Super7 Spindle run out opinions

[chrisemParticipant @chrisem]

Hi all.
New on here, so apologies if this has been asked 100 times before.
I have owned my father’s Super 7 for at least 30 years. He had had it a good 10-15 before that. I have never seriously looked at setting it up beyond throwing a bubble level on the bed either end & leaving it at that.
I recently set to with some centres, a precision test bar & a morse taper test bar, and these are my results: spindle nose OD – 10 microns. Shoulder 6 microns. Taper 20 microns.
100mm Morse taper test bar: spindle end – 50 microns. 100mm out – 150 microns.
Centres in head & tailstock, 200mm test bar: Indicator on top of bar – headstock end 0, tailstock end +30 microns. Indicator on side of bar: headstock end 0, tailstock end + 25 microns.
My feelings are between centres & on the nose it’s not too bad. Run out on the headstock morse taper is not good. It looks in clean condition with no dents.
So really after some opinions on this. I was going to set up the lathe ‘properly’ using the MT test bar, but I’m not going there with that amount of run out. Spindle bearings are all good and adjusted as per the manual.

Regards: Chris.

[SillyOldDufferModerator @sillyoldduffer]

I’d expect a Super7 to do better than “100mm Morse taper test bar: spindle end – 50 microns. 100mm out – 150 microns.”   150 microns in real money is 0.15mm.  In comparison the Chinese 3 jaw chuck that came with my Chinese lathe managed 0.02mm when new.

Don’t panic though. As Myford’s have bendy beds, it’s important to make sure they’re straight on the bench.    Often misleadingly called ‘levelling’, but think ‘alignment’.   I suggest the first thing to try is Rollie’s Dad.  Then measure again.

Be careful with the measuring too.  All too easy to be confused by getting sub-millimetre measurements wrong, and the technique has to be practised.  Don’t damn the lathe on the basis of measurements; if it delivers OK on real work, the operator may be the problem,  He causes most of the trouble in my workshop!

Dave

[Martin of WickParticipant @martinofwick]

Ignore the test bar for the time being and set the lathe up as best you can by  taking test cuts at each end on a piece of 1″ x 6″ bar adjusting out any twist using the bed setting screws until the cuts at each end are the same dimensions.

Then if you need to, revisit headstock to tailstock parallelism.

If you have to use a test bar, take multiple readings rotating the bar 90 degrees or so.

[Charles LamontParticipant @charleslamont71117]

I recently set to with some centres, a precision test bar & a morse taper test bar, and these are my results: spindle nose OD – 10 microns. Shoulder 6 microns. Taper 20 microns.
100mm Morse taper test bar: spindle end – 50 microns. 100mm out – 150 microns.

SoD and MoW, I think Chris is talking here about runout on rotating the spindle. This has nothing to do with any possible bed twist. I would not be happy if I had spindle runout that bad.

Chris, these tests were rotating the spindle by hand, with the belt slack?

You may want the spindle taper adjusted a bit tighter than the manual says. With mine, if I try to use top speed from cold it will sometimes bind. I have to warm it up for a minute or two at a lower speed. I rang Myford (Beeston) for advice a good number of years ago and was told “that sounds about right”. I was also told to tighten the nut at the back of the spindle as tight as I could grip it by hand to ensure enough preload on the angular contact bearings at the back.

[peak4Participant @peak4]

Hello and welcome, whereabouts roughly are you; there might be a local member happy to assist

A picture is worth a thousand words.
It would help others to assist you, if we knew exactly what you’re measuring and how.
The very first thing to  establish, is whether the Morse taper in the spindle is damaged, using engineers blue, or even a bit of lipstick.
Only then can you trust any readings from a test bar inserted into the taper.

Bill

p.s. I do have a Myford, test bar, DTIs etc

[Neil LickfoldParticipant @neillickfold44316]

I assume that you have the manual, and the correct procedures to adjust the spindle position correctly. The adjustment is done on the pair of castle nuts in the back. It will pay to adjust it all forward, some 1/4 turn or a little more. Then loosen the screw of the shaft collar at the back, and see if it can tighten by hand any amount. re tighten the locking screw. This is setting the preload on the thrust bearing pair in the back. Then loosen  the LH or castle nut rotating it so it un winds. Then rotate the rh castle nut , pulling the shaft assembly back into the taper. The belt lever tension, must be loose like when changing the belt position. Keep slowly pulling in the assembly until the taper binds on the bronze bushing. Then you wind in the LH castle nut and tighten it together. As the LH one comes in, the small backlash on the rh nut is taken up and then the spindle will be free to rotate. You can place a DTI on  the front of the spindle. From being tight and only just able to be turned by hand, to a free running state, is in the 10um range to a max of 14um, or 4 to 5 tenths of a thou movement from tight. Any more and the spindle will be loose and could cause wear in  the bearing or on the shaft. It rides on a thin oil film. Having the belts too tight is also another cause of wear .  The above setting will give about 3um of clearance for the oil film that it needs.

The numbers in the error of the spindle runout seem to be unusually large that you are reporting.

Another check, is to place a bar in the chuck, preferably a 4 jaw chuck. 1 inch or so Al or free cutting steel and sticking out of the chuck about 100mm or so. The longer the test length, the sharper the cutter needs to be. 4d of material is a good stable length. You need a very sharp tool and very fine cuts. I relieve the testing bar in the centre section. So leave a piece 10mm long at the front and similar at the chuck end. Take a cut on the front first. Measure it. Then set your dti on it, and see how true it runs. Is it round or some other shape? Does it measure the same in many different positions . Then take a cut at the chuck end on the same setting of the cross slide. How do they compare? Are they round as each other? The difference is size is the error of the bed position in most cases. Very seldom, it is the headstock that is not correctly aligned to the lathe bed. But at this stage we assume that the headstock is correct. You either adjust the headstock end or the tailstock end adjusting the riser feet. If the end of the bar is smaller than the chuck end, then the bed will need to be twisted a very small amount towards the operator at the tailstock end. Operator side adjuster is wound down a  little and  the behind adjuster is wound up a little. Keep adjusting until the test cuts are even in size.

I prefer to use the low pressure finger style indicators when setting up machine tools. You will notice that if you have 2 DTI’s and stands, you place the 1st DTI onto the bar and set to 0 . Then set the other DTI on the opposite side and have the same load from zero on each, you will see the 1st indicator no longer be zero.  Then zero both, and rotate the bar. It will show any sideways irregularity or roundness errors.

On my S7 from the early 1970’s pre power cross slides came out, I recently replaced the thrust bearings, as well as made a new Taper bronze bush, and setup on a big lathe with a very precise centre, recut the front taper of the spindle back to being round again. I also corrected the small runout of the back area that the angular contact bearings sit on. The movement of the housing means that the inner of the tapered bush is required to be scraped back into true with the taper of the spindle. And then the scraping for the oil film will need to be done.  A big job for sure. I can see why they stopped selling just the bush on it’s own.

The roundness error in my spindle was around 0.01mm.  Now it is back to the original runout of around 1um or so if adjusted tighter than normal. But at that setting , can not drill onto stock, that 0.006mm form tight on the taper. It needs to be looser to be able to drill into stock or take reasonable cuts. At around 3um or 0.014 from tight spindle position, drilling is just fine. Roundness should be better than 4um on the Myford S7 setup of it is not worn.

[DiogenesParticipant @diogenes]

All your readings are pretty good – I’d suspect the test-bar, you must test that first, unless you are sure of an immaculate provenance.

Parallelism, concentricity, and roundness, paying attention to the tapered part and it’s relationship to the straight part.

[chrisemParticipant @chrisem]

Many thanks for all the replies.

Yes – my first test was with a high quality 4” or 100mm #2MT test bar in the spindle nose. I had the belts loose and rotated the spindle carefully by hand. Up near the headstock end I pushed with thumb & pulled with finger on the stationary bar, to make sure I had no major visible side play. DTI is Mitutoyo finger type.
In previous moves, it’s just been bolted down on a builders level set bench. About 13 years ago it was moved again and I put an engineers bubble level on the bed out of curiosity. There is approx 1 division of twist on the level. I’ve always just used it – mostly with a 3 jaw for bits and bobs, nothing that needed more than a pair of callipers to measure,  & it turns OK. If I want to do anything serious, I have a whole machine shop at work.Way back when my father used it, he went through the spindle setup, made a new cross-slide nut, then screw cut a new cross-slide screw. He had no fancy DTI or levels. He just tweaked about until the lathe cut true. He was a very skilled man.
Good point made about blueing the taper – I will do that first. I have neglected this little machine, and have a project in mind. So decided to kit it up with a few accessories that my father could never afford – a quick change tool post etc, and give it a bit of a birthday. Felt bed wiper changed & a pair of new belts fitted. I was going to buy some collets – but that’s on hold until I get this sorted. Spindle checked and adjusted as per the manual & googling the process described above. I’m based in West Sussex & many thanks for the offer of help. Will take a photo later.
Regards: Chris.

[Charles Lamont]

On Charles Lamont Said:

I recently set to with some centres, a precision test bar & a morse taper test bar, and these are my results: spindle nose OD – 10 microns. Shoulder 6 microns. Taper 20 microns.
100mm Morse taper test bar: spindle end – 50 microns. 100mm out – 150 microns.

SoD and MoW, I think Chris is talking here about runout on rotating the spindle. This has nothing to do with any possible bed twist…

Charles might well be right!   In part, I posted ‘Be careful with the measuring too’ partly because I wasn’t clear what Chris was actually doing.   Judging by other responses, the forum doesn’t have a consensus on what the numbers mean this morning.   That makes showering Chris with fix suggestions a little dangerous, unless they are of the ‘do no harm’ variety.   I suggested Rollie’s Dad for that reason.

Run-out is one of those terms that’s used loosely, probably because the symptoms are similar.  The far end of a stiff test rod is seen to be not running true – wobble.  And the lathe cuts tapers, not straight.  I  believe ‘run-out’ can mean: ‘error caused by a loose spindle’ AND ‘error caused by a misaligned spindle’.   In an older lathe both are likely.

We can ask the horses mouth!   Chris: in this topic, what do you mean by run-out?

Also useful for Chris to post photographs of each test so we can see exactly what’s what.

I’d also like to know how exactly the lathe is misbehaving on the job.  If it’s not cutting properly, use measurements to pin-down the problem, not the other way round.  Too easy to chase non-existent faults based on duff measurements.    For example, Diogenes mentions the need to be sure about the test-bar; he’s right – second-hand bars are likely to have been bashed about, and some new ones are ‘too cheap’…

Dave

[Martin of WickParticipant @martinofwick]

SoD and MoW, I think Chris is talking here about runout on rotating the spindle. This has nothing to do with any possible bed twist. I would not be happy if I had spindle runout that bad.

True, but it is not clear what the op is measuring or how.  Frankly, without fully understanding what has been measured  it is difficult to be conclusive. But this is suspect…. if we assume that op measured TIR at the nose as .0004in and yet for the MT bar close to the  nose it is double that at the nose, then I would suspect the taper or the test bar or the fitting or the measurement first. I always find in these circumstances it is beneficial to take multiple measurements, fitting and re fitting the test bar.

Unless the spindle is rattling around in its cone bearing like a pea in a cocoa tin, I can hardly believe the measured runout 100mm out from the nose, that is why I may have misread the – 150u as being minus 0.006in as a lateral displacement (ie the more common problem).

Would still recommend making a test cut as described above, if problems are apparent then pushing and pulling on the chuck with a dti in place to measure movement will highlight spindle to bearing fitting issues.

 

 

[chrisemParticipant @chrisem]

Good morning & thank you.

Spindle run out – I am describing the error I see on the dial indicator running on a test bar when the headstock spindle is rotated.
You are dead right about the MT test bar – I didn’t blue the taper, & I didn’t pop the test bar between centres. I did buy it new & checked a cylindrical grinder at work with it, so I think it’s ok.
Going looking for problems – yes – guilty. It turns OK. Little bit of taper towards the chuck – component smaller at headstock, light cut, sharp tool. I have my relieved test piece ready for that little experiment in the future.
At the moment problem #1. It’s got a wobbly hole & I want to use collets.
Regards: Chris.

[chrisemParticipant @chrisem]

Hi Martin.

I was a bit surprised myself. Sorry about my mixture of units & terminology – it’s just what I’m used to working in. A DTI up the bore about 1/2” from the front showed about 8 tenths TIR. 0.0008” Did the push / pull on the bar and spindle, as I thought as you did, that  either the test bar must be not seated or the spindle had play

Regards: Chris.

[Martin of WickParticipant @martinofwick]

At the moment problem #1. It’s got a wobbly hole & I want to use collets.
Regards: Chris.

So if your test cut is OK and you have checked for movement of spindle relative to bearing and that is OK are you saying you suspect the Morse taper bore inside the spindle is the problem?  I would find it hard to believe the spindle taper could be badly misaligned unless it had suffered some kind of gross damage.

suggest make a wood 2MT shaped plug, slot and put 600/800 grit silcarbide paper in to thoroughly clean internal taper

or

just purchase an 2MT collet of your choice and see if you can do what you need to do with without issue (MT2 collets don’t cost much). If it turns out to be a disaster, you can proceed from there to determine remedials.

[chrisemParticipant @chrisem]

<p style=”text-align: left;”>👍🏻 Good suggestion.</p>
Further investigation needed.

Regards: Chris.

[chrisemParticipant @chrisem]

SK 75185 in all its glory.

I need to reduce the photo size.
Fixed centre in headstock and tailstock. MT2 test bar between centres after cleaning them. Extreme ends of taper & of test bar – 0. Clock doesn’t move. Headstock end of parallel bar where it transitions to a taper – 0.0005” run out.

I suppose with spelling like that, I might have known. Will invest in a proper one that comes with a cert.
Blue being at work, I did the pencil test, and it does show the taper being slightly tight at the back, but that relies on the mandrel having been ground correctly.
I think as suggested, buy a collet & try that. Go from there. Do the test piece turning & shim the feet to correct any error. Lastly align the tail stock.

Regards: Chris.

[DiogenesParticipant @diogenes]

Under the circumstances you describe and with your stated objective in mind, I’d be inclined to stick some work in a collet with say an inch of stick-out, and take a light skim down it.

Mic. both ends and see whether the results are acceptable or not.

 

[peak4Participant @peak4]

Before you invest in a new test bar, why not see if there’s anything wrong with this one first?

Regarding the lathe, I would suggest that the initial things to check are the chuck register, on both the round and left hand faces, and the concentricity and condition of the bore in the spindle.
Try a variety of Morse taper tooling, as well as the test bar, and/or centres, plus some blue/felt tip pen/lipstick; also check the male Morse tapers in other sockets, such as extenders, tailstock etc.
At the moment, no-one knows if the bore has been damaged, got a bit of swarf embedded, or suffering from over enthusiastic cleaning. External eccentricity of the chuck register would certainly need further investigation.
Also just to confirm, what do you mean by microns, and how are you measuring micron accuracy?

Checking the tailstock end of the test bar for concentricity, whilst supported by a tailstock centre, isn’t checking the lathe; it’s just checking the concentricity of the end of the test bar compared to its drilled centre.

There’s a bit of a discussion on test bars here, but unless used with care and the correct methodology, it’s easy to be mislead by the readings; I should stress I have no formal training in metrology.

Would you chance it – bluefox tools MT4 alignment bar

[chrisemParticipant @chrisem]

Thanks Diogenes.
I will do exactly that. If I can put clean material up in a collet sticking out an inch or so and get a thou or two runout, I’d be very pleased with that. I just have no idea how good they are as standard.

Peak4 – I’d say cautiously yes, the test bar has an error.
The chuck register OD has about 0.0005” run out, and the face 0.0002”

A quick check using stripes of lead pencil in place of blue, shows the taper is tight at the back. But that is with an unchecked and possibly dud arbor. If the manufacturer didn’t grind the parallel bit correctly, then who knows if the taper is correct.

Microns – as in 5 microns is two tenths or 0.005mm is 0.0002” ish. My DTI is marked in 0.002mm divisions. Measuring to micron accuracy – no you won’t be doing that in a domestic garage. At work I can measure microns & even sub micron with a bit of care. Temperature really comes into it – but we are going a bit off piste here.

Checking the test bar between centres – yes exactly, nothing to do with the lathe, I wanted to see if the test bar was straight and ran concentric – it’s not & it doesn’t. The end of the bar and both ends of the taper are spot on, but the parallel part of the bar adjacent to the taper runs out. This is where you would be taking a concentricity reading next to the spindle nose.
I had a play trying to remember how to sharpen tools & screw cut. A quick and dirty 32 tpi on 0.470 bar turned out OK. I will have to delve into the beginners section. I really am very rusty.

Regards: Chris.

 

[peak4Participant @peak4]

p.s. from my previous post, the embedded link in one of Ketan’s replies no longer works, as it’s to the old forum.
This is the thread, which you may also find relevant.

Looking for lathe mandrel test bars.

Bill

[chrisemParticipant @chrisem]

Thanks Bill.

The link that worked had another to Testing Machine Tools. A publication I have been after for years. So many thanks for that!

Regards: Chris.

[Author]

Posts