I'm buying a new lathe chuck for my S7B, and the one I want is not available in Myford fitting so I'll have to machine a backplate for it. I've done this once before years ago but am considering the job again new. It looks like a suitable 5" backplate is available pre-machined to fit the spindle (Chronos list one), and while last time I machined everything from a raw casting this might be a time-saver if the quality is good.

The chuck is a TOS 5" 6-jaw, selected mainly because I do a fair bit of work on alloy tubes and a 6-jaw is better suited to gripping thin tubing. I would love a Griptru Burnerd 6-jaw chuck but the price new is silly (over £800 new), and I already have a 5" TOS 3-jaw for the S7B and I've never had an issue with the build quality of it.

I was thinking for alignment, after machining the backplate register to fit the recess, I could put a 2MT test bar in the headstock and grip that with the chuck jaws while the holes are spotted through (it's a front mounting chuck), remove the backplate for drilling and tapping, then re-mount and use the test bar again to hold the chuck whilst the bolts are tightened.

I know I should be relying totally on the rear machined register of the chuck to provide the alignment datum point but then thought it might be an idea to use an actual workpiece clamped in the jaws as being a more representative/useful datum? In fact, if the backplate were machined a couple of thou undersize there would be room for subequent adustment should the chuck start to run out after a period of use. I figure the chances of it shifting in use are minmal with 3x8mm bolts fully tightened.

If anyone has their own ideas as to how to go about this job for best results I would be interested to hear them.

Chris

Edited By Chris Heapy on 01/06/2013 10:23:45

I specifically want the 6-jaw for holding alloy tube which tends to distort in a 3-jaw, but it will come in handy for other work too – particularly delicate rings and discs. Yes, there are alternatives and I've been using them for years – which is why I recognise their limitations and want the chuck.

It isn't too heavy for the S7, especially with fairly lightweight workpieces. I would be reluctant to run it at top speed perhaps but all in all the 5" 3-jaw I posess has performed faultlessly and is the same weight. I would rather not have the added overhang caused by fitting a backplate but it's unavoidable.

The register on my existing chuck I machined the backplate a close fit in its recess so it cannot readily be adjusted by the 'loosen the bolts and tap with soft mallet' method. That is why I was thinking of not having it such a close fit on this one and indeed having that technique available to me. I think we have to assume a decent chuck is machined axially true – there isn't much we can do about it (well, not easily anyway).

Martin, the register is only nominally concentric with the gripping circle it may not be optimal in a 'less than premium' chuck. The quoted accuracy is 0.025mm TIR and there is no Griptru-type of adjustment. If it were a Burnerd I would be more inclined to trust it (but note that even with such a premium chuck there is apparently a benefit in having it adjustable – Griptru/Set-tru etc.) There may be something to be gained by tweaking the mounting – but quite possibly not if the run-out is not stable at all diameters as can be expected on a cheaper chuck.

Bazyle – you've stirred the memory now! I seem to recall a method for a DIY Griptru ajustment published years ago. I'll have to search the archives.

I fitted the backplate and tested the chuck this afternoon. Maybe I was just lucky but there is negligable run-out near the chuck jaws and about 0.06mm TIR about 4 inches out. I don't know that I could make any improvements on that so I'll just use it for a while and check again.

Another attempt at a video

Edited By Chris Heapy on 07/06/2013 17:24:15

I meant to mention the logic behind using a 2MT test bar to 'tune' the chuck alignment (mentioned earlier). As it turned out the process was not necessary with the TOS 6-jaw because the chuck proved to be very accurate straight off, in fact – about as good as you can reasonably expect from any scroll chuck. Certainly better than the 'reasonably priced' (to quote Top Gear) ER32 collet chuck set I have anyway.

Chucks are assembled from pre-machined parts and I'm not certain how much work is done after assembly to correct for possible errors. Perhaps the more expensive chucks are tweaked (like Pratt-Burnerd 'super accuracy' versions) and will be tested to ensure they fall within defined accuracy parameters but remember, chuck jaws are available as spares and are expected to fit right in and produce the same accuracy as the originals (just as the inside jaw set are expected to provide the same accuracy as the outside set), so there isn't much that will be done on an idividual basis post-assembly. On cheaper chucks it's a certainty you'll get whatever accuracy is present post-assembly, there simply isn't the production profit available with inexpensive chucks to pay for that kind of QC. I didn't know how good a TOS chuck would turn out to be – I was assuming 'middle-road' – but I discovered later TOS actually make Pratt-Burnerd chucks

Anyway, the idea with the test bar was to correct for any axial misalignment rather than run-out (the latter can be compensated for when mating to the backplate, and this should be carried out first in any case). The procedure depends on your lathe being perfectly setup and aligned (headstock spindle parallel to lathe bed, tailstock aligned to headstock) – you'll be wasting your time if that is not the case. Ideally, the 2MT test bar needs to be able to pass through the center of the chuck and be supported at the tailstock end with a live or superior quality rotating center. The bar is inserted into the spindle, the new chuck (without backplate) is mounted facing the spindle close to it, and the jaws tightened onto the bar, whilst the tailstock is brought up to support the other end. In this position a skim can be taken across the mating surface at the rear of the chuck body – the thinnest necessary to face it off cleanly – then it can be removed and re-assembled.

Doing this ensures that when a workpiece is gripped by the jaws it now lies exactly parallel with spindle (and lathe bed) and corrects for any potential inaccuracy within the chuck body/jaw assembly. Testing run-out on round work should now not rapidly deteriorate with distance from the chuck jaws and will be as good as possible with this type of chuck.

Well those were my thoughts (FWIW )

Chris

PS., I didn't mention I bought the chuck from Rotagrip Online who were a good company to deal with.

Edited By Chris Heapy on 11/06/2013 19:26:43