Can you guess what's happening?

Just wasted an hour of my life, up until now I have been very happy with my Chester DB10 lathe. It's not perfect and needed a lot of setting up to get it to cut to 0.01mm taper over 120mm, but I got there in the end, the leadscrew nuts constantly need tweaking to take out the backlash but all in all, it's not bad.

Today I decide to make sure that the compound slide was also cutting parallel, so piece of ally round into the chuck, first cut and then a light cut, micrometer showed it bob on, parallel, as did the indicator in the tool post.

So using the indicator in the tool post I clocked the compound slide on the ally round just turned. Just over 0.3 out when lined up on the markings, now matter how I trammed it, it was always showing light at the chuck and heavy at the tailstock end.

After several attempts to get it right I twigged what was wrong, can you guess?

You were using the saddle traverse to do the cut not the compound feed ?

No, it wasn't that.

Another clue, I zeroed the clock at the tailstock end.

No matter how I rotated the compound, it got thinner (less cut) as I travelled towards the chuck, the error getting progresively worse as I got nearer to the chuck !

compound slide is running uphill

Well down hill actually, but yes same fault as in not horizontally true to the chuck, well figured out Duncan.

I Will have to sort out why and fix it now I know about it.

I cant even imagine what you have managed to do.

Are you saying that even if you turn the compound slide to 5 degrees you still get the same reading on the dial indicater over the same path? Surely not? I must be misunderstanding something. (It's happened before!)

It does sound as though, possibly, the compound slide is not mounted parallel to the lathe bed, so is in effect going "downhill" as you wind it towards the chuck, which if it was low by sufficient amount, would vary the diameter of the finished job, or the reading on the the dial gauge mounted at centre height at the beginning of travel.

Maybe a bit of swarf under the base of the top slide? Or a burr or paint etc?

Edited By Hopper on 21/06/2018 03:33:29

If I were a betting man (which thank the lord I'm not, sir   ), that's where my money would be.

Edited By Mick B1 on 21/06/2018 06:45:07

The type of lathe that I would buy, etc
top-slide move, cross-slide move,
both of them together.
We'd be alright in the middle of the night
top-sliding together.
Sorry, couldn't resist, triggered a long-forgotten memory there, Mick B1.
DaveD

none of which merriment helps the OP - any update ?

Hi,

Hopper, no I was trying to ensure that setting the compound slide to zero gave a true (indicated) run along a parallel test bar and was measured as follows.

Test bar mounted in collett. Clock mounted in tool post and passed along test bar in both directions from both the side (the same as acutting tool would be) and then again along the top. This showed that the tool post travelled very accurately in relation to the chuck and any workpiece in the chuck/collett.

Next the compound slide was checked to be showing a zero degree setting and wound fully back towards the tailstock. Carriage positioned so that the indiactor was at the far end (tailstock end) of the test bar and positioned so that the plunger was at mid point (where the cutting tip would be).

As the compound slide was moved forwards (towards the chuck) the indicator showed the cut getting shallower, no amount of minor adjustments to the angle corrected this error and the error always ran the same way.

Carefull observation showed that the plunger was at mid point of the diameter at the start of the travel, but as the slide moved forward the plunger was dipping lower and thus indicating a shallower cut.

Further investigation showed that the fault is not being caused by swarf, or burrs or paint, just crap maching of the bed or way of the compound slide.

I have corrected the error using shims as stated but it shows how far out the maching is for the compound slide. Now with the compound slide set to zero degrees, it does indeed run 100% parallel to the workpiece, so when required to cut a taper I know it will be accurate.

Wow. Pee poor machining indeed. [Insert comments about Chinese quality control and the celestial merits of Myfords here] It must have been out by a good bit if it affected the turned diameter so much. Usually a few thou there will not make a huge difference, the surface of the job being close to vertical at centre height.

I would mount the body of the compound slide unit on the faceplate, mounted by the flat surfaces the slide follows, and take a cut over the base so it's parallel between base and the sliding surfaces.

Although, first you would want to check the surface of the top of the cross slide is parallel to its slide way surfaces. Easy enough done with a clock mounted to the bed and reading off the top of the cross slide surface while racking it back and forth with carriage traverse handle.

And you might want to face off a test piece first to make sure  your lathe faces square or slightly concave and not convex. On a machine that poorly built, there are no guarantees!

Edited By Hopper on 23/06/2018 05:14:11

With so many errors apparent from initial assembly you have to laugh at the accuracy reports that suppliers give with their product when sold from new, have they any value at all, I don’t think so.

Dave W

It does depend a lot on the diameter of the workpiece which I don't think you mentioned. You also kept saying the topslide was set to a graduation which is like expecting your car odometer to be accurate to a yard in ten miles. You do have to set the topslide with a dti each time you move it. Really it doesn't matter too much if you can do all parallel turning with the saddle and only use the topslide for tapers and short cuts.

Your quite right, diameter was 22mm. I was trying to set the top slide to with a DTI see how accurate the zero was, and then test with a known taper. As I couldn't achieve a zero and get the clock to run over the test length without error (no run out) with the dti and not even get close I knew something was wrong.

Closer inspection showed that after setting clock to zero at tailstock end and ensuring it was mid line/point of the diameter, it would start to back off as I moved towards the chuck. Any adjustments resulted in the same thing making it impossible to tram and that's when I did a bit more testing and discovered it was running down hill.

As I mentioned it turns 0.01mm taper over 120mm length which is good enough for the work I do (hobby stuff). However, if I were to upgrade my lathe (I will at some point). I want one that has more adjustment to achieve this and better levels of accuracy, getting mine to this level took several days and several strip downs of the head and use of 1200 wet and dry.

Ian,

Have you tried putting the DTI stylus on the top of the 22mm bar in the chuck/collet. (12 o'clock position ) and then traversing both the saddle and the compound slide in turn along the bar. This would confirm that the main bed was true to the lathe spindle and that your compound slide is running "level". A sloping slideway is far easier to spot with a DTI reading on the top of test.bar.

Bob D

You'll struggle to get much better accuracy than that out of most home shop machines. Even the Myford factory test sheet specifies an allowable tolerance of .015mm over 150mm. (.0006" over 6 inches.) (And they were the Rolls Royce of lathes, you know. )

Once you start wanting accuracy of much less than a thou, you are getting into the realm of cylindrical grinders, not lathes.

Hi, yes I confirmed the error by tramming along the top of the aluminium bar and again with the 16mm test bar, deffinetly the error is with the compound slide which I have shimmed for now.

Hopper your right, but if the lathe had a means of adjustment I could at least try to improve or maintain accuracy without resorting to stripping the head and messing about. I guess doing my apprenticeship at Rolls-Royce has something to do with it.