Lathe run out

Hi Robin,

I feel your pain, I have been there and done that as they say. I agree that if your lathe is not sitting securely with the head and bed lined up you will have the sort of problems you are currently getting, I started out with the same problem on my DB10. Don't get me wrong, it wasn't out by much, but it was out and giving a slight taper. I used the ‘Rolly’s Dad Method’ as it is sometimes referred to which amounts to twisting the bed slightly. At first I thought that it had provided the cure and that my lathe could not have been set up truly in the first place. However, I produced some test pieces and found that I had exchanged my taper cut for a bobbin effect whereby the centre of the piece is narrower and thus more material has been removed from the central area of the work when compared to the chuck or tail end, again we are talking very small discrepancies but I wanted as good as I could get. So looking again what was happening and the penny dropped, in twisting the bed the tool travels along a helix around the workpiece, in my case the tool was slightly higher than the workpiece centre line at the start of the cut, on the centre line of the work in the middle of the cut and then slightly lower than centre at the chuck end of the cut. So when the tool was cutting the middle part of the cut it was removing very slightly more material and hence producing a bobbin effect. Altering the amount of twist resulted in either the taper returning or a greater bobbin effect. I suppose I should really have left it at that but it ate away at me, in the end I decided to take the head off and check it, I very carefully and very lightly scraped the corner I thought was causing the problem and then reassembled the lathe, BINGO, no taper and no bobbin or any other problems or was there?

So happy that I could cut normally using the feed and carriage etc I decided to look at the rarely used top slide, I wish I hadn’t, it runs downhill so when wound back towards the tail the tool is higher than when it is wound forward towards the chuck, I used a shim to provide a temporary fix and to my shame it is still there, one day I will get it re-ground ….maybe.

Hi All,

here we go down the rabbit hole again! A couple of observations:

First, never enter this rabbit hole unless the lathe is not producing the parts you need to the accuracy you require.

With an engineers level, check across the bed at either end, and shim the cabinet untill the readings are as near identical as possible. you need to shim at four points, front and rear, at the headstock and tailstock end. The idea of this is to twist the bed into as near perfect alignment as possible. this process is best done over a few days to allow the bed to settle, and then be checked again, when you are happy with this, proceed as follows

It is pointless to try to align using the three jaw chuck, the very best of which are usually at least 3 thou out WHEN NEW!

Remove the chuck, check and clean the spindle bore, fit the bush and a known good centre, and then put a DTI on the spindle, and then on the centre.

If there is run out on the spindle, tighten headstock bearings to eliminate it.

If there is runout on the centre, remove the centre and bush, check the bore again, check the bore in the bush, and re seat the bush and the centre.

Buy a test bar,, remove the centre from the headstock and fit the test bar in its place, and check with dti at the remote end, you may need to remove and reseat the test bar till you get the absolute minimum possible run out.

Place the DTI on the saddle with its foot on the front of the test bar at centre height, and wind the saddle from the headstock towards the tailstock, If the headstock is misaligned you should see the DTI needle move steadily in one direction, if the reading is erratic, you may be looking at bed wear.

Take note of the runout at the remote end, rotate the spindle 90deg, and try again, note the reading and repeat till you get the spindle back to its original position.

You should now have a good idea about headstock to bed alignment, and will be able to make a decision as to whether you need to realign the headstock front to back, and whether the bed is worn or not

Now move the Dti to the top of the test bar and test from headstock towards the tailstock to see if the test bar is parralell to the bed, again be aware that you are looking for a constant downward or upward trend, and any erratic movement in the needle may be bed wear.

You now have a good idea whether you need to shim the headstock to correct any "nod" in the alignment.

You will usually find that by slackening the headstock bolts slightly, and bumping the headstock with a soft faced mallet in the desired direction, front to back alignment can be corrected.

Sometimes, selective retightening of the bolts can correct up or down nod, but if not you will have to shim, or scrape, but it may be a case of removing the headstock, cleaning the plate it sits on, identifiying and removing any dings or burrs with a flat stone or small scraper.

Now you have the spindle sorted and the headstock alignment as accurate as possible, you can move on to the tailstock. Whilst any of the above tailstock methods will work, I usually put my best centre in the tailstock, and with a dti on the side of the test bar, wind the centre into the centre drilling whilst watching the reading. Any discrepancy here can be corrected with the tailstock alignment screws.

Now repeat with the DTI on top of the test bar, and you will see if you need to shim the tailstock, by how much, and whether it needs to nod down or up.

Now move on to the turning tests ans see what results you get!

I must add that this is the way I do it, based on as much logic as I can muster at the particular moment. I am not suggesting that you go this far, indeed, you should only correct a fault if it occurs, and not go looking for trouble, because it is all to easy to imagine there is a problem. As I said at the top, only dive down the rabbit hole if your machine is unable to make the parts you want to the accuracy you require, and I will add to that that most lathes and associated chucks/tooling etc benefit from a complete strip down, clean, debur, lubricate, and carefull reassembly

The best of luck!

Phil

Very good!

It looks like you might be best to replace the two wooden riser blocks with a couple bits of 4" square steel tubing, thick walled etc. Bolt them to the floor and then bolt the lathe cabinet to the square tube, after inserting suitable shims to stop any rocking if the lathe is not sitting flat on the tubes at all 4 points. Then you can start your turning tests and shim either the cabinet base or the lathe base accordingly to get a nice straight cut.

Some of these larger Chinese lathes I have seen have some very thin cross sectional areas along the bed just below the ways, so can be quite flexible. They are nowhere near as rigid as they appear at first glance. I was recently looking at a bed repair done to one that fell off the back of a truck -- literally -- and the casting below the front way was only 9mm thick, so weak under load or twisting forces etc. This one had snapped clean off.

Another thing you should check on yours: It looks in the pics like you have a removable gap piece in the bed at the left hand end. These can cause trouble if the securing bolts and dowels come loose. There are two tapered dowel pins that locate it. Each has a threaded section on top with a nut for extraction. I can see one of them in your first pic. You need to loosen the nut and gently tap the threaded section with a brass drift to make sure the tapered dowel is "home". Then check tightness on the allen bolts holding the gap piece down. And check the removeable piece is not obviously misaligned at the join on the bed ways etc.

It's best not to remove the gap piece unnecessarily as they can be a pain to get seated correctly again. If you do, make sure all burrs and grit and swarf are meticulously removed before reassembly.

New here but always interested in this topic...
Am I right to roughly assume the following:
-As soon as you bolt the lathe Headstock-to-Bed combination(Maybe new from factory) down to something, in most cases the cabinet, your bed twist is linked to that, if all that weight is placed on the cement floor the Bed straightness/twist is linked to the cabinet and weight pulling it down to floor, so the bed is naturally not as straight as at the factory without the cabinet as grinded. So now you always forced to shim it at some stage of the fitment.

I my case I fitted a 12mm steel plate(70Kg) between cabinet and lathe bed, forcing me to shim it at the bed end points to cut a testpiece sticking out straight, luckily not at the headstock side, it took some time(Long).
I bolted the cabinet to the floor as well, but did not do my adjustment at the floor due to the plate stronger than the cabinet

Edited By Chris Mate on 12/05/2021 19:46:47

Wasn't sure if I wanted to admit it, but I do have a level, does that make me weird?

It is making me think I may want a floor mount I can adjust while looking at the bubble.

Not really, it depends upon the lathe. My secondhand lathe came bolted to the manufacturers cabinet. Early on I had issues with it turning tapered at the headstock, although strangely not when boring. The consensus on the internet was that it was a bad idea to fiddle with the lathe to cabinet bolts. Instead I spent ages shimming and fiddling with the cabinet bolt down points to no avail. In the end it turned out (just like Hopper mentioned above) that I hadn't been careful enough when replacing the gap piece. Having triple cleaned the gap piece and refitted it according to the instructions I loosened all the hold down bolts and just let the lathe sit on the concrete floor. The lathe now turns parallel to a couple of tenths at the tailstock and a thou or so at the headstock. Not bad for a lathe that is over 40 years old and is unshimmed.

Andrew

..............I do have a level, does that make me weird?

No, but using the level across the tops of the prismatic guides doesn't tell you anything useful, as those surfaces play no part in controlling how the saddle moves, which is what you're really interested in. In addition the saddle is controlled by one prismatic guide, not by both.

Andrew

Whilst maybe not strictly correct placing it across both prismatic ways at both ends is at least consistent.

I think you can (initially anyway) assume the prismatic ways relationships do not alter along the bed - it there would be all sorts of fun with the tailstock/saddle relationship.

That implies that the reading should be the same at each end assuming the lathe is not twisted.
The photos show 2 quite different readings, so there is at least a bit of twist to look at.

Dave

I have the same machine, and also fought the taper for a frustrating time. I made jacking feet from M12 bolts as my floor is not surface plate quality... Things improved a lot and have stayed true... I bought an MT3 testbar and got it to .0005" over 15" (actually it worked out about .0007" over 30" when I turned the back axle for my TE)

If someones had the gap piece out they are a real b****r to get back properly. I use a DTI to see if the carriage lifts when passing over it... If the needle moves it's not bedded properly.... do it again. The mating surfaces are not as well ground as they might be.

At the end I faffed around with the tailstock to get it as close as possible... and I've left it alone... IF I need an offset (only needed it a couple of times) I use a 2" boring head from the mill in the tailstock on an MT3 arbor

Good luck, the basic lathe is fine (especially for the money) some of the accessories are "sub optimal"
oh and you'll have to pull the coiled leadscrew guard off to skim the backplates on the chucks

Good start. I would first get it bolted down firmly to the floor. Then set it "level" (or at least the same bubble position) at both ends of the bed proper, not the removable gap piece. U-shaped shims slid in under the mounting bolts will do the job. Usually only takes a few thou here and there of shim. You will soon get a feel of how much shim causes how much movement of the bubble.

Once you have got it level, you can then repeat your turning test and see how it compares. Any further shimming to bring it to turn dead parallel can then be done.

Setting it level does not always absolutely guarantee parallel turning because as noted above, it is unknown if the tops of the inverted V ways are dead true to the angled load bearing surfaces at the front or the flat surface the carriage rides on  at the back. Plus these cheaper hobby lathes are often made from "green" castings that have not been aged to relieve casting stresses, so can move and distort a bit after being machined in the factory. Plus there can be the movemet of the carriage under the load of cutting forces. Also there is the factor of the removable gap piece being properly made and installed.

So your turning test will be the last double check just in case. Then, on with the tailstock but that is a different job.

Edited By Hopper on 13/05/2021 03:18:45

I might get away with glue but drilling holes in her garage floor is not really on the cards, we have to keep this credible. She has X-ray vision and knows instinctively if I am hiding something.

I like Mark's M12 adjustable feet, but how does that square with your "few thou here and there"? How about three m12 feet and one fine thread foot at the lighter end? I have a set of M17x1.0 taps and die which need trying out.

Someone in New Delhi is selling MT5 alignment test bars for £70 including carriage on fleabay. Worth taking a chance?

Consider a lathe on a ship

All well and good, but lathes are not built on ships.

They are usually built in factories with concrete floors and precision levels are used duing the build process to a national or "house" standard. Using a precision level to put the machine back to the condition in which was was checked out at the factory should produce the same test sheet results.

Warco sell adjustable mounting feet

There are holes visible in the machine base that will take these and allow adjustment of the machine base on the floor.

Nigel B

Crikey, you need to get 'er better trained. But that's a longterm project and beyond our remit here.

Four adjustable feet, the type with large round rubber or urethane etc pad at the bottom of each would do the job. Search for "machine feet adjustable" or "Machine Levelling Feet" and you have plenty to choose from. They come in different diamter mounting threaded studs to suit whatever holes are already in your cabinet base. They have two nuts on each threaded stud so can be adjusted to suit your needs. No need for ultra fine threads etc. Whatever they come with will do the job. Not as good as bolted down but better than perched on lumps of wood for sure.

I would not buy the test bar from India. Unknown quality -- some are very poorly made -- and not at all necessary to set up your lathe. The two turning tests I described earlier will do the job to a better standard.

Others here can probably recommend where to buy suitable machine feet in the UK and that should get you going.

It depends....the one that I bought was tri-lobed on a very coarse spiral down the bar, bent and (as far as it was possible to measure accurately across lobes) tapered. In essence, an expensive MT3 taper since that was the only functioning part. eBay agreed and refunded me immediately suggesting that my supplier might have already had some complaints...

The replacement from China (yes, I know, I never learn) appears straight and measures the same with my digital micrometer at each end but is not the finest piece of work. I use it with caution and remind myself that I didn't spend a fortune on it so shouldn't expect any more.

After further investigation RDG Tools have an HBM 5MT LATHE PARALLEL TEST BAR for £79.50 incl., but the description next to it is for a 'tween centres bar so I would have to enquire what it actually is

You don't need a test bar, just shim the lathe until it cuts parallel.

Tony

Alignment bars from India - yes it does depend.

I took a chance and paid £80 for an MT5 one with Express delivery (Arrived in 5 days)!

There seemed to be several suppliers and I was lucky.

Mine measured OK and the results matched those done by a professional, with an ex Harrison works real thing, on my M300 the week before.

My only criticism would be the lack of proper centers in each end, but still worth the money I think.

In your position, I would not worry about fixing the cabinet to the floor. You could fit adjustable feet, and adjust the level using those. Some cast iron lathe stands have tappings for just that purpose..

If you have to concentrate your efforts on shimming between the lathe feet and the cabinet. to remove twist from the bed, you need to read how to do it as per Ian Bradley in "The Amateur's Workshop" or the "Myford 7 Series Manual" as a guide to where to adjust the feet.

If you don't like shimming, you can use studding and nuts between the lathe and cabinet. One nut each side of the cabinets holds the studding captive, while a nut above and below each lathe foot provides the adjustment to take twist out of the bed.

Ian Bradley advocates leaving the Headstock end alone and making all the adjustments at the Tailstock end..

Once twist is out bof the bed , the machine should cut parallel on an supported bar (Follow Ian Bradley's advice and only take fine cuts, on a reasonably thick bar, to minimise risk mof the cuttingb pressures deflecting bthe bar.

If, after this, the lathe cuts a taper when working between centres, it is possible that the Tailstock is offset across the bed. To check, and if correct this, firstly you need two good centres. The hard one goes in the Tailstock, and it is worth trimming up the soft one in the Mandrel, immediately before use.

Fit an alignment bar (Or a substantial bar known to be straight and parallel. ) into the four jaw chuck and adjust the jaws until the DTI shows it to be running absolutely true..

Mount a DTI (The more sensitive, the better ) in the Toolpost, at centre height and set to Zero. Move the Saddle to the end of the bar, and note the reading Unless the Headstock is out of alignment, the reading should still be Zero..

Then mount the Alignment Bar between centres and set the DTI, still at centre height, to Zero. Move the Saddle to the tailstock end of the bar and note the reading..

If the reading is anything but Zero, use the adjusting screws to move the Tailstock, (lock one against the other each time ) until the DTI reads Zero at both ends of the travel of the Saddle.

You can check the centres for being at the same height.,by bringing them together so that they grip a thick feeler gauge or steel rule, placed vertically between the centres. If the centres are at the same height, the feeler will be vertical. If it is not, the centres are not at the same height. Hopefully this will not be the case, since your lathe appears to be fairly new.

Unless the error produces large taper errors, with a prismatic bed, you might be better to live with it.. Scraping or shimming the flat underside, might introduce other errors because of the prismatic way, so sleeping dogs may best be left to lie. Totally correcting the error may be difficult.

Howard

Howard

Sounds good and I agree with this Ian Bradley bloke, fix the headstock, adjust the tailstock. However, I won't be reading his book because education is for young people with their lives ahead of them, not for old gits living on anti-stroke meds.

I like shopping with my granddaughter, drawing CAD on my computer, and exporting that CAD to the machine tools in the garage.

This week I have bought a hoverboard, made some machine feet, and ordered some steels to fit those feet to the bottom of the lathe cabinet. One foot in each corner, one tie bar across the middle to stop the bed sagging like a hammock.

Once made and fitted I will reawaken this "Topic" and I will be most grateful if you could be here clutching your copy of "The Amateur's Workshop" just in case it all goes horribly wrong, as it probably will, confidence being that feeling you get just before you understand the problem

I make adjustable machine feet by welding large thick washers to the heads of bolts The washers go against the floor and I use flanged nuts underside of the feet, unless the feet are themselves tapped.

As for putting the level on the flat tops of the prisms - This won't allow for wear but it will let you see if the bed has twist. T

he prisms will have been ground in the same setup as the ways (or at least they should have). There's no other efficient way to do it.