiers]

Eh, there is the actual question, and there is the implied question, and there is the question that should have been asked (QTSHBA).

Answering the one that should have been asked is often the most direct and useful way to handle it.

In this case the QTSHBA is "I have what seems like a lot of backlash, is that bad, and what do I do about it?".

And, the answer is that 0.020 isn't bad, backlash is a fact of life, and you deal with it by moving final moves in the "takeup" direction.

That answer was given, at least once, and probably several times if you put more than one response together.

Answering the actual question is often to go flying off on a tangent which is ultimately just "noise" when it comes to dealing with the original problem.

Example: the familiar "I discovered my spindle is bent, what now?". And the discussion uncovers the fact that there is actually a quite different problem. Advice on straightening, replacing, etc, is just "noise", even though it technically answers the actual question.

[bellyupfish]

Cross slide play? How to get rid of it?

Well I broke down the cross slide today and was able to reduce the backlash quite a bit.

Thanks for the help guys.

[Richard King]

What did you? Where was most of the backlash? How did you repair it and how much do you have now?