Setting up the lathe accurately

Help / advise needed again please.

I have been trying to set up my Harrison M300 to turn accurately. I had noticed I was taper turning slightly.

I have read all the information about turning a bobbin shaped shaft and taking light cuts etc. By adjusting the levelling feet on the lathe stand I have managed to get results down to 1.5 to 1.9 thou over a six inch length. No matter how many times or slight the adjustments are I cannot get any better.

Am I expecting too much? Is this within realistic limits? Could I assume that I cannot improve on my results due to bed wear ? If so how do you measure bed wear in the home workshop using the general measuring tools available.

The lathe has a removable gap bed which I have indeed removed and thoroughly cleaned before replacing.

Should I be content with my results and just be aware of the machine limits for future use or am I missing a trick somewhere.

Thanks in advance

Dave

Is the tailstock a thou over to one side as this will also give the results you are getting

Hi Jason

I am not using the tailstock. Initially I just thought I would do some checks to make sure all was well with the lathe. When turning a 1.25" diameter length of bar over a length of about 3" I noticed the taper of about two thou which I considered to be excessive.

This led to me turning the test bar and making the adjustments. To now get 1.5 thou over six inches is obviously a big improvement. Another point is that the lathe stand is quite a substantial construction. Although I can get immediate results by adjusting one levelling bolt at the tailstock end I can only get so far. Would it be reasonable to raise the corner and see if the stand will twist over time giving a better result eventually ?

thanks

Dave

I am adjusting the bolts at the bottom of the stand not the tailstock. I have been advised not to adjust the bolts to the bed as this can cause even worse levelling problems as these are set in some kind of bedding agent when initially set up at the factory ??????????

Edited By Dave C on 15/11/2014 17:25:54

The OP did not say, but when I read the post my take was that the 'bobbin' was a bar just held in the chuck (not using the tailstock)

If the tailstock is involved then it obviously would be the first culprit.

If its just chuck held then it needs to be a decent diameter to eliminate springing.

Ian P

If the bobin is sticking 6" out of the chuck its liekly to be deflecting, which end measures larger?

The test bar is 1.25" diameter at the centre part. I have a sharp cutter and a fine feed rate removing no more than a couple of thou at a time.

Dave

The tailstock end is the large end Jason. I thought it could be deflection hence the very light cuts.

Dave

Have you checked the runout on the three jaw chuck? Even the best new 3 jaw chucks are only good to about 3 thou concentricity. Find a good parralell round bar of 3/4 or more, set it up in a 4 jaw with a clock guage to as near to zero runout as you can, then do the turning test, and remember the old adage "the most accuraute chuck you have is a 4 jaw"

Is this a new m300?. If you mean the leveling screws on the pressed metal cabinet, they are for levelling the bed only, not taking twist out of the bed. If you suspect the bed, do the above with a 4 jaw, and then slacken the bolts that attach the lathe to the cabinet and see if there is a difference. Try placing a bar of known goodness between centres and then running the DTI along it from end to end (lathe stationary). now tighten the bolts that hold lathe to cabinet whilst watching the DTI and see if there is aany movement on the dti. If there is you need to place shims under the bolts, and watch again as you tighten down. Ideally there should be no movement from the dti as you tighten the bolts, which means that the bed is not being twisted as they tighten. Now you have the bed in its "relaxed position, try the turning test, work out which way the bed needs to twist to improve the readings, and adjust the shimming to suit.

Phil

Forgot to say you need to lock the carriage and cross slide (tighten gybs if no lock) and use the compund slide, which must have minimal play in it!

Phil

Thanks Phil

I will have another try as soon as I get chance to spend some time on it and concentrate.

I was wary of undoing the lathe bed / tailstock bolts as I had been advised against it. I had used this technique on my last lathe, A boxford AUD with good results.

Many thanks to all.

Exactly right Dave, it is a job that requires time and concentration, and it can be very rewarding, but always have in mind the accuracy you NEED to achieve as opposed to that which you WISH to achieve. If you shim with ally beercan you will find that an extra tweak on the holding down bolts makes an adjustment, but above all, don't sweat it! engineers work to a tolerance, not perfection!

Phil

You should also try levelling the bed with a precision level. This is to ensure the bed is not twisted.

Put the level on the saddle, across the ways, move it to one end and note the bubble position. Slowly move the saddle to the other end - ideally the bubble shouldn't move. If it does, bed is twisted, or worn, which, as far as the saddle is concerned, amounts to the same thing. Adjust the feet or whatever, to eliminate the twist.

Note, it isn't necessary for the bubble to show that its level, only that it doesn't move as the saddle moves along the bed.

I don't know whoi is kidding who but for a lathe of Harrison M300 size you are already in the ball park of Schlesinger limits ,at least edition one.

Sure you can get it better,but who are you producing work for ? NASA? Are you in mass production where your parts have to fit those made miles away by another company? I would suspect ,in fact I would bet , that any skilled turner ( of whom I am not one) could produce more accurate work on the lathe as you have it than most of your respondents, with a toolroom lathe, and that with as much backlash as you like to specify..At one time this would have been done this with old fashioned callipers and maybe a micrometer as they got close. If you get it down to well within Schlesinger limits how much time are you prepared to invest in keeping it there?

Oh , and then don't lean on your lathe, M300 size or not.

Your lathe's pretty good. Regards Alan

Edited By alan frost on 15/11/2014 19:40:44

Hi Dave,

Don’t be disheartened by Alan’s put down, I for one would not be happy with the results you are getting and would make it my business to put it right.

Google 'levelling a lathe' and there is more than enough information to help you out.

Tony

Yes Alan, you are right, but is this an old lathe or a new one?, and it has just been moved and installed. If it is old, we don't know the state of wear, if it is new, it is likely that the bed was cast in china, and finish machined and assembled in the UK. Bed castings used to be matured for 5 or ten years to alow them to normallise, this is no longer done. A machine tool Manufactured to "Schlesinger" limits will not retain those limits if it transported and resited badly. The OP is merely trying to do the job properly My Colchester manual goes into detail on siting the machine, levelling, performing these tests, and how to correct any errors found. This should be done with every machine tool, new or old, after delivery and installation

Phil

It is not a new lathe it is probably about twenty years old I would guess.

Accurate machine work is difficult enough to start with for me at least and even more so with an inaccurate machine,

The whole intention was to get advise on whether It would be possible to improve my readings and get the lathe as accurate as possible and learn something along the way.

My thoughts were that if you start out as accurate as possible then you've more chance of finishing as accurate as possible

Yes Dave, It is possible, but it is not paticularly easy, and requires patience.You need to learn your machine, and start making or turning things on it. You will know when the lathe is less accurate that you are. As I said before, don't get hung up on accuracy, just have fun, play, and you will get better and more accurate. Do you have anything you want to make? What do you intend to use the lathe for? I will stick my neck out and say that a twenty year old Harrison is better than a new one because it was all made in England! If I am wrong, I am sure someone on here will correct me

Phil.

I have a Harrison M300, built in 1980, and I had a very similar issue. After I got it, and got used to it, I found that the machine turned taper near the chuck. If I recall correctly about 3 thou over 4 inches. A precision level confirmed that there was an uneven travel to the saddle. I spent ages bolting the cabinet down, adjusting the hold down bolts, measuring, adjusting and so on ad infinitum. I got it a bit better, but not significantly so. I was resigned to the performance and took to turning longer pieces to get the parts I needed, as the taper wasn't quite so bad further out. That should have been a clue.

When I turned the journals on the front axle for my traction engine I was operating about 30" from the chuck. The journal was parallel to within a tenth or two; I can't really see any difference on the micrometer. Then the light began to dawn. In between starting to turn, and noticing the taper turning, I had taken out, and replaced, the gap piece.

So I took out the gap piece again, gave it a thorough clean, cleaned it again, stoned the mating surfaces and made sure I followed the correct procedure for replacement. I also made a cut down Allen key so I could properly tighten the rear hold down bolt on the gap piece. Well blow me down, the lathe now turns parallel near the chuck to better than half a thou over 4". The simplest solutions are often the best.

Mind the gap!

Regards,

Andrew

Sorry Mr. Frost, but your answer made me shaking my head in disbelief. Possibly you think that the OP's numbers are in millimeters? I admit I don't know the Schlesinger limits for a lathe, but I'm quite sure that a diameter difference of nearly 0.05 mm over a distance of around 150 mm is not good. (sorry again for using these newfangled units, but that's what I need to make me a mental picture)

So it happened just a couple of days ago that I also wanted to know if my lathe (not far from 50 years old) is still usable or crap. I now have the test piece in front of me, I noted the numbers on it. A piece of free cutting steel, originally around 21 mm in diameter. HSS tool (yes in one of my new Wimberley holders), fine feed and a small depth of cut). My turning length was around 100 mm, or 4" if you prefer. What I got - measured with my best metric micrometer which has a vernier to 3 digits - is as follows:

- on the free end: 20.661 mm

- in the middle: 20.665 mm

- on the chuck end: 20.665 mm

What would Mr. Schlesinger say to that?

I'm very glad I didn't have to tamper with the mounting of the bed on the cabinet, this would be very difficult for this lathe (I even had to search for the screws...)

Regards, HansR.

I imagine he would say,yes,lathes were manufactured like everything else to a tolerance. Some lathes, by the nature of statistics, from the same manufacturer were more accurate than others. Many manufacturers charged a premium price for the accurate ones and gave them a separate title or part number. You obviously have one of the good ones. Schlesinger does not say that every lathe has to be bang on. Just that these are standards that a lathe should meet to enable reasonably accurate manufacturing practices. Dave's lathe is n't meeting schlesingers standards but its in the ball park.

As I also said yes ,Dave can improve the accuracy of his lathe but does he need to, how long will it stay this accurate, and how much trouble will he have to take to maintain this accuracy.? Lathes turning to the accuracy of your lathe are usually pampered by operating in temperature controlled rooms and are certainly not leant on while cutting. Quality watchmaking lathes ,not the most rugged admittedly, but very accurate, demonstrate very measurable flexing of the headstock bearings under very small forces, far less than normal cutting forces, and the inaccuracies induced into the most rugged of lathes by just leaning on a part of the lathe are well documented. Maybe you were leaning on yours at just the right place,maybe not. Either way is an error such as Dave describes important in the work he is doing.. IIs very easy to get carried away into working in thousandths of a mm but this is a very very small distance both in reality and in most applications.and usually does n't matter a hoot. As Phil ,I think said,, earlier, don't sweat about it, even the Swiss work to tolerances.

There was no put down I am aware of , Dave's lathe is inanimate-I was just trying to save him a lot of work that may or may not be successful or necessary.

Phil. above , sounds like an experienced lathe man  , and has given much the same advice.

Regards Alan

Edited By alan frost on 16/11/2014 01:12:01

Edited By alan frost on 16/11/2014 01:14:34