Screw-cutting – Is There Something We’re Not Told?

[HopperParticipant @hopper]

Well done. As often happens, an accumulation of, ahem, oversights, most of which you have overcome with a bit of gumption and perseverance.

Re the remaining one or two thou taper along the shank of your 3/8″ diameter shank: You should be using a tailstock centre on a job of that diameter and length, so it should be just a matter of adjusting your tailstock offset by a tiny tiny amount to compensate.

Details on how to do that are in the ML7 Myford Owners Manual, available for free download all over the net. And there has been at least one article on it in MEW in the past 10 years or so. In a nutshell, there is an adjusting screw on either side of the body where it joins the base and a locking bolt or two under the base (from memory). But with a worn lathe, it can be tricky to get it dead nuts on due to wear on the bed shears in different places and wear in the tailstock body so the barrel moves when the clamp lever is tightened. There is also some adjustment on the key in the base that runs inside the bed ways to take up slack as things wear.

Part of being a good machinist is knowing how to work around on old worn machinery. So measure the taper on your job before you get to the final cut stage and adjust the tailstock accordingly, sometimes on a worn lathe by unclamping it, giving the nose a bit of a bump in the requisite direction and reclamping.

Otherwise you can cheat a bit if the taper is only one or two thou. A quick lick with some emery paper and oil with the job spinning at high rpm will quickly remove a small taper like that if you focus its attention on one end of the job more than the other. Keep measuring as you go along. You can also achieve final size to within a couple tenths of a thou with emery paper and oil, with of course a very nice finish.

Emery papering for final finish and size seems to have gone out of fashion with the advent of modern insert tooling taking heavier cuts and giving good off-tool finish and sizing in the process. But traditionally it was common practice, especially on worn machines.

For severe tapers you could also use a file to correct the taper, such as a 10-inch flat single-cut mill saw file. But be prepared for derisive comments from your peers if you get caught.

[Martin KyteParticipant @martinkyte99762]

I was a little concerned when you said your bolt would fit one slot and not another. They will have been machined with the same cutter so should not vary. T slots do pick up muck and need proper cleaning out to function well. If you have muck in the recess of the T it’s going to push the bolt to one side and jam. Check for the odd burr on the edges (internal and external) too and remove with a fine needle file if needs be. I made some hardwood TEE strips a while back that live in my cross slide TEE slots which keeps all the muck out and eliminates the need for cleaning every time I want to use them.

You are doing OK. Just slow down a bit and check what you are doing. If something doesn’t work try and find out why and then you won’t do it again. We all made mistakes to begin with and most of us still do from time to time, I do especially if I rush.

regards Martin

[HopperParticipant @hopper]

The T head and shank of a T bolt is usually a loose fit in the T slot for this reason. Making things size-for-size means there is no room for machining variations or swarf or dirt. As Mr Harley famously  said to Mr Davidson, a little extra clearance never got in the way.

[JasonBModerator @jasonb]

Might be of interest that the shank of that Myford Tee bolt I showed was 0.372″ and the thread measures 0.370 over the crests. Th enuts are an easy fit but have little play.

Not sure what you are going to be using these two bolts for but you mention them going into reamed holes. Given the problems with the bolts it may be better to allow a bit of room in the holes to allow for not placing them exactly. Better to be able to clock the item in true to the lathe than have it sitting a bit p****d due to the tight fits pushing it out of line.

At least we have a good thread to refer back too the next time the debate that old machines are best comes up. 🙈🙉🙊

[HopperParticipant @hopper]

I’d be very careful cinching down on 3/8″ bolts on a Myford T-slot too. They can exert a lot of force and tear lumps out of the upper part of the T-slot, or distort the topslide and cause it to bind etc. Use the spanner with your hand right up next to the nut and just nip things up.

A lot of original Myford T-bolts and T-nuts etc were 1/4″ BSF or maybe 5/16 at most for this reason.

What will your 3/8 T-bolt/s be holding down?

[Martin KyteParticipant @martinkyte99762]

Wise words Hopper. George Thomas advocated using T nuts and studding which is a lot easier to generate different lengths as needed. Clamping force really doesn’t need to be high at our rates of metal removal.

regards Martin

[Nigel Graham 2Participant @nigelgraham2]

I did describe these bolts as “specials”

I have made two sets of T-bolts, for two projects, and to the published drawings. Since they are basically similar it was sensible to make all five in one series.

They are:-

– One-off 3/8″ short-shank and two-off 5/16″ long-shank, for a ‘Stent’ Tool & Cutter Grinder.

I made the short, fat one from solid as drawn, but for economy the long thin ones in two parts screwed tightly together with ‘Loctite’.  These are not critical as long as they fit, and will not be heavily loaded: rather than ordinary hexagonal nuts on the 5/16″ ones I may make knurled brass thumb-nuts.

– One-off 3/8″ diameter shank X 2″; ditto but 2.3″ long, for a “Dore” (Arnold Throp) designed Slotting Attachment (keyway shaper), a Hemingway Kits product. The drawing specifies 1/2″ A/F heads, and I’d no reason to think this is significantly narrower than the Myford-made bolts.

The longer needs be as close as possible to fit a reamed hole in the aluminium-alloy casting, to act as a pivot when the attachment is angled to cut a keyway in a tapered bore. This was the one that caused all the grief!

The shorter one, I realised later, works through an arcuate slot that the instructions happily suggest filing to shape, so is not so critical. I will probably mill the slot though.

I still wanted these bolts right though – to drawing anyway!

That larger diameter give a larger load-bearing area within the attachment itself.

..

Of course I am aware of the risks of over-tightening a bolt in a small Tee-slot. Though this is reduced to some extent by the slot flanges bearing against the underside of the attachment.

”’

This evening I cleaned the lathe, removed the top-slide for space for a magnetic indicator stand to investigate where all the inaccuracy is coming from, and started to examine it.

No wonder I could not get that bolt parallel.

.

I observed the two T-bolts holding the slide-base are indeed only 5/16″ BSF, but of course the thrust is taken by the large spigot on the base’s underside. The plain part of a milling-cutter shank as a gauge showed the upper parts of the Tee-slots on my lathe are indeed tight on width, and the “gauge” does not easily enter one end of each.

.

First discovery was slight yaw in the saddle, observable by hand and eye, so I adjusted the gib-strip.

Centres in the two ends met by eye, so I processed to more formal tests.

My test-bar is a length of stainless-steel originally part of a very high-precision (or accuracy?) printing-machine so far more reliable than stock bar or anything I could make. It has a spigot on one end and a small centre-drilling in the other, so I held it between an ER collet and tailstock centre.

This proved the tailstock was slightly off-line horizontally but not vertically. I spent a very long time trying to adjust the tailstock but each time I thought it was within 0.002″ over about 8″, so 0.00025″ per inch, it wasn’t.

That bolt finished with an error 4 times that, of 0.001″ over an inch.

It seemed as if that adjustment I thought I had achieved will be the limit, but it was a static test, using the cross-slide and DTI against a stationary bar…..

Further investigation:

Spindle register flange (so the bit outside the chuck): run-out within 0.001″.

Register itself: perhaps 0.001″.

Any centre in the spindle: 0.004″ run-out. Worrying. This on three different centres and different areas of each.

My best chuck, a 4-jaw S.C: 0.004 – 0.005″ run-out on my test-bar. You can see it wobble at fast speed, and the eye is very good at seeing tiny eccentric running.

An old 3-jaw that has had a hard life including collisions: surprisingly not much worse though its jaws are badly worn.

The ER collet: nearly 0.004″ run-out on the bar, although I will revisit this with a much shorter test-piece.

These might suggest a worn register – really the only cure short of expensive treatment is new back-plates for the self-centring chucks. They do not account for the errors on the centres, and I need re-test this with an unused centre.

.

I will have to adjust the tailstock as closely as possible; but in future turning anything between centres may need a temporary live-centre held in any chuck. It will be correct only for that task but can be any short length of suitably-sized mild-steel bar.

If the spindle taper is slightly damaged, there is no visible sign, and both the spindle and tailstock grip tools well enough to need a nudge from the back end with a rod to release them.

.

Incidentally that 4-jaw chuck and its independent-jaw brother are massive things, very heavy and deep, backplate rather than direct mounting so with a huge overhang. I do worry about using them on the Myford lathe despite their being sold for such machines. They’d be better on the Harrison!

[HopperParticipant @hopper]

You are in danger of disappearing down the rabbit hole that entraps many beginners and not-so-beginner hobby lathists. IE trying to set up a lathe with parallel bars and such without really knowing the full story on setting up from basics, it is not a simple job and is another story in itself.

Suffice to say, you are better off doing as I suggested and measuring the taper on your actual job before getting down to finished size and adjust your tailstock to get parallel turning before you need to take the final cuts. Setting up with ground parallel bars and dial indicators does not take into account bed wear in different positions on old lathes and it does not take into account deflection caused by cutting forces. You are trying to set up for a dynamic situation by using a static test. Not good. (And printer rods are nowhere near stout enough. 1″ diameter and up is required.)

The other issue as you have found is that chucks and cheap collets don’t always run true. In fact, any self centring chuck , 3 or 4 jaw, is considered good if it has 2 or 3 thou runout when new. The are not designed or made to run dead true. If you want to hold something dead true you need either top quality collets and holder or an independent 4 jaw chuck as came with the Myford when new. Also, 4-jaw self-centring chucks are usually a woodworking chuck and nowhere near true to metal working standards.

The standard workaround is you machine something like your T bolt in one setting in the chuck. Grip it by the large end where the T will eventually be milled, face and centre drill the far end and hold it steady with the tailstock centre. Then proceed to whittle the 1″ OD down to your required 3/8, checking the taper as you go and adjusting the tailstock to suit. If the key in the base of the tailstock that runs against the vertical shear has any slop, or if the bed ways have wear there, moving the tailstock can change its alignment. Once down to size, you then do your screwcutting and whatever. So it does not matter if your chuck is running a few thou out of true. The shank and thread will be true to each other along the length, all machined in one set up.You can if you wish machine the centre drilled end off the shank by holding the shank closer in in the chuck. Concentricity is not critical there.

Don’t ever expect to be able to machine one end of a bar or shaft or whatever then flip it around in a self-centring chuck and machine the other and have it come out all nice and true. To machine in two settings like that you need top quality collets (cheap ones are unreliable) or an independent 4 jaw chuck to set the job true iwht a dial indicator. Usually, method 1 above is quicker and easier.

If your collets are running out, it may be the taper in the lathe spindle that is out, if the collet holder is MT2 mounted there. Cure is to take a skim cut down the taper and true it up in situ. BUT your headstock bearings first need to be adjusted so there is 0 detectable play in the spindle when yanked up and down with a long bar held in the chuck.

New backplate s are probably not needed to fix your chuck runout. Back plates are finish machined in situ — so you can do that with your existing ones. But it is probably the self centreing chucks themselves at fault, not the back plates.

My advice would be to get a good four jaw independent chuck and learn how to use it. Setting a job true is a minute or two’s work if you are methodical about it. But before doing that, focus on doing jobs in one setting in the existing self centreing chucks as described above.

 

[JasonBModerator @jasonb]

The 4-jaw self centering woodworking chucks that I have are a different cattle of fish to the ones used for engineering, go look at a Bison or Pratt Catalogue or even Sanou

Also why do people say that a 3-jaw is only good for 2-3 thou when new, I only have a “standard” quality Bison 3-jaw which from their data is good for 0.03mm max runout but actually gives half that so about half a thou.

A similar size  standard quality SC 4-jaw is also given as 0.03max run out. Good enough for ironwood!

 

[JasonB]

On JasonB Said:

 

Also why do people say that a 3-jaw is only good for 2-3 thou when new.

 

Hi Jason,

Maybe because the chucks have been strained, or the jaws are bell-mouthed from gripping work right on the end of the jaws.

The humble second hand Unimat Lever 3 Jaw Chuck holds my work closer than that. The Compact 5 has a TOS chuck, “standard quality” and this holds work under a “thou”, (0,02 mm measured). The test certificate that came with the chuck quotes 0.03 mm for the same size work.

Regards

Gray,

[JasonBModerator @jasonb]

Gray, Hopper said “FROM NEW” so hopefully not supplied with strain or wear.

[Dave HalfordParticipant @davehalford22513]

Nigel,

If you have a DTi on the cross slide you are measuring both runout and saddle angle at the same time.

What does the Harrison give with the same test bar?

[Howard LewisParticipant @howardlewis46836]

Stray comments

The reference to “Full Form Inserts” immediately made me think of a 60 degree Metric thread!

Whitworth form threads are not flat topped (truncated like Metric or Unified) but have rounded crests.

Being idle and a coward, if there is a Tap or Die available, they are used instead of screwcutting the thread. And less hassle with withdrawing the tool for winding back and ren engaging the half nuts (Although my practice is to withdraw the tool and reverse with the half nuts still engaged, but far enough to take out backlash on the Leadscrew when running forwards again)

Plus it avoids the risk of errors in selcting change wheels, (Not unknown!)

A Tap or Die is likely to cut a more accurate thread form than I can produce, even using a cutter grinder.

Dies are set by tightening onto a good quaility bolt (Allenn screw, where possible) to give the right “feel”.

The Die is carried in a sliding Tailstock Die holder. There is a small clearance between the Holder and the Arbor, so that the Die can self center on the work. Always with a lead in chamfer, preferably of thread depth.

Lubrication is either Rocol RTD or old fashioned bacon fat, used on every occasion.

If supporting the work with a tailstock centre, unless the Tailstock is correctly aligned, the lathe will taper turn.  (Offsetting the Tailstock is a standard method for turning longer tapers.)

HTH

Howard

[Howard Lewis]

On Howard Lewis Said:

The reference to “Full Form Inserts” immediately made me think of a 60 degree Metric thread!

Whitworth form threads are not flat topped (truncated like Metric or Unified) but have rounded crests.

The 3/4″ BSW thread shown in a previous post was screwcut using a full form insert; the roots and crests are formed correctly. Here are some inserts:

On the left is a partial form insert; note the ‘sharp’ point that will form the root. The middle insert is a full form Whitworth, with rounded root and crest. On the right is a full form ISO metric insert, note the flat to form the crest.

Screwcutting is fairly simple for me as my lathe has a full screwcutting gearbox. For most common threads it is simply a matter of setting a few knobs. I have extra change gears if needed, but thus far have never needed to cut a DP or module “thread”. I also have an Ainjest high speed threading unit so, for imperial threads, I don’t need to worry about the thread dial indicator. On the few occasions that I need to screwcut a metric thread I simply set the QC gearbox and leave the half nuts engaged.

Andrew

[JasonB]

On JasonB Said:

Gray, Hopper said “FROM NEW” so hopefully not supplied with strain or wear.

Sorry Jason,

I read your post as a generalisation. No I too would not expect such results from a new chuck.

Regards

Gray,

[Nigel Graham 2Participant @nigelgraham2]

I am now testing the lathe by turning a length of substantial bar.

I tested the spindle run out on various centres again and seemed to obtain better results – down to about 0.002″.

‘

Hopper – I was not using “printer rods” if you mean mined from a scrapped domestic computer-printer. My test-piece had been made for a very high-grade industrial electronics-manufacturing printer, and is solid stainless-steel bar 22mm diameter.

.

Dave – I have not investigated the Harrison yet. I want to sort the machines out one at a time and am working on the Myford first.

If as you suggest I am measuring a combination of tailstock adjustment and saddle wear I cannot hope to obtain accurate results.

I may be able to test the tailstock setting by an indicator in a chuck or faceplate pointing to the surface of a good-quality centre in the tailstock or the tailstock barrel itself, rather like centring a rotary-table on a milling-machine. That assumes being able to read the indicator properly when it’s facing the wrong way: a possible source of more error. Still, it may be the quickest and least inaccurate method.

I think it’s going to be get the thing as near as I can and accept it’s not RR Aerospace quality. I can adjust the tailstock setting but I cannot do anything about uneven wear in the lathe generally.

.

That 3-jaw chuck I mentioned is badly worn, with years of hard and sometimes probably bad use since it came to me nearly fifty-four years ago. I had thought or re-grinding the jaws but it’s not worth it as the poor old thing is too worn generally. A bit like its owner. I’d be better buying a new one for Sunday Best – a bit like Grandma’s front room. I can still use the old ‘un where its internal accuracy is not so critical, which is for a good deal of basic items like collars, short pins, ordinary studs, pipe-fittings, etc.

[Nigel McBurney 1Participant @nigelmcburney1]

The taper in the tailstock should never be used to drive a die holder,The die holder can run on a parallel shank held in the tailstock chuck ,the dieholder just slides on the shank and the tailstock should be clamped tight on the bed, a tommy bar placed in a crosshole in the die holder,now on 3/8 bsf its best to lock the lathe spindle ,easy on a super 7, then pull the die holder round with a long tommy bar, I would not use power to drive the spindle. Nowadays with the poor carbon dies I would use hss ground thread dies,if there is trouble centering the die in the holder then the holder has too much clearance,its easy to make die holder with less clearance. On a larger lathepower can be used,same set up with a free sliding dieholder and tommy bar, to take the torque on the tommy bar, let it rest on a length of square mild steel in the tool holder,dont use a tool bit a slip with ones fingers can get cut on the tool bit. If I was cutting 3/8 bsf I would o it with a sliding die holder in my Colchester,if a larger thread I would screwcut using machine chasers ,which have 5 degrees top rake ground on them.I do have a lot uk industrial quality carbide tip tooling which luckily cost me a few pounds but rarely use them

[Nigel Graham 2Participant @nigelgraham2]

I was not expecting to drive the dire-holder from the tailstock. The problem was that the die-holder shank would grip the tailstock properly. I found eventually the cause was a tiny burr where the shank had slipped at some time past.

.

All that grief with that T-bolt came because the tailstock was slightly off-centre.

I spent several hours and wore out a wind-farm or two today verifying this then trying to correct it.

Just as I thought I had done so, with an error of about 0.004″ on diameter over 4″ of turned length which I thought would be as good as I could hope for, I managed to lose it.

The adjustment is not easy when you can’t walk round the end of the lathe, you need loosen the clamp that holds the unit to the bed to adjust it then that looseness makes it impossible to set the adjustment to the DTI clamped to a magnetic stand “stuck” to the chip-tray…

Then I notice a tiny scale on the end of the tailstock, showing about 1/32″ off-centre. So I set it back to the 0 point, took another test cut…. 0.02″ error on diameter so 0.010″ out.

I gave up, locked up and came indoors for tea.

.

A thought occurred to me… Align the tailstock to a gauge!

I had measured the tailstock barrel’s o.d. as 0.999″.

So, 1) Put the DTI etc. away.They are leading me astray.

2) Bore a convenient bit of steel bar to a close fit on something accurate enough to act as a one-inch plug-gauge. Leave this newly-made ring gauge in the chuck.

3) Remove the top-slide and rear tool-post, for access.

4) Bring the tailstock up as closely as possible – the saddle will still be in the way.

– I don’t know if it’s possible to slip the tailstock off the bed and re-install it ahead of the saddle. That would be a better approach.

5) Adjust the off-set until the barrel, with a spot of oil on it, will wind very gently forwards into the ring without catching the edge of the hole.

6) Lock the adjusters to each other, and if the ring-gauge test still works re-fit the top-slide & tool-post.

7) Try to find the ring-gauge a safe home for future general-purpose turning use.

8) Have another cup of tea.

 

[HopperParticipant @hopper]

Well done. A smart solution. The proof will be in the pudding when you turn a test piece and measure the taper.

Often when adjusting those tailstock offset adjuster screws you can do it in situ without a dial indicator, just  by “feel” until you get it right. For that last thou or two you can sometimes just tighten the one without loosening the other first. Seems to often give it just that last tiny nudge.

Another thing to be aware of is that piece of 3/8″ square key steel held to the bottom of the tailstock base that bears in the lathe bed vertical surface can be adjusted to take up side-to-side movement.You slacken off the allen screws that hold it and set the tailstock on the least worn part of the bed at the far right then while pushing the tailstock one way on the bed and the key steel the other, tighten up the screws.

You can even go a step further then and use a feel gauge to measure wear on the bed between that key steel and the vertical shear of the bed down in the high wear area nearer the chuck. At least then you know how much potential movement you have when setting the tailstock in that zone.

And another thing to be aware of is that wear on the extendable quill/barrel can make a thou or two of difference when you tighten up the clamp, so make all your measurements with it clamped consistently tight.

Have fun.

[Nigel Graham 2]

On Nigel Graham 2 Said:

I was not expecting to drive the dire-holder from the tailstock. The problem was that the die-holder shank would grip the tailstock properly. I found eventually the cause was a tiny burr where the shank had slipped at some time past.

.

All that grief with that T-bolt came because the tailstock was slightly off-centre.

I spent several hours and wore out a wind-farm or two today verifying this then trying to correct it.

Just as I thought I had done so, with an error of about 0.004″ on diameter over 4″ of turned length which I thought would be as good as I could hope for, I managed to lose it.

The adjustment is not easy when you can’t walk round the end of the lathe, you need loosen the clamp that holds the unit to the bed to adjust it then that looseness makes it impossible to set the adjustment to the DTI clamped to a magnetic stand “stuck” to the chip-tray…

Then I notice a tiny scale on the end of the tailstock, showing about 1/32″ off-centre. So I set it back to the 0 point, took another test cut…. 0.02″ error on diameter so 0.010″ out.

I gave up, locked up and came indoors for tea.

.

A thought occurred to me… Align the tailstock to a gauge!

I had measured the tailstock barrel’s o.d. as 0.999″.

So, 1) Put the DTI etc. away.They are leading me astray.

2) Bore a convenient bit of steel bar to a close fit on something accurate enough to act as a one-inch plug-gauge. Leave this newly-made ring gauge in the chuck.

3) Remove the top-slide and rear tool-post, for access.

4) Bring the tailstock up as closely as possible – the saddle will still be in the way.

– I don’t know if it’s possible to slip the tailstock off the bed and re-install it ahead of the saddle. That would be a better approach.

5) Adjust the off-set until the barrel, with a spot of oil on it, will wind very gently forwards into the ring without catching the edge of the hole.

6) Lock the adjusters to each other, and if the ring-gauge test still works re-fit the top-slide & tool-post.

7) Try to find the ring-gauge a safe home for future general-purpose turning use.

8) Have another cup of tea.

 

The problem with the ring gauge approach is that an assumption has been made that the Tailstock Barrel outside diameter is concentric with the Morse Taper. I have known instances where this is not the case. Also it assumes that the Tailstock Barrel centre-line is at the same height as the Spindle centre-line.

It would be far easier to mount a DTI in the lathe chuck and indicate the nose of the Morse taper bore, by revolving the chuck, By Hand. It only needs to be the same reading in the horizontal plane. Wear on the tailstock base will give an error in the vertical plane.

Regards

Gray,

[Howard LewisParticipant @howardlewis46836]

Basic assumptions, if you are looking for precision.

The tool is on centre height.

The Cross Slide gibs are correctly set, and the Cross Slide is locked while turning.

If the machine is turning a taper, how about checking some basics, such as twist in the bed, and  Headstock alignement for a start?

One assumes that the bed (Hopefully, not a cantilever bed) is not twisted?

If you have any doubts, check using “Rollie’s Dad’s Method”, and correct. The method is that advocated by Myford, among others.

Ian Bradley details the method in “The Amateur’s Workshop” and “The Myford Series 7 Manual” . This will tell which side of the tailstock foot needs to be shimmed.

(On a purely personal note, I prefer mounting the lathe on riser blocks, so that screw adjustment can be used for greater precision.)

Ideally, use an Alignment bar with a Morse taper, to locate in the Spindle.

Do NOT use a chuck. Because of all the clearances necessary to allow the jaws to move, a 3 jaw will not be able to hold work concentric, at any setting.  (Backplate to spindle register, Backplate to chuck body, scroll to body, jaws to scroll, jaws to chuck body). If you allow a thou for each (Quite optimistic) the errors will soon add up to 0.005″ or more!

Once the Headstock has been aligned parallel to the movement of the Saddle (Be wary of bed wear across the bed!)

I had a friend whose recently acquired ML4, turned tapers.  Someone had cross threaded one of the four 1/4BSF studs that secured the Headstock. All was well until the nut on that was tightened. The cure was to drill and tap 3/8BSF and bush back to 1/4 BSF.  That allowed the Headstock to be relaigned and secured, to turn parallel.

THEN he could use the Alignment bar to align the Tailstock, between centres. (Ideally trim up a soft centre in the Headstock before using the bar between centres. If any any doubts exist about the hard centre for the tailsock, get a blank arbor, set it in the headstock and turn that into a centre, for use in the tailstock.

Unfortunately, if you have to use sleeves to adapt the Tailstock size taper to the Headstock, that might introduce a concentricity error; hopefully very small, but less than the one that you are currently trying to eradicate.

When you finally have everything aligned (AND you may have to reiterate as each error is reduced or eliminated) yoiu should be more confident of being able to turn parallel, and more precisely.

Howard

[Grindstone CowboyParticipant @grindstonecowboy]

Random thought just occurred to me – and it’s probably ridiculous – but could some form of cutting tool, set at 30 degrees to give a sixty degree included angle, be held in the chuck and used to form a center held in the tailstock. Would this ensure concentricity?

Going to lie down now… Nurse, the screens!

Rob

[Nigel Graham 2Participant @nigelgraham2]

I spent today going through those various tests as best I could.

Hopper –

I found the key under the tailstock. There was about 0.003″ sideways play, by feeler-gauge, to remove.

My lathe is an earlier one than that illustrating the Myford handbook, as the adjustment screws for the key are underneath the unit and not obvious. I found them only when I had the tailstock upside down, held very gently in the bench-vice, and was faffing about with vernier calipers and feeler-gauges.

Their location means you have to lift the unit from the lathe to set it, so there’s quite a bit of trial-and-errors, but I managed to reach a point where the unit will park rather tightly in the unworn end of the bed but has no shake play when moved up as far as the saddle lets it.

””’

Howard –

Having re-set the tailstock’s’ fit between the shears I did indeed adjust all the gibs: compound slide, cross-slide and saddle. They were not too far out, but this did reveal how badly worn are the slide screws or nuts, or both.

The bed is the sort with its feet at the corners, and it is on riser-blocks. It stands on a proper Myford cabinet, but its previous owner had bolted it to a massive wooden bench via a hefty cast-iron tray of unknown purpose and having no machined surfaces.

I thought I had set it correctly when I bought the cabinet and risers but might not have done quite right, so I tested it with that turning trial in the book. This did suggest some twist and I put that right, though will probably have to revisit it.

I have a Morse-taper test bar … somewhere. I have not yet managed to find it!

That hefty stainless-steel bar I was using has a centre-hole only at one end, so I’d need set it to run true in the independent 4-jaw chuck, and set the steady at the chuck end, before centre-drilling its other end.

My turning tests yesterday were on a seven-inch length of 40mm dia steel that had started to become a connecting-rod for my steam-wagon before yet another “design” change, so it had a centre-drilling in both ends. I had to improvise somewhat to mount it between centres on the ML7, because my drive-dogs are all too small or much too big and the wrong shape!

.

I still need centre the tailstock correctly, but after today’s work I will examine whether setting it to its own scale will have largely removed the error. With its base having 0.003″ lateral play between the shears, I have quite unknowingly clamping it at a tiny angle, enough to move the tip of the centre in the extended barrel quite a bit further across the lathe.

My idea of turning a tailstock gauge does depend on leaving it in the chuck . Even if the chuck is is in poor condition the ring bore will be on axis while it is still in its manufacturing position.

I realise I can only adjust the tailstock sideways, not vertically. If its centre has “sunk” there is nothing I can do about that.

…

So I am making some progress but not out of the woods (or swarf thickets) yet….

The sample for the Myford-book turning test is not wasted. I turned the other end to be a plug-gauge for further examining the state of the EW lathe’s headstocks.

 

[HopperParticipant @hopper]

Good stuff. Sounds like you are well on the way to getting it sorted. I am a big fan of the “Myford method” of aligning your lathe by turning a 4 to 6″ test piece and measuring it. Particularly for older lathes with a bit of wear in them that static measuring does not always account for.

Don’t worry about slack in the cross slide and top slide feed screws and nuts. Does not affect lathe performance significantly as cutting forces push the screw up against the nut solidly. Just remember to allow for the backlash when resetting a cross slide dial etc and wind it past the end of the backlash and then back again.

Likewise don’t worry too much about vertical alignment of the tailstock. It does not affect the end result anywhere near as much as the horizontal alignment and many lathes are made with the tailstock a thou or two high so they come into alignment as the tailstock base wears over the years.

One more thing to be aware of when setting tailstock alignment, besides having it all clamped up solid on both the bed and barrel, is that the amount the barrel is extended from the tailstock body can change your alignment if the barrel is not being held perfectly parallel to the bed axis due to wear etc. So best to pic a point about halfway extended, do all your tests there, then use that point when turning critical jobs. (There is a whole article on Myford tailstock alignment in MEW a few years back. Search for the Pete Barker byline. Another article outlines bed alignment after mounting on a new bench.)

If you don’t have a drive dog to fit your 40mm bar, try this: Put a bit of half inch, or whatever, bar in the three jaw chuck and turn a 60 degree point on it to make a live centre that runs dead true as long as not disturbed. Then drill and tap a hole in the end of the bar near the OD. Then screw a bolt into that hole and let it engage with one of the chuck jaws to provide drive. A 1/4″ Allen head bolt should do the job for taking light test cuts. (A rubber band or zip tie around the bolt to the jaw can be helpful to stop it flapping about when not under cutting load. )

[Howard LewisParticipant @howardlewis46836]

You are grasdually eliminating errors and play, so things can only get better.

As Hopper says, the effect of vertical alignment of the tailstock is FAR less than that of horizontal.

(More akin to having the tool set a few thou above or below centre height.)

Getting rid of the twist, will definitely improve out of parallel.

As Hopper says, turning a 60 degree centre on a bar in a 3 jaw chuck, and not disturbing it, will eliminate any errors in the chuck and be as good if not better than a Morse taper centre in the spindle.

Every small improvement will produce a much greater benefit.

Keep up the good work! The sows’s ear is becoming more silky every time!

Howard

 

[Author]

Posts