T51 Dixon style tool holders?

[JasonBModerator @jasonb]

It’s a wonder those Multifix holders works being constrained by those multiple scallops 😕

I wonder if someone wanted to unconstrain the Dickson type would it be better to just have say the two outer faces of the Vees making contact ? That would spread the single opposing surfaces as far apart as possible and unlike the one shown with the knurling tool if you were using it with say a boring bar along the lathe’s axis you would still have constraint at the end where the load is.

[JasonB]

On JasonB Said:

It’s a wonder those Multifix holders works being constrained by those multiple scallops 😕

…

It’s only a matter of making them sufficiently accurately

Samuel Colt had this sussed, and so did Dickson

… back in the day.

MichaelG.

 

[ChrisLHParticipant @chrislh]

Jason,

My take is that using only the outsides of the vees would be more problematical than trying to use all four. Viewed from on top, the mating surfaces would form part of a cylinder with the holder settling in different positions every time it is reclamped. Only the parallelism (or lack of it) of the very short mating surfaces would determine the position of the holder.

Michael,

We all seem to get along with the current designs so one assumes that they are made to an acceptable tolerance. I’ve yet to get round to doing a repeatability test though. Could they be cheaper if the constraint of tight tolerances between the two vees were removed ? Am I missing something though ?

[Michael Gilligan]

 

On Michael Gilligan Said:

It’s only a matter of making them sufficiently accurately

Samuel Colt had this sussed, and so did Dickson

… back in the day.

MichaelG.

 

Exactly!

Having checked a few Bison holders on two posts I can say that some of them only seat on three of the four V faces. But the holders still work and repeat position well. All the holders work on all the positions on the tool posts.

I suspect that some of this stuff is just made too cheaply!

Cheers

John.

[Clive FosterParticipant @clivefoster55965]

In practice a Dickson is not overconstrained. It would take ridiculous levels of manufacturing precision or lottery winning tolerance build up luck to make it so.

When in good order and correctly used they always find a stable “3 something” of contact lines or points. The correct way of fitting the tool carrier is to hold it in position and lock it. Using the locking system to pull it into engagement risks bending the adjuster stud a little. After which it will never work properly until the stud is fixed.

The design is actually very clever because the only requirement is that the male and female Vees are parallel in both axes and sufficiently close in separation to engage with each other. Any spacing error likely with a properly set up finish grinding arrangement is well within sufficiently close tolerance. The thing works so long as you have one line contact on one Vee pair. In practice it normally settle either on one Vee or on two inside or outside faces rather akin the a WW pattern watchmakers lathe bed.

A decade or more ago when I had apparent fit issues with my collection genuine and industrial quality clone toolholders shared between a genuine Dickson T2T and Rapid Original (Italian clone) T2 tool posts I did a bit of bluing and fit examination. A considerable variety of contact lines could easily be seen but no overconstraint.

I eventually tracked the problems down to bent toolholder adjusting studs. Turns out that each maker has their own views on the size of the flange at the bottom of the adjuster bobbin and the slot in the snail shaped part of the locking nut.  So not all bobbins will fit the  in all tool posts. After replacing all the bent studs and adjusting bobbin bottom flange depth and diameters to fit the Rapid tool post, which has a smaller slot than the Dickson, all works fine. The fit must be absolutely free. It was fairly clear that the bends on most of my collection was due to things being just a bit stiff where slot and flange didn’t mate easily so the operator resorted to a decent heave to ensure seating. Which wasn’t good thing to do.

Dicksons don’t to well when the innards are dirty. Mine seem to be total magnets for really fine swarf. Fortunately the are very easy to clean. Under a minute to pop the locker out wipe and replace.

The difficult bit when making Dickson holders is getting the inside flange face that the locking device pulls on in the correct plane as all tolerances stack up there. All too many of the less expensive imports have the flange too far back so they cannot lock properly. I suspect that real Dickson or industrial quality clone makers have / had cunning way of ensuring the flange position is correct during manufacture. There appear to be some fairly obvious ways of doing this but I know nothing about surface and production grinding and haven’t a clue as to what should actually done. I’ve a darn good idea as to what a not quite good enough way could look like tho’.

That said I imagine that if you reference off both sides of both Vees when grinding teh inner face of the flange it will be very close.

Clive

[BazyleParticipant @bazyle]

The question of ‘two outer faces’ of the V or two next to each other is exactly the same conundrum as the flat bed Myfords saddle running on the two faces of the forward sheer, ie close together, or the compensation for a worn bed using front and rear faces that has often been detailed on the forum and the magazine.
The dimensions given in the drawings earlier are irrelevant to the problem What I must do some time is measure the green marked face in Jason’s sketch wrt the Vs (using two round parallel bars in them) in all holders. This will  establish the range of variation, possibly dividing the holders in to sets from different sources. Then with reference to the locking lever cam offset the range of lever angles to expect the holders to lock decide what is acceptable or needs correction.
The simplest correction would be to loosen the cam so all fit easily then instead of needing to remove metal from the green area the correction would be to add a shim there.

[ChrisLH]

On ChrisLH Said:

[…]

Michael,

We all seem to get along with the current designs so one assumes that they are made to an acceptable tolerance. I’ve yet to get round to doing a repeatability test though. Could they be cheaper if the constraint of tight tolerances between the two vees were removed ? Am I missing something though ?

Sorry, Chris … your comment is rather lost upon me

This forum seems to have more than enough discussion about problems of interchangeability.

.

All I can usefully add [without reciting the basic principles of designing for interchangeability] is one example from my personal experience:

When Myford was in Beeston, I bought one of their ‘Dickson’ sets at a factory ‘open day’

Several years later, I bought a secondhand ML7R which came fitted with a similar Myford/Dickson set.

Before selling-on the ex-ML7R set [to a forum member who had ‘seen the light’ and wanted the real thing] … I did a quick check … every tool-holder fitted all of the stations on both posts to my entire satisfaction.

Not statistically significant, I know … but a very good sign that the Dickson was properly designed and appropriately made.

MichaelG.

[JasonBModerator @jasonb]

Fair to say the Dickson ones made by Wessex are made to a suitable standard to be interchangeable, as are the Bison and any others that cost a bit more.

The Dixon style made to a budget can suffer with interchangeability problems.

Having said that I have not found the need to alter the angles surfaces, all the ones I have bought from the likes of Chronos and RDG that had issues of fit were related to the retaining method not the constraining surfaces.

I seem to remember that some of the smaller Emco QCTPs also just has two angled surfaces at he ends of the holder much like the sketch I marked up so must have worked for them.

[JasonBModerator @jasonb]

Quickly “blued” an original Dickson, fitted and removed half a dozen times then slid it up and down the post several times.

Contact on only 3 of the 4 surfaces and even that is not fully over any surface.

Well it has produced the goods for me over 37yrs so I don’t think I will be making any alterations.

[ChrisLHParticipant @chrislh]

Jason,

I copied your test and got a similar result i.e. touches here and there but on all four surfaces in my case. No doubt there would be better mating under clamping load but difficult to demonstrate. My holders have the same markings as yours, I wondered what the W stood for (Wessex ?).

Michael,

Sorry not well expressed. Using the single vee and flat design (see Martin Connelly’s post in this thread) there is no  requirement for the very tight tolerance that is neessary in the double vee design (between the two vee features). So on the face of it we could have had a design that is no less effective (than the double vee design) but easier, and therefore cheaper, to make. So why is the double vee popular, is there some property of it that I’m missing ? Which design would Samuel Colt (or Eli Whitney come to that) have gone for ?

[ChrisLH]

On ChrisLH Said:

[…]

Michael,

Sorry not well expressed. Using the single vee and flat design (see Martin Connelly’s post in this thread) there is no  requirement for the very tight tolerance that is neessary in the double vee design (between the two vee features). So on the face of it we could have had a design that is no less effective (than the double vee design) but easier, and therefore cheaper, to make. So why is the double vee popular, is there some property of it that I’m missing ? Which design would Samuel Colt (or Eli Whitney come to that) have gone for ?

No problem, Chris … I think we are all heading towards the same understanding

The contact patches that you and Jason have both found are ‘fit for purpose’ … the trick with designing interchangeable parts is that all examples within the envelope of all the tolerances must be fit for purpose. …. The Dickson design, made to Dickson’s standards, demonstrably works … made to lower standards it may not.

The alternatives that are being proposed might well work adequately when made to looser tolerances [but we must check whether the baby has then been thrown-out with the bathwater].

In my earlier post, when I wrote “It’s only a matter of making them sufficiently accurately” … the key-word was ‘sufficiently’. … If the design is right, then that level of accuracy need not be implausibly difficult to achieve.

The reality [for real-world permutations of the Dickson tolerances] is almost certainly that the two notional vee locations, will resolve to some sort of Kinematic threesome which is adequate for the intended purpose.

John [elanman] got the point, I think

MichaelG.

 

[elanmanParticipant @elanman]

MichealG, I hope I did.

My take on things is that only a couple of points are critical. 1, the distance between the V’s and 2 the relationship of the V’s to the undercut lip ( green on JasonB’s marked up drawing). The first point is easy on a tilt head miller or a horizontal mill with the correct cutter but there are other ways. The second is not difficult to achieve if you think about how to measure it, with rollers, depth mic and slips (packing pieces).

When I bought my M300 it came with a Dickson set and I wanted some more tool holders. I rung up some of the suppliers and a few said they “should” fit, but only one said they “will” fit. And they did.

They are not difficult to make at home provided you have a miller bigger than a mini mill. I have made a baker’s dozen. Obviously I did not grind the V’s or harden the block but I’ve found that my home made ones fit just as well as the bought ones and with my limited use will last me out. I made them out of EN8.

 

Cheers

John.

[ChrisLHParticipant @chrislh]

A feature of my Wessex tool holders which has caused me problems on occasion is that the seat against which the tool is clamped is not always perpendicular to the vee clamping faces. Only a degree or so out but with, for example, a parting tool where side clearances are necessarily small I can be left with no side clearance at all. Obviously, it can be lived with by judicious use of packing but irritating on an article which is beautifully ground on 4 of its non-functional surfaces.

[Clive FosterParticipant @clivefoster55965]

I’d say that Jasons result suggest that the male and female Vees on his sets aren’t exactly parallel.

If memory serves me right all mine showed line or symmetrically intermittent line contact on opposing Vee faces. Effectively the same contact pattern as a WW pattern lathe bed.

What is clear from Jasons results is that the so-called over-constraint of four potential mating faces allows the system to cope with any small manufacturing errors and still achieve kinematically stable positioning as reported by ChrisH. It’s fairly obvious that if the various Vees are sufficiently parallel, of sufficiently similar angle and sufficiently similar in spacing you must get line contact on two opposing faces, subject only to the limitations of surface finish. Actually defining “sufficiently” in that context might be a problem but, if my memory its correct, all my industrial standard tool holders achieved that on both Dickson and Rapid tool posts.

Turning the problem round it would seem interesting to figure out what the maximum tolerable positional and dimensional error in the Vees is before the system cannot find a stable 3 point or 2 line position. My guess is the Dickson set up is least tolerant of any out of parallelism between the pairs of Vees in the vertical plane. Effectively requiring the tool holder and tool post to take up a skewed position which would seem to guarantee inadequate contact.

The single Vee and flat set-up proposed as an alternative is impractical as tight tolerances on the position of the flat is needed if both sides of the Vee are to be engaged to give a stable position. Any error in the flat position will tilt the toolholder Vee so the engagement points on the two sides will move up and down in opposite directions. Too much tilt and one side will run off the Vee completely so simultaneous engagement no longer occurs and the device will be unstable. In principle it’s possible but in practice the geometry involved is very unfavourable to accurate production.

Female Vee running on a half cylinder would probably be better as giving more scope for small rotations due to flat position error whilst still engaging on both sides. But back in my lab rat days I had a couple of “high precision” adjustable optical mounts made along those lines that were totally miserable devices. Desperately unstable and approximately unusable. The last resort of desperation when everything else was in tied up. Quality manufacture and the geometry looked sound but they just didn’t work.

Clive

[Author]

Posts