How about some garage stuff?
I few years back I was getting frustrated with the runout of my 6-jaw Buck chuck. It's an older USA made chuck and they're expensive to replace - like $1500. It wasn't quite holding things repeatably. Chucks can have a little runout but they need to be consistent and mine wasn't.
Runout is basically how much out of true something is when placed in the chuck. If you have a lathe there are three, well, maybe five, ways to hold parts.
First is a three or six jaw chuck. Basically a much larger version of what's on your drill press. It's fast and can generally be counted on to hold within .003-.005" or as we like to call it "thou" and that amount is about the thickness of human hair. So not much. 6-jaws are more expensive and better at holding thin stock.
If you need to be more accurate you move to a collet. These are sized sleeves that hold a much greater tolerance but are limited to smaller diameter work generally. There are some specialty ones but they're special. Collets should hold work under a thou or .001"
Finally if you really need to be absolutely accurate you hold your work in a four jaw chuck. These don't self center so every time you put work into them you need to get out your dial indicator and slowly tweak each jaw one at a time nudging the part into concentricity. It's tedious but accurate and with patience you can get your part running to a few tenths or .0001"
There is also holding work on a face plate and between centers but it's just not something that is done much. I've used my face plate once and centers never. Most work on a lathe happens in a 3-jaw or 6-jaw chuck.
Since it's what gets used most often you'd like it to be accurate and after a while it wears and just isn't. Mine wasn't. You could turn a part and have it be round but if you removed it and put it back it wouldn't be concentric again.
My solution at the time was to cobble together a grinding wheel from a Dremmel extension.
I figured it was off and if I ground it with some really light passes I could get it back to serviceable shape. The shot above is me trying to grind it. Well, it didn't work. In fact I made it worse through my cobbled method. This is when, as an amateur machinist, you start to realize how much you don't know. Machining is interesting because it's the art of making parts and so as a craft you're sort of at square one - you often can't
buy a solution to your problems - you can only
devise (or
make) a solution. The better the machinist the better the solution.
My solution was weak. And as you'll see one problem often requires many solutions.
So I watched a bunch of YouTube videos and called my friend Scott who is a master machinist. He explained that my solution simply wasn't rigid enough and the only way to fix the chuck was with an actual tool post grinder. So I started looking for one on eBay and after a few months snagged a deal on a used one.
As the name suggests it's a grinder... you attach to your lathe's tool post. The reason you are using a grinder instead of just the lathe itself is because the jaws of the chuck are hardened and can't be cut with standard tooling.
So to give you a sense of the cascading problem/solution situation I'll run you through what it took for me to work this solution out. First off the grinder had a broken switch. Now don't expect a lot of photos here because I never expected any of this to become the hassle that it was. Anyway, it draws 11 amps so all the hardware store switches were 4 amps so I had to find a vintage switch on ebay and rob parts. Of course while the cord was being repaired I rewired it with a new cord.
Next it was missing a spring. Buying the spring would set me back $50 and since I didn't want to spend $50 or wait the week for it I instead decided to wind my own spring. Something I've never done before.
I made this small arbor and an arm and used my boring tool to hold the wire. Again, no action shots because I was sort of frustrated and just wanting to get the dang thing working. Several attempts later I fashioned a workable spring.
With the grinder running I turned my attention to the chuck. In order to accurately grind it you have to place the jaws under tension like they would be holding a part. My first attempt used a brass sleeve that I moved to the back to grind the front part of the jaws and then to the front to grind the back. Another weak solution.
There are other solutions but I found none for doing a 6-jaw chuck so I modified one idea I found to work with the 6-jaw but since I'd never seen it tried had no idea if it would work. But basically you make a plate and drill (very accurately) 6 holes for the chuck jaws.
Then you open the center hole enough to break into the six holes so the tips of the jaws can sneak through and you tighten the jaws. What you're not seeing here is that center hole was opened while the plate was mounted in the 4-jaw chuck...
Next we go back to the grinder. In order to use the grinder the wheels have to run true and be fresh so they need to be "dressed" which is basically cutting them to expose fresh abrasive. Since we're using a grinder because the steel is too hard to machine we now need something harder than the grinder wheel - which is a diamond.
And we need a way to hold the diamond.
I drill some holes into a 1" bar stock and drill and tap for set screws. I order a diamond dresser.
Now we have a way to hold the diamond tipped dresser.
I also need to make a shaft extension for the tool post grinder because it won't reach to the back of the chuck jaws without that. So I make that only to discover that it wobbles. I assume it's me so I make it three more times each one more careful than the last. Nope. Something is very slightly off.
I decide to reverse grind the shaft using the lathe so I take an extension shaft I've already made and mount it in a collet and then put a dressing tip in the lathes tool holder and dress the wheel.
Then I mount my turned shaft and attempt to grind out the wobble. I get it close but not perfect but figure I'll have to dress the wheel anyway so I move on.
My lathe chuck is adjustable so I put in a precision ground bar stock to check it's run out and then decide to center the chuck body and fix the jaws from there.
So let's review. These are the parts that I had to make to be able to use the tool post grinder to try to fix the lathe chuck. Don't think this happened in a few days. Each time I hit a road block I'd work on it until I got frustrated or ran out of time and then came back later. So this process has taken weeks.
Then, to make sure you're not grinding a cone or at an angle you put a piece of precision ground stock into a collet and mount an indicator to the grinder and run it back while nudging the grinder into place.
Finally, I dressed the new smaller grinding wheel, mounted it up in the improved but not perfect extended shaft and tightened the jaws of the chuck and started grinding away. You'll notice that I've covered the lathes ways with towels to help keep grinding dust off them. That dust mixed with cutting fluid is like valve grinding past - bad for the lathe bed.
After many passes I kept getting the same error when I'd measure either test stock or the actual inside jaw faces. Every time the same exact error - .012 out or +-.006" so I'd somehow spent all this time and accomplished absolutely zero. Worse, I've probably ruined the chuck to the point where I couldn't even sell it to get my money back.
At the point where I'd given up and started shopping for a new chuck to replace this one I decided to re-true the chuck body to see if that might help. I didn't understand how I was getting the error and thought maybe if I try to true the test stock with the chuck body and regrind I might be able to save it still.
To my complete and utter surprise the chuck was perfectly true. Okay, not "perfect" but less than a thou or about .0003" which is about as good as you could ever expect a chuck to get. I took out the bar and reset it a few times, swapped ends and it was always the same. Yeah! I don't think I can explain it but I'm not going to argue with it.
So, after several weeks of work, and dozens of hours I've somehow managed to save the chuck. Along the way I learned a lot and now I'm a marginally better machinist who still can't quite understand how I managed to stumble onto my solution. But I'll take it!
Gregor
