Wobbly chuck, how to diagnose?

[AlexPeel159]

Why backwards? I face my backplate by putting it on as it would normally be used.

First face the face, use light cuts. Use power crossfeed to do most of the facing, at the end go manual .
You should be less than 1/2 a thou ideally 0 run out on the face. Then do the registration boss for the chuck.View attachment 498240

So the backwards is because its not seating right against the spindle and no amount of facing it on chuck side will fix that. I guess the argument can be made that if the chuck side was faced to the miss-seating then the miss seating does not matter.

That "gap" is just a shadow in the picture. Thats the threaded spindle coming through but its fairly tight fit there. There is no real gap.

I do think I will do this face as well after a pass on the otherside. Question, the face you have marked "do this second" has a tight fit right now with the chuck. Is it a problem if I reduce that and loosen the fit? I guess the bolts still hold it in place but will it cause any problems?

 

[AlexPeel159]

Also, pic 5613 seems to show a chip embedded in the plate.
that could have cause the plate to bend,

Thats not a chip there, thats the dimpling (really a protrusion, with a small broken hole in the middle) from the head of a drill bit coming in from the other side and just barely not breaking through. Its about 25 thou proud but its in an area where the corresponding chuck face has about a 26-28 thou deep recess in the center of the chuck. I expect when I face it it will remove those bumps and just leave the end of the hole.

 

[woodchucker]

So the backwards is because its not seating right against the spindle and no amount of facing it on chuck side will fix that. I guess the argument can be made that if the chuck side was faced to the miss-seating then the miss seating does not matter.

That "gap" is just a shadow in the picture. Thats the threaded spindle coming through but its fairly tight fit there. There is no real gap.

I do think I will do this face as well after a pass on the otherside. Question, the face you have marked "do this second" has a tight fit right now with the chuck. Is it a problem if I reduce that and loosen the fit? I guess the bolts still hold it in place but will it cause any problems?

No it won't really cause an issue. You are now relying on that being dead on, but if the chuck has run out, you will be able to losen the screws, slightly and tap the chuck true. I did this with my Bison. Was able to get it to zero. And it's been repeatable. So nice to have a chuck that is dead on. It wasn't before the fix.

 

[AlexPeel159]

Next question is what rpm should I be trying to do this facing at? I see that the simplified formula is rpm = sfpm * 4 / diameter. Problem is I dont know what type of metal this is. Also the diameter at the outside is about 3.25" for the first facing I am going to be doing (just the registration portion of the back face which is a bit proud). So picking a sfpm of 125 (seemed like that is on the lower range for a variety of steel) I would get about 154 rpm. That sound like I am in the right rpm ballpark? Or should I be using a smaller diameter because its only going to be 3.25 at the start? 2 being a bit more in the middle would give me 250 rpm. Going to read through the south bend lathe guide tonight before attempting this but wanted to hear from people who are experienced.

 

[woodchucker]

I don't usually worry about the speeds and feeds. I do it by feel. So I can't answer. I like a medium speed to start out.. it could be steel, semi steel or cast iron.. Start cutting and see how it behaves. Cast iron is messy, and you should cover your ways and as much as you can.. aluminum foil is a good material to block the chips from fouling your ways. Do a good cleanup after, vac it all up. then use paper towels, then alcohol on a paper towel.. feel the ways and other areas with your bare hand. If you feel grit keep cleaning.

 

[AlexPeel159]

Was not able to get to it today but I did think of a couple things I wanted to test.

Put the faceplate spindle side down on a surface plate and checked the clearances. Wanted to confirm that the faceplate was slightly tilted. Feeler gauges confirmed there was more of a gap on one side than the other. Which makes sense, we already know the plate is not level. Odd thing was that there was less clearance on the side that was farther towards the tailstock when I measured it on the lathe. I was expecting the opposite.

Started testing with the micrometer and the plate is thicker on that side than the other. 0.5863 vs 0.5889. So the plate is all sorts of out of whack.

Does not really change anything but its interesting.

 

[WobblyHand]

Was not able to get to it today but I did think of a couple things I wanted to test.

Put the faceplate spindle side down on a surface plate and checked the clearances. Wanted to confirm that the faceplate was slightly tilted. Feeler gauges confirmed there was more of a gap on one side than the other. Which makes sense, we already know the plate is not level. Odd thing was that there was less clearance on the side that was farther towards the tailstock when I measured it on the lathe. I was expecting the opposite.

Started testing with the micrometer and the plate is thicker on that side than the other. 0.5863 vs 0.5889. So the plate is all sorts of out of whack.

Does not really change anything but its interesting.

It does mean that the back plate is out of spec in several areas, either due to some lathe incident, or that it wasn't made correctly. So if it were me in the same situation, I'd determine if there's some surface that was trustable (use as a reference) and figure out a way to machine it trueness again.

I did make my own back plate from raw stock, which was fun, but I'm not sure how to advise you on recovering a funky back plate. Hopefully some better machinists than I can provide more details on what to do. Strangely enough, I found it far easier to start from scratch than to machine a partially machined back plate. When I did it at least I knew which surfaces I could trust and the features that were truly concentric as opposed to what "looked concentric".

 

[AlexPeel159]

It does mean that the back plate is out of spec in several areas, either due to some lathe incident, or that it wasn't made correctly. So if it were me in the same situation, I'd determine if there's some surface that was trustable (use as a reference) and figure out a way to machine it trueness again.

I did make my own back plate from raw stock, which was fun, but I'm not sure how to advise you on recovering a funky back plate. Hopefully some better machinists than I can provide more details on what to do. Strangely enough, I found it far easier to start from scratch than to machine a partially machined back plate. When I did it at least I knew which surfaces I could trust and the features that were truly concentric as opposed to what "looked concentric".

Yeah I am just hoping that I can face the spindle side face and get it to sit flat on the spindle register. Then that can be the reference for everything else.

I have two things giving me some confidence. Its so messed up I dont feel like I can do much worse. Second, a buck chuck might have fallen into my ebay cart yesterday. I am still going to try getting this one fixed but even if I can get rid of the runout its a very worn chuck.

 

[Jimsehr]

After 70 years of machining my first thought is the register might be the cause.
I would bore it out a little bigger.
Then mount and check plate to see if it is true.
If that does not work I would get a new back plate.
As I said before don’t worry about the chuck until the back plate is dead true.

 

[AlexPeel159]

Was able to work on it the last few nights. Been quite the learning experience.

Made my first real cuts with the machine. Regular cuts, facing cuts, boring cuts. Reduced the speed to very low after I dulled one of my tools and just ran it at 50 rpm, 0.0018 on the power feed and only a few thou depth while doing all of it. Tried to use cutting fluid the whole time.

Learned the plate was cast iron. Also learned why people disparage lathing cast iron.

Sharpened my first tool bits after I dulled one facing the plate.

Faced the register to the spindle, but that did not quite do it. Noticed that just before it hit the register it seemed to get resistance on something. Figured the threads did not quite have enough clearance on the spindle. Bored out out the faceplate and removed the last thread. Its a little harder to start on the threads now but it gets all the way flush now. Flipped the plate over and faced this in many passes. Noticed that after my supposed to be final cut which was only a thou or two I had to do three additional spring passes across the last 1/2 inch of the outside of the plate before the tool finally stopped cutting. Not sure why this happened. Is cast iron springy? Would it flex at the outer edge? I had my carriage locked down. Then I took off a few thou of the chuck registration cylinder in the middle until it was concentric and just a bit smaller.

The surfaces I got never seemed quite smooth even though I was using minimal depth, lube, and a tool with a large radius. It was not grooved but instead felt like it was just a bit rough like sand paper. Not really bad just a slightly gritty texture. I am not sure if this was just because I was running it too slow.

Cleaned and relubed the chuck put it back on the lathe and remeasured using the 2" .500 pin gauge. Unfortunately I was only able to get it to about 9 thou out even with loosening the chuck to facing plate bolts playing with the wiggle room that I gave myself on the last step. Immediately regretted not testing the face before I reassembled the chuck.

Overall the chuck no longer visibly wobbles. Seems very straight now while spinning, like my 4 jaw chuck. The 9 thou while not great is much better then what I was getting before.

I am probably going to take the faceplate back off and dial test it again when I get more time. I could try cutting a bit more relief into the chuck mating cylinder to see if it will give me more adjustment but not sure how much I can get in the end from that because the bolts wont let it wander too far.