Mounting a 5C chuck on my lathe

[Ironman]

I have acquired a used collet chuck and spent the coin for a backing plate. I faced it off true and then cut it back for a register and achieved a tight fit on the register of the chuck.
After assembly I put in a 1" turned and ground shaft I use to check lineup with and I had a TOR of 4 thou. so I loosened it and tapped it a bit and the runout went to 1/2 a thou. Tighten it up and it went back to 3 thou. I could accept a 1/2 but not 3.
I've been told to cut the register undersize and snug the bolts and tap it true and then tighten. This seems the wrong thing to do. I've also had the suggestion to shim the low side of the chuck at the backing plate with a paper shim. I'm wondering if tere were stresses in the casting that were released when machining and this is showing up now.
I'm at the point where I think this back plate needs a skim taken off.
Any bright ideas?

 

[PeterT]

I don't see any harm in tapping a chuck true & re-tightening. Even the $$ Set-Tru type chucks need this occasionally. Now if you have to do this real often & are chasing your tail with perpetual runout, then something is amiss. It could be that your back plate boss is perfect but you inherited a chuck with non-concentric relief vs collet seat. Or it could have become slightly distorted over time for some reason. Of course this assumes your collets & seats are in good condition.

One thing you could do as a sanity check is: grip some perfect dowel stock in the collet chuck, but hold the dowel in a 4J chuck which you dial in with zero runout. Now, with the back of the chuck exposed, measure the chuck recess with DTI. Does it have runout, or eccentricity, or back face runout, or...? Any problems here would just be translated to your perfect back plate.

 

[thestelster]

Have a look at this thread when I bought and installed a 5c collet chuck.

Tool - 5C Collet Chuck

I've been humming and hawing about getting a 5C collet chuck for my lathe. I want a new Hardinge Sjogren chuck....that's not going to happen any time soon! I narrowed it down to either Bison or Accusize, 5C with integrated D1-5 mount. Accusize: $422.00 Bison: $1,881.86 I'm a poor starving...

canadianhobbymetalworkers.com

 

[thestelster]

Which style head stock mount does your lathe use?

 

[Ironman]

Which style head stock mount does your lathe use?

L00

 

[thestelster]

Did you check run-out of your headstock taper, front and back? It's best to be less than 0.0002" or better for maximum accuracy. And probably better if the runout is the same, near and far.

 

[Tom Kitta]

First figure out where exactly the problem is. I.e. is it the backing plate - chuck itself, collet?

So check runout of the backing plate - it may have different runout when you cut it and now.

Next test the chuck itself - if its Chinese it may need some grinding.

Finally test with different collets - even best collets do have runout that is quite annoying.

I make my own backing plates from steel stock so no problems with them changing - but they usually do have some runout - like under 0.001 if my D1-3 has right pins engaged. The issue is the taper in the 5C - need to grind it true. then only the best collets will give me good results. Should be within 0.001 say and inch or two from the chuck. Getting 1/2 will be hard - your collets better be new and of better quality & everything else error wise needs to cancel out - even spindle runout.

Remember there are set true collet chucks for a reason. https://www.globalindustrial.ca/p/5...Ok5rlv8PG1fRZ9WUp6SZ3WSLWCgjO9_hoCaPYQAvD_BwE

Even that Bison claims precision of only 0.0004 i.e. roughly 0.01mm

 

[Ironman]

Thanks for all the responses. I have "Poor Man's set trues" on my 3 jaw chucks. I will check the headstock runout, and then the backing plate runout. I have a shit ton of collets that were given to me on the death of a friend in Wyoming so they can't all be bad, some have never been unwrapped. I will have to get some turned and ground shafting in different sizes to check the collets If it is the chuck itself, well I have so little invested in it, I can afford to have it ground.

 

[Susquatch]

Some random thoughts. It seems that tightening it just settles it back. So something isn't right. I'd start as Peter suggests and find the problem before I cut anything. Reverse mounting the chuck is a good start. So is testing with a number of collets.

The chuck is used. So it's entirely possible that it's deliberately out to match the original backplate. The back of the chuck might need a skim cut with sharp hss. I know lots of guys trust carbide for that but I'm not one of them (not yet anyway ...... Inserts for aluminium might yet win me over though).

At any rate, a 5C collet chuck is worth the effort. My 5C is my favorite chuck by a wide margin.

I don't like the idea of shimming. Shimming could fix the runout but mess up concentricity. For me at least, concentricity is just as important as runout.

 

[Ironman]

I did some more work on the Collet chuck today and I can only conclude it is junk. I was given it with a bunch of machine tools that were left to me when a friend died, so it's history and brand name is unknown.
I unbolted it from the backing plate and checked the plate again. It was dead true and hardly caused a ripple on the indicator. So I reinstalled the chuck and ran a dial indicator on the outer edge of the chuck next to the back plate. This showed a small deflection, 1/4 thou? I can live with that.
Jumped down to the next step on the housing and I saw 2 thou deflection. I stepped down to the nose cone and read 15 thou deflection. This is visible to the naked eye at 130 rpm.
If the manufacturer cannot machine the outer shell any better than this, it is the hight of optimism to assume they got the inside and the cone to be true with the back of the chuck.
Oh well, I have 25 collets to use someday.

 

[Tom Kitta]

I did some more work on the Collet chuck today and I can only conclude it is junk. I was given it with a bunch of machine tools that were left to me when a friend died, so it's history and brand name is unknown.
I unbolted it from the backing plate and checked the plate again. It was dead true and hardly caused a ripple on the indicator. So I reinstalled the chuck and ran a dial indicator on the outer edge of the chuck next to the back plate. This showed a small deflection, 1/4 thou? I can live with that.
Jumped down to the next step on the housing and I saw 2 thou deflection. I stepped down to the nose cone and read 15 thou deflection. This is visible to the naked eye at 130 rpm.
If the manufacturer cannot machine the outer shell any better than this, it is the hight of optimism to assume they got the inside and the cone to be true with the back of the chuck.
Oh well, I have 25 collets to use someday.

Did you check whatever your plate is nice and flat and the back side of the chuck goes nice and flat to it? If your chuck is not precisely flat to the surface then it will wobble.

Also check the inside of the taper and see how much it wobbles.

If it is poorly ground you can always re-grind the whole thing as a project.

 

[Ironman]

Yes I checked the backside and wiped it all clean with light oil and then clocked it and it ran true.
I suspect it is a chinese cull and for the cost of having it ground true, I can get another closer for 250 bucks from Amazon. It's not worth it to have a piece of junk tooling around

 

[terry_g]

I was considering a 5C chuck off Amazon but even the better sellers chucks don't have good reviews.

 

[PeterT]

If the collet axis is perpendicular to the rear face & repeats with various collets, you might consider using it in the mill or on a rotary table vs. relegating to a shiny door stop, because you would dial in the part separately anyways. 5C chucks are a bit more compact that jaw chucks, but obviously limited in what they can hold. Maybe more of a production thing where you drop in like parts & they repeat concentrically. I've seen some guys mill flats on their chuck bodies to hold in vise jaws.

 

[Tom Kitta]

I was considering a 5C chuck off Amazon but even the better sellers chucks don't have good reviews.

Well I have cheap 5C and it does have out of the box a runout of over 0.001 - I actually forgot how much it has - as measured on the inside taper. I was planning to grind the taper true.

I think people have too high expectation for a collet chuck here - Even Bison has limits.

 

[Susquatch]

I did some more work on the Collet chuck today and I can only conclude it is junk. I was given it with a bunch of machine tools that were left to me when a friend died, so it's history and brand name is unknown.

Since your runout grows as you move toward the nose, I suspect it has an axial concentricity issue.

Assuming your lathe spindle is concentric to your bed, you could try cutting a new clean bar as big as your biggest collet in a regular chuck and then without removing the bar, mount your collet chuck in reverse on the bar. This will allow you to check the back of the collet chuck and perhaps even cut it square and concentric to the collet receiver.

But, sometimes it just isn't worth it. A wise man doesn't beat dead horses. I have a 5C Bison Collet Chuck that I love. I bought it new 10 years ago. They have gone up a lot since then, but if I knew how much I would love it, I wouldn't hesitate to buy one today.

 

[Tom Kitta]

Well, if you are in Calgary I would buy off you this collet chuck in hopes of fixing it as a project.

 

[Ironman]

I might have a way to get it to you in spring, but I'll fight with it some more first.

 

[Ironman]

Well, I did it. I solved the problem. I bought a new backing plate and collet chuck for $222 Cdn from Bostar Precision Tools on Ebay
The backing plate had to have a skim taken off for final fit. Out of the box runout was .001 on the face and the register have to have .100 taken off.
I disassembled it to check for chips and gunk and it was really clean inside. After completing the install a test bar put in the nose read .00005 runout so I am good with it.
And Tom, I'll get the old one to you in spring.

 

[Tom Kitta]

Sure I take it. My D1-6 could use a collet chuck.