Collet runout. Is this normal

[JDSParticipant @jds]

First off thanks to everyone who has offered advice, I know have a new spanner that is much better than the C spanner, I have found the indeed the tightness of the nuts that secure the chuck to the backplate can make up to 5 tenths difference and I have learned a new technique. Tighten collet chuck to bar to check the back of the chuck is square to the taper.

Sadly all this I am slowly losing the will to continue (not really, but if i had enough hair i would be pulling it out)

I started again from scratch. but i also skimmed the back of the chuck on advice from RDG. Unfortunately this has not made any real difference.

But what i have noticed is when it is all bolted up tight with hardly a flicker on the dti at the front of the taper the back of the taper and the face of the threaded spigot have up to 1 thou runout.

I have uploaded a short youtube video (32 seconds) to show the measurements

Do you think the chuck might be faulty or is there something else I can try.

Once again really big thanks for all you help

[Pete RimmerParticipant @peterimmer30576]

That taper is running out from the body. The OD, thread, face and taper should have been machined on one operation. That one certainly has not, you can see it in the face chamfer. Most likely it was turned and threaded them moved to another machine for the taper to be turned or ground.

First thing I would do is blue up the rear mounting face to make sure it's flat and has no raised burrs around the bolt holes. Then I'd mount the chuck on the backplate and if it's hardened take a PCD insert and turn the taper. If it's un-hardened I'd use carbide.

To set up for turning, use your DTI on the inside face at the point where the cutting tip will be. Watch the dial as you turn the spindle and mark the point on the taper where the dial is mid-swing. This will be the closest you can get to the true collet angle. Now swing your compound round to the taper angle and run the DTI in the taper. Adjust the compound until you can run the DTI up and down the taper with no swing. Make sure your gibs are well adjusted and check it again. Now just paint the taper with a marker pen and turn it true until the pen is all removed. Remove the minimum required and remember that when cutting a taper this way the centre height of the tool is critical.

You might have to face and chamfer the front edge when you're done to make sure the nut doesn''t bottom out.

[old martParticipant @oldmart]

You could try chucking up a steel bar in the lathe and skim it to a size to match one of your collets. Then fit the collet chuck backwards and check its runout in every measurable aspect (concentricity of OD and register, also flatness).

[not done it yetParticipant @notdoneityet]

One forum I frequent seems to have quite a few members that have a very low opinion of the quality of items from that supplier. I always ask for a receipt for any new tools sold/bought at the shows/exhibitions. Good luck if it needs to be returned as faulty (let us know how you get on if you do need to).

[Neil LickfoldParticipant @neillickfold44316]

Posted by Bill Phinn on 29/01/2021 00:39:46:

Posted by Neil Lickfold on 28/01/2021 21:38:54:

I have lately been buying the high precision ER collets from Regofix

 

Would you mind telling us where Regofix products are to be had in the UK, Neil, assuming they are?

Thanks.

I think this the UK  retail agents. The wholesalers are MA Ford. 

http://www.brunnermachine.co.uk/rego-fix-toolholding.html

Neil

Edited By Neil Lickfold on 31/01/2021 09:58:15

[Neil LickfoldParticipant @neillickfold44316]

The only thing not covered in your video is the thread concentricity. If the outer thread is concentric to how it is mounted, then you can give the inner taper a light skim cut to clean it up. The very front of the holder does nothing. When you recut the inner 8 deg per side, also recut the inside edge break so that you have a concentric start to the 8 deg taper.

[John HaineParticipant @johnhaine32865]

Another thread here on this subject mentioned that the tightening spanner needs to apply torque but not sidways force, which can also push the nut of concentric. So a big ring spanner is not the way to do it! For my ER16 collets on the Unimat I have invested in a large socket to fit the nut to allow this. Spindle is locked with a tommy bar through the hole at the back of the mandrel. Also use a ball bearing nut.

[HOWARDTParticipant @howardt]

If you look at the tightening torque Regs-fix recommend for ER32 using their nuts it varies between 24Nm for the small bore collets with rear counterbore to 136Nm for the rest. Accuracy also varies depending on the collet type from 2-10 micron maximum. ER collets are primarily used for tool holding in industry where the runout affects cutting size and tool edge life when trying to achieve reasonable economical piece price, higher accuracy etc moves you into HSK and shrink fit. For a hobby machinist, and I am talking non precise here, look and function rather than production fitting, tool life will be far less of an issue.

Using ER system for work holding in place of a chuck is useful for small diameters. I do it, but in most case the part can be machined in one mounting to produce a part which will be concentric to itself where cut. It is only where second operations from the non cut end that the tolerance comes into operation, or moving to say collet blocks. Whenever you release a part from a holding device and refit it there will be issues with concentricity, no matter how accurate the holding device is. When designing work holding you have to be aware of the stack up of tolerances across the whole process used to produce a part, that is one of the reasons there are muti-axis machine tools parts can be produced from two setups, one to produce the datum face, holes/edges and the second to finish the part.

So if using collets for work holding rather than chase microns on a cheap system, use a work holding method to allow you to tune the part to the required level of accuracy.

[Alan JacksonParticipant @alanjackson47790]

What do you think of this spanner and ER32 fitting? There are 16 grooves and 15 holes in the nut, so there is always a zone for tightening/undoing the collet without locking the spindle.

Alan

Edited By Alan Jackson on 31/01/2021 11:35:02

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Alan Jackson on 31/01/2021 11:32:26:

What do you think of this spanner and ER32 fitting?

Alan

.

I admired it principle, as soon as I saw it in MEW, Alan

… I thought it was very clever

My only reservation is that the load on those pins is [if the prescribed tightening torque is used] potentially enormous; so I will be interested to see how well it works in practice.

MichaelG.

[Turbine GuyParticipant @turbineguy]

Posted by JasonB on 28/01/2021 17:25:43:

Picture not showing up but I suspect you can see a hole with an offset cut out this is so the collect can be snapped into the nut so I will ask again are you correctly fitting the collet into the nut before screwing the nut onto the chuck. If you don't do that it will push the collet sideways and give the measurements you are getting.

The nut should also pull the collet out of the chuck when undone, if the collet is staying stuck in the chuck that is an indication you have not fitted it correctly into the nut.

See pic at the top of this page the lip at the bottom of the section through the nut is smaller than the lip at the top.

Edited By JasonB on 28/01/2021 17:32:58

I just received a collet set that was sold in Ebay for use with Umimat 3 &4 lathes and was listed as made by JJG Turning Mlling Devices. I quoted this post by JasonB because I read it before my collet set arrived, and despite his excellent instructions, I had the same problems with runout as JDS. The collets and collet chuck were still in sealed bags so I don't believe they were ever used. I followed JasonB's instructions for putting the collets in the chuck and tried different sizes at the upper limit and lower limit printed on the collets. I also tried tightening them as much as I could using the bar and wrench included in the kit. In all cases I tried, the runout was worse than what I would get using my EMCO drill chuck and an additional drill chuck made by another unknown manufacturer. I mentioned the manufacturers name to caution anyone from purchasing a collet set shown as being made by them. I hope this helps someone from making the same costly mistake.

Byron

[JDSParticipant @jds]

Cheers Turbine Guy that’s good to know

I follow up on my original post, I still don’t have a resolution after weeks of back and forth with RDG and different suggestion of what I need to try, including machining the Chuck they have asked me to return it for test and inspection incase the error is related to my skills rather and their parts.

Even though I sent them several videos of the runout on the backplate, runout on the Chuck back front and taper, Stuart’s blue on mating surfaces to show contact etc etc.

They have received it but not yet come back with an answer so I wait to see what happens.

I will update you guys with the outcome.

Thanks for all your advice and help

 

 

 

Edited By JDS on 14/02/2021 20:46:42

[Turbine GuyParticipant @turbineguy]

JDS

I was luckier than you. I bought my collet set on Ebay and several people suggested that I request a refund. Ebay's refund policy is well thought out with protections for the buyers and sellers. I followed their instructions exactly, and although the seller did everything he could to avoid giving a refund, Ebay forced him to give a full refund. I hope that you can find a way to repair the collet holder. My runout was even worse than you got and one of the things I found that may apply to your collet holder, was the bore of the mounting hole of my collet holder had quite a bit of clearance with the lathe shaft it mounted on. Any machining done to correct the runout would not have worked without adding a sleeve and making a tight clearance fit between the chuck and lathe mount shaft.

Hope this helps,

Byron

[JDSParticipant @jds]

Ok so to wrap this up from my perspective.

After lots of back and forth RDG have agreed they the chuck I returned to them does not run true. Initially they said that this was because I had over machined the backplate. I reminded them to re-watch the videos I had sent them and check the messages. When the register was a spot on fit the back of the taper was running out so it had to be undersized to get the taper to run try at the back.

Long story short they sent a new Chuck. The new chuck is immediately visually better, for a start the chamfer a the front of the taper is a uniform thickness unlike the first one.

The new chuck has about 2 tenths runout so I am happy with that

Thanks again for all your help and input.

[Author]

Posts