Corbetts Little Jim Lathe restoration – newbie needs advice

[Andy Thompson 3Participant @andythompson3]

Next I needed a rough headstock and tailstock alignment. I do not yet have a clock – that is another discussion.

The headstock is held by three bolts and it does not have an adjustment mechanism – just loose bolt holes. I stuck a long bar in the chuck and a square bar in the topslide. Roughly adjusted the headstock by running carriage up and down the bed. Rotate bar and look for same gap.

Tailstock is held by a pin and one bolt. Again that can swivel and I did my best to point it at the headstock taper.

I know all this needs redone by Rollies Dad.

I sharpened a knife tool on my £15 grinder. See pic – don't laugh – I did this for woodturning chisels and it works fine. Rough grind on stone followed by belt is definitely the way to go

Phew,

So after all the above I had a working lathe. I was going to post an impressive video of making swarf – well its my first swarf so I am impressed. But not sure of where/how to host video. See picture – the finish is not great and that was going to be my next question.

I intend to remount it all on a better worktop and do the alignment etc. I will get thrust bearings for the slides. What else I look at.

Cheers

[Andy CarlsonParticipant @andycarlson18141]

Looks like you have made great progress with resolving your issues – excellent ingenuity so well done!

Used DTIs crop up on fleabay from time to time… or…

Given your proven levels of ingenuity you could look at some more old school 'project' ideas and make something to do the job – basically a sprung pointery thing with a very long pointer to magnify any movement. Back in the day there was no fleabay and perhaps not a lot of ex industry clock gauges to be found easily so people had to make their own. You don't need accurate measurements, just a sensitive way to tell if something is runnning out and which side is 'high'. IIRC there was one in a Unimat projects book by Gerald Wingrove… I know you dont have a Unimat but worth a look anyway. You will probably need to join the group but it's free… and then look at the Rex Tingey projects book for some seriously ambitious Unimat projects.

https://groups.io/g/Unimat/files/unimat%20manuals/Unimat%20Lathe%20Projects.pdf

https://groups.io/g/Unimat/files/Projects%20for%20the%20Unimat%20Part%201.pdf

https://groups.io/g/Unimat/files/Projects%20for%20the%20Unimat%20Part%202.pdf

To assess your spindle and bearing wear… you'd need to take it apart again but if it's working then I'd leave it be…

The spindle can easily be measured, preferably with a micrometer. You will probably figure out that the nominal size is a convenient fractional inch measurement. Then you will have a rough idea how much has been lost.

The bearing shells are more tricky but you could use the internal jaws of a caliper to find the smallest and largest diameter. Unlike the spindle, the shells will wear very unevenly. In the case of the bearing nearest to the chuck the wear will happen on the rear and top faces because the cutting forces push the job and spindle backwards and upwards. The caliper likely won't give a great internal reading but comparing the min and max should give you some idea.

Oilite… I'm not so sure that you will find one with the correct internal and external diameter to match the original. They can be turned although this does tend to cover up the pores. To turn the OD you would first need to make a mandrel with a shallow taper on it… time to practice some tailstock offsetting.

My tailstock clamp is a similar design – it's just a small iron casting with 60 degree face and a stud that goes up through the lug on the top part of the tailstock. It's just cast iron on cast iron but it's not really a bearing.

[Howard LewisParticipant @howardlewis46836]

Being pedantic, am not thrilled at the prospect of Metric Leadscrews on an otherwise Imperial machine, as a permanent solution.

But you have done a great job in making the lathe operational again, and very neatly..

Now, possibly could it be used to make the new Imperial Leadcscrews for the Cross and Top Slides?

But, before the days of graduated dials, craftsmen used their skills, with a six inch rule, plain callipers, and pencil marks, to turn things precisely to size, so in practical terms it may not really matter..

Howard

[Andy Thompson 3Participant @andythompson3]

Thanks guys, making more progress with your advice.

Good News and Bad news

Installed lathe on bench – see picture – it is a bit narrow so mods may be required. I have ordered a no volt switch.

Installed thrust bearings on slides – see pic. What a difference, makes it a pleasure to wind them in and out.

Ordered a missing screw gear from Myford – 20T and 14.5 but sems a little different. Turned a spigot to stick in the spindle and mounted the gear on it. It drives the saddle at 5 thou per revolution or 0.13mm. Does this sound right – seems fast to me but what do I know.

But now the concerning news. Ordered a dial gauge and mounted it as per pics. With dial on backplate when I rotate by hand I get TIR of 0.08mm/0.003". However when I stick a long bar in the chuck and apply a fair bit of pressure, I get sideways movement of 0.25mm/0.01" and up and down movement is similar. Are my bearings and spindle knackered!!!!!

Is this why I get a finish as shown in last pic. On the right was HSS knife tool (self ground) with and without feed. The start of the shiny bit is carbide with manual feed, not all of this is shiny. Screwfeed is chewed again, then I went back to slow manual feed and it bit a bit better.

Any idea what is the problem. Any other advice.

Cheers

[Andy Thompson 3Participant @andythompson3]

Forgot to add picture of the finish as below

[Andy CarlsonParticipant @andycarlson18141]

There are loads of reasons for bad finishes – the material and the tool amongst others.

If you want to eliminate the bearings from the equation then take a long piece of bar that will go through the spindle, centre drill it, entend it a long way out of the chuck and turn the OD using a tailstock centre for support.

Bell mouthed chuck jaws are another thing to check for before deciding that your bearings are the problem. TBH most old chucks have this issue to some extent. It comes from being overtightened on short workpieces in the past. Some shim around the workpiece at the outer end of the jaws can compensate for this, albeit that it's a bit of a faff to set it up each time.

Apart from the measurements I mentioned in my previous post you could use your DTI at a point on the spindle close to each bearing in turn. Then you should be able to figure out if the play is coming from one bearing, both or (hopefully) neither.

[Andy Thompson 3Participant @andythompson3]

Thanks Andy. I should have said that finish was on a 1/2" piece of EN1A that was also supported by the tailstock. You can see this and the tools I tried in the picture.

The main reason for my concern was the TIR when I hauled the bar back and forth and up and down. What is normal for a test like that. The TIR when rotating is about .08mm/0.003" which sounds ok to me, but when I pull the bar I get 0.25mm/0.010".

[Andy CarlsonParticipant @andycarlson18141]

I can't see how far away your work area is from the chuck but I'd say that being so close to the tailstock the finish is telling you very little about the state of your bearings or chuck. The tailstock is providing most of the support here, so the potential causes are the tool, workpiece material, the support of the job (i.e. the tailstock in this case) and the rigidity of the tool holding and everything beneath it.

The HSS tool doesnt look terrible – a bit sharp though – you should very slightly round off the corner of the cutting edge which will make it last longer and probably cut better. Check that the tool is exactly on centre height. If it's EN1A from a trustworthy source then that should be OK too. Check that your toolpost can't wobble in any direction (play in feed screws can be a cause), lock everything that doesnt need to move, keep the tailstock barrel 'stick out' to a minimum and nip up the tailstock barrel clamp (but no need to go mad).

I'd avoid power feeding TBH – most of these lathes don't have a fine enough feed to give you a nice finish – that's why you sometimes see some quite Heath Robinson fine feed contraptions attached to them. Another observation I'd make is that if you want a very smooth finish then you'd be better off with a round nosed turning tool rather than a general purpose pointy one.

Carbide… not something I have much experience with TBH. I use HSS for 99% of my lathe work. I don't think these old lathes will derive any benefit from carbide.

But mainly… keep at it, I don't think your results look too awful and with practice and careful checking of all of the variables your results should improve. You won't get perfection every time though so don't get too stressed over it.

[Andy CarlsonParticipant @andycarlson18141]

To answer your question about runout, 2 or 3 thou measured near to the chuck jaws on the OD of a bar held in the chuck is pretty average for a 3 jaw but it could be a lot more with an old chuck. You'll probably get a different answer with a different diameter workpiece because you will be at a different place on your chuck scroll.

Moving further along the bed and with a bar held only by the chuck… anything could happen due to the jaws being out of true. It will probably run out more but it could even run true at some point along its length.

If the job moves when you lean on it then I'd suspect bell mouthed chuck jaws before wondering about play in the bearings.

I'm not sure where you had the clock gauge on the backplate. If it was on the OD then that's pretty unimportant. If you took the chuck off the backplate and clocked the OD of the register then a few thou of runout is not ideal because this will be added to any error in the chuck itself. Thinking about fixing an inaccuracy in the backplate would probably not be a good idea until you have gained some more experience though.

[Andy Thompson 3Participant @andythompson3]

Thanks Andy. All good points.

I had the clock on the OD of backplate, what worried me was that when I hauled on a bar in the chuck that dial would move by 0.25mm or 10 thou. I remember reading about that type of test to check a second hand lathe but cannot find it again

I will experiment some more and see if I can crack the problem.

[Howard LewisParticipant @howardlewis46836]

If the OD of the backplate shows 0.003" runout, the three jaw chuck will show at least as much when fitted to the backplate.

How much deflection you measure on a bar gripped in the chuck will depend on the diameter of the bar, and how hard you push or pull.

Obviously, if you push hard on a small diameter bar it will bend, apart from any bell mouthing in the chuck jaws.

The important figure for runout is that for material, measured close to the chuck. This may be a combination of Bearing clearance (may be different in one plane from another! ) Runout of backplate OD (Which is the register that locates the chuck ) Check that both are absolutely clean before reassembling. – Might be worth giving the backplate a very light skim on the face to JUST clean it up, with a small chamfer on the edge Wear between jaws and chuck body, differences between one jaw and anothe, and any bell mouthing of the jaws.

To improve matters you have to eliminate, as much as possible each error, until runout of a round bar in the chuck is minimal. It might be worth checking that the bar OD in several places, on the same line to be certain that it is circular.

There have been cases where steel has bee centreless ground is lobular rather than absolutely ground.s circular.

This lathe is old enough to be Imperial, so I would measure and work in Imperial units. Using Metric leadscrews will involve you in calculations at almost every turn (No pun! )

A lathe with 3 mm pitch leadscrews, so there are 118 thous for every complete turn of the handwheel..

A M12 leadscrew is 1.75 mm pitch, so each complete revolution will be 0.0689"

If you work in Metric, 70 divisions will make each one 0.025 mm, which is 0.000984", so just under a thou., which may get you close to the best of both worlds. ASSUMING that commercial studding is accurate and consistent in pitch.

With commercial studding and nuts, you will have to live with whatever backlash results.

Howard

Edited By Howard Lewis on 02/08/2021 08:34:22

[Andy Thompson 3Participant @andythompson3]

Thanks Howard.

Think you may have misunderstood my deflection test. The clock was on the od of backplate so bending of the bar or the bell mouthed chuck would make no difference. The deflection of .25mm/10thou must be from the bearings. As you say it was a bit different horizontally and vertically but still 0.25mm. That was what concerned me.

[Andy Thompson 3Participant @andythompson3]

Duplicate post

Edited By Andy Thompson 3 on 02/08/2021 10:18:31

[Andy CarlsonParticipant @andycarlson18141]

Sounds like you do indeed have some bearing play.

I have a spindle and bearing shells from a high mileage scrapped Faircut lathe stashed away so I decided to dig them out and do some measurements. Your spindle and bearings may have a different design but this should give you a few ideas…

One thing that I had not thought of previously… the front end of the spindle bearing area protrudes by a few mm from the bearing. This part of the spindle surface is unworn so it provides a way to get the original diameter. There is also a smaller ridge inside the rear end of the front bearing shell but this is less useful.

As expected, getting repeatable and reliable measurements inside the bearing shells is difficult without any proper internal measurement kit. It's possible to get some measurements with calipers but inevitably they will be wrong. I was able to detect some degree of ovalness in the front bearing shell.

I think the best approach is to figure out the nominal diameter of each bearing and work from there. You should be able to get an accurate idea how much diameter has been lost from the spindle. For the bearing shells… a not very accurate idea.

My front bearing has a nominal diameter of 1 1/8 ins (28.575mm) and the rear one 3/4 ins (19.05mm).

I dont have a 1-2 ins mike so can only use calipers on the front bearing.
On the unworn ridge I get 28.55mm suggesting that the spindle was originally 1 thou below the nominal size.
On the worn part I get 28.40mm so it has lost about 6 thou of diameter.
Inside the bearing I was getting 28.62 to 28.65mm

On the rear bearing I got 19.00mm on the shaft with the calipers and 18.99mm with the mike. I got 19.10-19.14 using calipers inside the bearing.

The wear on the rear bearing is numerically less severe… but that's the one where the bearing housing snapped and resulted in the lathe being scrapped. The failure is due to the smaller amount of material in the rear bearing housing along with weaknesses arising from the oil holes (modified to take bigger oilers in this case) and any imperfections in the (not very brilliant) casting. Basically you can't predict the point at which the casting will fail but if it happens it's pretty much 'game over'… so tread carefully.

If you want to go down the oilite route then you should be able to figure out which ID to buy. Likely you will have some work to do to match up the OD. You will also need to figure out a way to provide any thrust flanges that exist on the originals. This approach would eliminate the bearing shell wear but you'd still have the lost diameter on the spindle itself.

The saga of my Cowells GEC motor bearings may help https://www.model-engineer.co.uk/forums/postings.asp?th=138249&p=1

Another option would be to get hold of some bearing material (there are whole threads on this topic – I decided on LG2 for my cross slide nuts) from M-Machine or elsewhere and machine new bearing shells. Buying hollow bar would be less wasteful than solid. You could make these undersize on the ID to compensate for the wear on the spindle. Measurement will still be a challenge. I did this by making my own plug gauges. As well as the required diameter, the gauges have a taper or a set of steps with slightly smaller diameters so that I know when I am getting close to size.

… but I still think that the bearing play is not the cause of any bad finish when turning so close to the tailstock centre. I'd put any work on the bearings some way down the queue until you have more experience of lathework.

[Andy Thompson 3Participant @andythompson3]

Thanks Andy – really appreciate your experience.

Pictures of my spindle and bearings below, clearly there is some wear on both. The nominal spindle od must be 1 1/4. There are flanged 1 1/4 " oillite bearings available here https://www.getyourbearings.co.uk/flanged-oilite/aj202624.html – very cheap. I would need to check the OD and maybe machine. I read that oillite bushes are oversize to allow for compression – that may not be great for my worn spindle – I presume smaller 30mm bearings would be too difficult to scrape.

However, I take your advice that I should try to sort out my finish and get some experience first. As you suggest, rounded cutters seem a bit better.

I read your thread on motor bearings – particilary that the bearing needs oil. My motor is a monster and I am not sure what sort of bearings it has. Took a cap off the back and stuffed grease in there. At the front there was a screw over the bearing so I took that off and stuffed grease in there. Should I have used oil?

Cheers

[RedsetterParticipant @redsetter]

OP- Your work on this lathe so far has been first class, but I think you should just make a few things on it, see how they turn out, and learn to use the tools before you worry too much about the bearings.

I don't think Oilite bearings are appropriate. The lathe is not designed for them. They are not meant to fit in split housings, and you are not supposed to scrape or ream them – or split them – to fit.

The Oilite bearing in your link has an OD of 1 5/8", so 3/16" wall thickness. Your existing bearings look thicker than that. Even if you find one which will fit straight in, it is unlikely to be much more accurate than what you have already, given the wear on the spindle.

Realistically, you do not yet have the skills or equipment (i.e. another very accurate lathe) to machine new bearings from scratch, so stick with what you have.

If your measurement of 10 thou play is accurate, then the bearings can probably be nipped up to reduce that. They have probably never been adjusted before, and the bearing caps look a lot stronger than on many lathes of that era, so you will very likely get away with a fraction of a turn on each clamp screw. It won't need much.

Just my opinion.

 

 

Edited By Redsetter on 02/08/2021 16:18:34

[Andy CarlsonParticipant @andycarlson18141]

If the oilite bearings are not expensive then I think I'd be tempted to get hold of some and experiment a bit. You can always bug out and put the old shells back in.

Have you figured out how the off the shelf bushes compare on OD to your old shells?

If it were me then my first strategy would be to forget that they are oilite and try to reproduce the old shells as closely as possible – i.e. cut the slot and drill the oil hole. Personally I don't see much of an issue with using them as a replacement for solid PB or gunmetal bearing shells. If it's going to need a lot of accurate machining then it's probably best postponed but I don't see any harm in finding out.

[Andy CarlsonParticipant @andycarlson18141]

Forgot to say…

Motor bearing lubricant… can be oil or grease depending on the motor. Of the three I've opened up the ball bearing one uses grease and the plain bearing ones use oil.

… which probably doesn't help much.

[AJAXParticipant @ajax]

Posted by Andy Thompson 3 on 06/07/2021 14:55:22:

Duffer, Ady1 – I have done domestic electrics and repaired things for decades so reasonably confortable. Earth continuity is fine, insulation is fine and circuit is RCD protected.

Not wishing to doubt your abilities (or in any way suggest mine are superior to yours), but how did you check the insulation? With a multimeter, or do you have a Megger-type 500/1000V insulation tester?

[Andy Thompson 3Participant @andythompson3]

Thanks guys, appreciate all comments, good and bad, I am here to learn (and stay safe).

Ajax – you suspect right, I just used a multimeter, its all I have. Body continuity to earth is fine. It is on an RCD circuit. Motor starts and runs fine. Do you think thre is a real risk? What else can I do.

Red, Andy, – I sure don't want to try running before I can walk – so I am in no hurry to change the bearings if they don't need it. I did nip the clamp up until the up and down movement was the same as side to side. Cautious about going further since it may not change the sideways movement. My concern was that it was so bad that it may be unusable and I didn't want to spend a lot of time and money a effort if that was the case. Clearly it useable to a certain standard and I don't need accuracy or precision at this stage.

I will improve my grinder/sharpener, tweak the gibs, and try different tools and speeds.

Re the oilite bearings – that site has 7 with 1 1/4 ID and different wall thicknesses – good chance of getting a fit – and only £12 each. Seemed like a good opportunity. However I cannot find my bearing OD without taking the spindle out again. Maybe revisit this in the future.

Cheers

[Andy CarlsonParticipant @andycarlson18141]

Posted by Andy Thompson 3 on 02/08/2021 21:55:15:

I will improve my grinder/sharpener, tweak the gibs, and try different tools and speeds.

Re the oilite bearings – that site has 7 with 1 1/4 ID and different wall thicknesses – good chance of getting a fit – and only £12 each. Seemed like a good opportunity. However I cannot find my bearing OD without taking the spindle out again. Maybe revisit this in the future.

Seems sensible. FWIW I did a quick check on the wall thickness of my worn shells before putting them away again. My first impression was that it was 1/8 in but looking more carefully I think it was less than that – around 2.7 to 2.8mm so not a very convenient number… even allowing for the loss due to wear. YMMV as they say in the States.

[Andy Thompson 3Participant @andythompson3]

Hi Guys. I promised to keep you up to date as I still need your advice and this may be useful for the next newbie.

First off I modified my workshop and bench. See photo. Added no volt switch – a pain that they use a mixture of spade and fork connectors. The bench is old worktop for free from kitchen suppliers who rip out old kitchens. Thought these would be ideal and resist the oil sitting on it. Need LED floodlights and desklamp to see a dammed thing. I was feed up with crunching on swarf and dropped tools so added the sloping bit at the front. This is really handy for stopping swarf and placing chuck key and tools. It is also perfect for leaning on with a cup of tea – reminds me a bit of those places we used to go to – bars I think.

Not sure whether to add shelves at back. Probably should now move the no volt switch – not sure where – on the wall to left would put me in the firing line of chuck keys etc. Any suggested improvements?

I have also realised that I can improve my range of speeds I run the belts squint. See photo of belts, the three lathe pulleys line up with the three rightmost pulleys on the countershaft. However if I run them one pulley i.e. squint, I can increase the top speed and infill gaps in the other speeds. Can you see any reason not to do so?

[Andy Thompson 3Participant @andythompson3]

Next I redid the tailstock clamp. See photo, the lump of steel on the left was the first attempt but I couldn't thread it, there was concern about it damaging the bed – and it is ugly. Replacement is two bits of aluminium screwed together. You might notice an extra screw. It is impossible to get the stud driiled in just the right position to leave the clamp flush on the bed but tight enough to hold. Solved by drilling the hole back and adding screw as shown on second photo. By adjusting this you can get it all to sit just right.

Now I remember being on the rigs and we were not allowed to use aluminium step ladders on the steel deck. Something about aluminium can leave a smear on steel – and this could lead to a spark and thus explosion. I kid you not. Not worried about explosion but could ali make my slide sticky?

[Andy Thompson 3Participant @andythompson3]

Next up while the ali was out I made a carriage stop. See picture. This has proved most reassuring as I often seem to be working very close to the chuck and this means that any slip does not mean disaster. However dont forget the screw for adjustment. Initially I just slid the stop against the carriage. Once when using carriage feed the whole lot jammed early as the space between the carriage and stop was filled with swarf. Very useful, but so ugly I may redo it in steel.

Thinking of stops I added a carriage depth stop – both ways. Again see picture. This seemed the easiest way. It doesn't interfere with toolpost and can be removed when not needed.

[Andy Thompson 3Participant @andythompson3]

Thought that a more secure toolpost may help my finish so made up the 4 way posts in the photo from scrap steel. Height for the small one is good for 8mm tools – are these robust enough? The big one works off the cross slide with 10mm tools (or 8 with packing).

Cannot say these are particulary time saving – at least not until I get more and shorter tools. Are two screws really enough to clamp tool securely.

Had to make more T bolts. See photo. The slot in the slide is 9.5mm. Best I could was turn and use M8 as a stud. Do I need proper milled T bolt?

[Author]

Posts