New Lathe

[Dave Harding 1Participant @daveharding1]

To day I used a length of  7/8" diameter silver steel bar stock 12" long to check alignment run out of the lathe.

I did not want to ruin this piece of bar as its for a project I would like to get under way.

The bar stock I used to turn down is 4" long piece of 1" silver steel. I used a glanze facing tool with a carbide insert to make the cuts. I used the parting tool to give me a end point to finish my cuts.

The measurements are the result of several passes I took small cuts.

I have no idea how to post pictures.

I have some mechanical knowledge so I am not splodging completely in the dark.

Thanks for the replies please keep them coming.

Edited By Dave Harding 1 on 28/10/2013 00:03:20

[michael howarth 1Participant @michaelhowarth1]

Dave……your latest set of measurements do not seem to be outrageously far out and were of the same sort of order that I was getting with my Warco 918. I was able to remedy the situation by adjustment (shimming) of the tailstock height relative to the headstock. Infuriatingly fiddly but got there in the end. Unfortunately, my lathe is on a wooden bench structure and I still need to do a bit of tinkering depending on the weather!

Mick

[Les Jones 1Participant @lesjones1]

Hi Dave,
Have a look at these posts for information about posting pictures.

Posting pictures

Posting pictures 2

Les.

 

Edited By Les Jones 1 on 28/10/2013 08:32:30

[Russell EberhardtParticipant @russelleberhardt48058]

O.K. so you now have 3 thou taper over 1", bigger away from the chuck. That could be twist in the bed (possibly a result of bolting it down to a non-flat surface) or, possibly, deflection of the work by the tool pressure (some of the carbide inserts are not very good with fine cuts and the chuck may not be holding the work well).

It's very easy to correct the bed twist using your 7/8 x 12" bar and a dial gauge. Try Googling "Rollie's Dad's Method" and you will find a very straightforward description of how to do it. The method doesn't rely on the chuck being accurate or the bar being straight. The bar only has to have the same diameter along it's length.

Russell.

[Dave Harding 1Participant @daveharding1]

I checked out the posting pictures links that's a awfully complicated way to post pictures.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Dave Harding 1 on 28/10/2013 19:27:40:

I checked out the posting pictures links that's a awfully complicated way to post pictures.

.

Ah but, it does give MTM the comfort that you have accepted the Terms and Conditions.

MichaelG.

[Howard LewisParticipant @howardlewis46836]

Hi Dave!

From what has been said so far, I agree with those who think that the lathe bed is twisted. If so, it will NEVER turn parallel.

If you want to turn using the Topslide, you need to set it parallel to an alignment bar using a DTI in the Toolpost, ONCE the bed is free of twist.

On my lathe having done this, I made up and fitted a dowel to ensure that was reset to parallel after any off setting.

You need to borrow a precision level, and check the level at Headstock and Tailstock ends of the bed

If the lathe has a vee bed, you need to mount the level on the cross slide.

Adjust the Headstock end, by shimming the lathe foot, or adjusting the holding down bolts (If it is direct onto a wooden bench, I would lift it up and put a sheet of at least 3mm steel under it, drilled for the holding down bolts to pass through, before trying to set it level and free of twist.

Having set the Headstock end level, across the bed, move to the Tailstock end, and adjust/shim until that end is level.

Then move back to the Headstock end and recheck,/ reset that level, and then move to the tailstock end and reset that level again.

Hopefully, each time the error will decrease. You keep on doing this until the error is either JUST acceptable or even better, Zero.

Try to be consistent about the torque applied to the holding down bolts/nuts, this can affect the readings. It did on my Myford 7 even when mounted on a steel bench on a steel sheet. The slightest tweak on a bolt would send the bubble flying around!

On one Industrial Lathe, the adjusting screws on the heavy cast iron base could not bring it completely level, and I had to resort to shimming between the Tailstock end of the lathe and the base to get the bed free of twist!

If you then want to check the alignment of the Headstock to the bedways, you need mount either an alignment bar or a piece of ground stock, (silver steel?) circa 1 inch/25mm diameter in the chuck.

The bar MUST be consistent in diameter along its length.

Mount a DTI on the toolpost, with the stylus at centre height, and set it to Zero at the Headstock end.

Without touching anything, traverse the saddle to the tailstock end of the bar. If all is well, the DTI will still read Zero.

If it doesn't the axis of the bar is not parallel to the bedways. This could be that for a number of reasons.

1) The chuck may not be mounted square to the axis.

a)Check that all the mating faces for the Backplate and the mandrel are all clean, and free of damage.

b) Check that the mating faces of the Backplate and Chuck are clean and free of damage.

If a) and b) check out as damage free, and the alignment is is still out;

(Hopefully, the rear mating face of the chuck is square to the work holding faces of the jaws, or vice versa.

If not the problems with the chuck need to be sorted out)

2) Remove the chuck from the Backplate, refit the Backplate to the mandrel and lightly skim the surface of the Backplate, just enough to clean it up. (This presumes that the gibstrips on the Saddle and Crosslide are correctly set, and do not allow either to move about when pulled or pushed across their ways).

Refit the chuck and recheck the alignment. Hopefully things will now be fine.

3) If the misalignment persists, then the Headstock mandrel axis would seem to be out of parallel to that of the bedways.

If the lathe is new, you have EVERY reason to complain to the supplier and demand a replacement, or at least to have your machine rectified at their expense, since it is not of adequate quality. You may not have bought a Toolroom lathe, so do not expect a tenth of a thou accuracy, but the figures that you are quoting say that something is radically wrong. You have every right to expect it to be within a thou from Head to Tail.

This too loo long for one post, so more follows.

Howard

[Howard LewisParticipant @howardlewis46836]

Continuation of the previous post

4) Once you have things aligned satisfactorily, you ought to check between centres.

At the the risk of being condescending, you do know that unless the Headstock and Tailstock centres align in both planes you will never be able to turn parallel between centres?

(Long tapers are turned by deliberately offsetting the Tailstock centre from the axis of the Mandrel)

To align horizontally, use the alignment bar between centres, and adjust the tailstock across its base (clamped down) until the DTI shows Zero at both ends.

If you use a piece of silver steel , I would suggest clocking it true in a four jaw chuck , before centredrilling.

If the piece that you are using is too big to pass through the Mandrel, use a steady to align the outer end until it does run true. Then centredrill it

Repeat for the other end.

To check vertical alignment, the DTI needs to be on the centreline of the bar and vertically above it.

If the Tailstock end reads high relative to the headstock, look for debris, or damage where the tailstock sits on the bed. If clean and still high, material can be GENTLY removed from the base of the Tailstock, until the Zero Headstock end reading is reproduced at the Tailstock end.

If it reads low, somehow, you need to shim the Ttailstock upwards. Try to do this between the tailstock body and its base.

Once you are happy with the readings, recheck and reset the Tailstock horizontally.

If you have any doubts, repeat the checks, and adjustments, and keep on redoing them until you are happy with the results.

It may be tedious in the extreme, but if you want good results, the machine has to be correctly adjusted, and the time will, ultimately, be well spent.

Sorry to be so long winded, but you have to eliminate each error in turn if you are get the best out of the machine.

When you have it sorted, you will have a great feeling of satisfaction, and much more confidence in the machine.

Howard

[Dave Harding 1Participant @daveharding1]

All interesting stuff I will read it a couple of times to take it in.

I am not going to do any thing else with the lathe until the week end. I bought the lathe new I have only had it four weeks so I am not a happy bunny at the moment. Its not what I had in mind when I bought it I did not expect to have to go to these lengths to turn a 1" piece of bar stock with out a taper.

[frank brownParticipant @frankbrown22225]

It could be that your chuck is out of alignment, due to its fixings or its jaws. With one of your "turned" rods, take a very light cut again. Mark the rod and a chuck jaw. Measure the diameter you have just turned. Now take your rod out of the chuck, rotate it 180 degrees and re-chuck it. If the chuck is running true your "old" turned diameter will still be running true in this new set up. Retract the cross slide a long way, cover the turned area will some marker pen (Sharpie?&quot, now take the tool into the work until it just touches the work, note the reading on the crossslide dial. Take deeper cuts until all the marker clout has been removed. Note new reading on the dial. Now if the chuck is running out as you removed material, it would have been removed less from one side. So you can sit down and work out how much out of true the chuck alignment is, if the cut was the same as before, then the chuck alignmant is OK.

Frank

[Russell EberhardtParticipant @russelleberhardt48058]

Posted by Dave Harding 1 on 28/10/2013 21:13:57:I bought the lathe new I have only had it four weeks so I am not a happy bunny at the moment.

If you bought the machine new it is most likely that you have twisted the bed by bolting it to a non-flat surface. Cast iron is flexible and this can occur with any lathe. Of course the heavier the construction the less this effect. The headstock is unlikely to be misaligned and the alignment of the chuck can't cause taper. As I have said try un-twisting the bed using "Rollies Dad's Method". It is dead simple and takes only a few minutes.

When I had set up my 9×20 using that method I had less than a thou error measured over the length of a 13" silver steel bar held in the chuck.

Russell.

[Dave Harding 1Participant @daveharding1]

Thanks for all the replies interesting stuff. I borrowed some alignment tools from work a dial gauge a alignment bar etc.

I slackened the holding down bolts then took a cut it was no different so I tightened them up again.

I put a 7/8" silver steel 13" long piece of bar stock in the chuck and clocked it turning the chuck by hand it was not bad at all.

I then took the top slide of disassembled it and measured every thing. I then reassembled it and set it up using a parallel bar and low and behold it is now cutting parallel using the saddle traverse. So the moral of the story is I should not rely on the vernier gauges fitted to the lathe. I am going to mark it up so I have a starting point for future reference. I am a lot happier now thanks every body for your help it is not a total disaster after all.

[VersabossParticipant @versaboss]

Posted by Dave Harding 1 on 29/10/2013 21:24:11:

….. So the moral of the story is I should not rely on the vernier gauges fitted to the lathe. I am going to mark it up so I have a starting point for future reference.

I'm glad to hear that your lathe now is turning parallel.

However, this remark above puzzles me very much. I suppose what you call 'vernier gauges' are the scale drums pn cross-slide and top-slide (and these rarely have vernier scales attached, but I don't know your lathe).

See http://en.wikipedia.org/wiki/Vernier_scale, also called Nonius.

But I don't understand what you did with these scales. Did you try to somehow measure the diameter after your turning exercises, by touching the bar with the tool? As I said' I'm puzzled, as I don't see what influence these scales can have on parallelism. If the scales are not precise (e,g, the tool moves more or less as what is shown), that's another problem then. But it would not influence the parallelism.

Greetings, Hansrudolf

[NJHParticipant @njh]

I too am a bit thrown by the "vernier gauges fitted to the lathe" and this only makes sense to me if this refers to the protractor scale on the top slide. However Gray ( as usual!) has it right and the topslide is for short tapers. Best however not to assume that the setting shown on the protractor scale is accurate if you need to generate an accurate angle.

Norman

[speelwerkParticipant @speelwerk]

Even on my 1998 British made Myford S7 that protractor scale zero setting is of by half a degree. Niko.

Edited By speelwerk on 30/10/2013 12:09:42

[mechman48Participant @mechman48]

The protractor scale on my 1 year old Warco WM 250 is also off by 1/2 a degree, I set the top slide true with a parallel bar (silver steel) & DTI & fastened everything down & left it set at that, so now unless I absolutely need to have a taper (for what I do at the mo' is very rare) it doesn't get moved.

George.

[speelwerkParticipant @speelwerk]

George; you can also scratch a new zero marking at 10 degrees, makes live a little easier. Niko.

Edited By speelwerk on 30/10/2013 13:54:35

[BazyleParticipant @bazyle]

On a lot of lathes the pin or stud about which the topslide rotates is not designed to be a precise fit or reference point so the degree scale can be variable even if you make a new 'zero'. At some stage you can make a genuine vernier scale to add to the degree scale if the above does not make it pointless.
If the toplide has a flat side face that is also accurately parallel to its movement you can set that against a square held against the face of the chuck. Another method is to set a centered holed bar between centres as a reference.

Another point is that on some lathes the whole saddle can twist slightly on the bed anyway. This does not happen on top v-beds but does on round and flat beds, including those with the v slope underneath a flat top.

In the end, if it is something important, you still need to check with a test indicator.

In many ways if the lathe has a leadscrew that can be disengaged from the saddle then it is just as well to deliberately keep the topslide offset to avoid temptation.

[NJHParticipant @njh]

If you set your top slide at an angle then you can get a finer control of the in feed for those times that you need to remove half a gnats whisker from the work. For example you can set it such that 1 division on the top slide dial advances the tool 1/2 thou. ( or any other amount depending upon the angle that you set). To find the angle you need you will need to revisit those maths lessons where trigonometry was on the menu. ( See they SAID it would come in useful one day!)

N

[Dave Harding 1Participant @daveharding1]

Sorry for the confusion its the protractor gauge I am referring to its way out. It needs to go in the bin.

The book I was advised to buy arrived in the post this morning I will give it a coat of looking at before having another go at turning expensive silver steel bars into scrap.

[Russell EberhardtParticipant @russelleberhardt48058]

Hi Dave,

The original measurements you quoted would equate to an error in the topslide setting of just under half a degree. That seems to be a reasonable error for the degree scale on this type of lathe. That scale would usually be used for things like 45 degree chamfers or feeding the tool in at an angle for screw cutting. Anything requiring accurate tapers would need setting with a dial gauge, a reference taper, or a sine bar.

Enjoy the reading.

Russell.

Edited By Russell Eberhardt on 31/10/2013 21:07:15

[Dave Harding 1Participant @daveharding1]

I have started reading the book I think its going to be a uphill struggle there is more to this machining business than meets the eye. I have loads of projects I would like to start but I am realizing there is a huge learning curve to get over. Thanks for all your input much appreciated.

[Author]

Posts