Headstock alignment - harrison l5

Does anyone have any links to a guide to setting up the headstock alignment on a harrison (or any for that matter) lathe?
My lathe cannot cut straight and everything I do I have to turn a fitting in situ to hold the final workpiece and not remove it. If I put a bar in the three jaw and centre drill it, I can get more accuracy by eye using the drill press and a regular centre punch its that badly aligned and dead centres in the spindle and tailstock are a few mm off centre to each other.
Now, I know in the past the headstock was dismounted from the lathe bed for repair to the spindle, so my concern is how to make sure it is aligned straight. I did have a machinist friend of a friend come over a while back, and he made a few sample pieces on it and proclaimed it badly aligned but then after a few hours of hitting the chuck with a lump hammer trying to get it to run true told me it was beyond his talents (he didnt slacken any of the headstock->bed clamping bolts).
Most of the guides I have found by searching mention setting over the tailstock (which as far as I can tell isnt actually adjustable)to suit the headstock, but Im thinking its a problem with the head alignment (or worse, potentially both, or a bent headstock spindle.)
Ive got rid of all the play in the various slides etc, and fitted a T slotted table and amassed tooling to suit it, so Id quite like to spend time getting it right rather than just buying a "better" lathe that potentially will have problems of its own.
It has the heavy fabricated steel harrison solid stand, (1/4 a metric tonnes worth!) so the mountings are onto a rigid platform.

Any links/suggestions?

First check the ‘level’. If you’re not familiar with lathe ‘levelling’, this does not mean getting the lathe to be truly horizontal, but involves the use of a precision (i.e. sensitive)spirit level across the ways at various positions. The level indication should show exactly the same at each position, otherwise it means the bed is twisted (assuming no wear, which would confuse things). This does not necessarily imply a permanent twist, the twist usually being the result of it not being properly shimmed under its support points.

Alignment of the headstock is another matter, and no doubt some advice will be forthcoming from other members.

Are you familiar with Tony Griffiths’ lathes.co.UK website? There’s a good Harrison write-up, from which I’ve extracted this:-

‘….great care was taken to assemble and shim the lathes onto their stands - and that it would be a Very Bad Idea Indeed to disturb their cosy relationship. Harrison also recommended that the stands, being rigid units, were not bolted down, but set on steel plates, on a concrete floor, with jacking screws fitted into the holes at each corner and the lathe "levelled" from those.’

Source:-
http://www.lathes.co.uk/harrison/index.html

Hi Mr. Fluffy sounds like you've got a bit of a problem. First things first, as Asquith says make sure your lathe is level ( I like to bolt them down as well ,that way they stay level longer ).Use a good quality level ( 0.001" per 10" is what I use).
When you've done that get a piece of free cutting round bar 18" long by about 1.250", grip about 5" in the chuck and leave 13" sticking out, do not support the free end, With a very sharp , pointed tool take a light cut down then bar for 12" if the tool cuts all the way down great, if it doesn't take another light cut until it does. Measure the bar at either end, depending on what readings you get you may need to "throw" the headstock. It's a long time since I worked on a Harrison but most lathes are the same, it will have bolts on all four corners to hold the headstock to the bed ( some fools even put the bolts inside the headstock so look out for that),it will also have a push me pull you type device at the end furthest from the chuck to move the headstock from side to side.I never undo all the bolts, leave ONE tight at the chuck end so you can pivot the headstock off it.
Now then, this is where it gets a bit complicated to describe, if your test bar is bigger at the tailstock end end you need to move the end of the headstock furthest from the chuck away from you bringing the tool more into contact with the metal, if it's smaller at the tailstock end you need to bring the end of headstock towards you. Tighten up and take another test cut, ideally what you want to achieve is a reading of 0.000" at the chuck end to + 0.0005"at the tailstock end of your test bar.
When you get to this point your headstock will be aligned true to your lathe bed, you then need to set the tailstock barrel true to the headstock spindle. Get another piece of round bar and turn it to the same diameter as your tailstock barrel , then by putting a test indicator on your saddle and winding from one end to another from you can determine how far out the tailstock is from true centre,you should then be able to adjust the tailstock over on its base until both spindle and tailstock barrel are in line.
In the real world your lathe would have a taper adaptor to fit in the spindle nose and two morse taper hardened and ground test bars to make life easier. We would also clock the top of the bars as well as the sides, I could go on and on but it can get confusing just writing it down.
All the best with your adjustments, I hope things go according to my plan, Reards Tyrone.

Mr Fluffy, I've never come across a tail stock yet that isn't designed to be adjusted from side to side, it's just that the makers can be clever at disguising the methods of adjustment.
Don't let your friend try to adjust it , make him check in his lump hammer at your door. Regards Tyrone

Allignment of the headstock is checked by cutting a bar aluminium or bronze about 50mm dia. over a lenght of about 200mm without support of a centre
It has to be done with a very light cut and a sharp possitive tool The headstock is good when the dia. on both ends is the same Then place 2 indicators on both ends of the cut (sideways)and zero them
Then correct the headstock until the difference in size measered (in radius) on the bar is also seen on the indicators So if you have a differenc of 0,03mm you move the head until you see a DIFFERENCE of 0,03mm on the indicators

Then you can adjust the tailstock
Place a magnetic based dial test indicator on the chuck and turn inside the morsetapered hole of the tailstock
Now you can see if the tailstock has to move up or down,left or right. Adjust accordingly
All this has to be done after you have levelled the lathe

To my way of thinking you could have a major problem or one as simple as a few adjustments.I've read and re-read your post and you don't say how much runout you have at the headstock.That's where I'd start by setting up an indicator and checking that but first recheck the bolts that secure it to the bed.Check the runout on both axis and also take a long pry bar and see if you get any up and down movement(indicating a possible loose spindle).Somehow your tailstock is adjustable.I've yet to see a modern lathe's that wasn't.If the problem is,in fact, in the headstock you should be able to compensate with the tailstock,up to a point.As Asquith said check to see if the machine is level overall.If it's a really long bed lathe then that could throw it off at either/or both ends.I guess if all else fails you could try a different 3 jaw chuck or go to the more accurate 4 jaw and compensate that way.Just thoughts.

Mr Fluffy,

I know this is a different model, but if you look at the second photo here:-

http://www.lathes.co.uk/harrison/page12.html

You'll see a tongue & groove arrangement between the tailstock body and base, and a small screw. There's probably a similar screw on the other side. These are used to set the tailstock over after undoing whatever clamps the body and base together.

Ray, I'm sorry but if your headstock spindle is not in true alignment with the bed ways nothing you do to the tailstock will correct the problem. Any component you turn in the chuck will always be tapered, I can see that work between centres or chuck and centre will be o.k. but this is making two wrongs into a right.
Regards Tyrone.

Occam's razor says the simplest explaination for
the problem is most likely the correct one.

If his part won't run true in the three jaw
chuck, and parts won't repeat in it either, that
says to me the lathe is fine.

The chuck is worn.

Hitting it with a lead hammer won't make it less
so.

Usually stories like this wind up with me asking
"how badly does the workpiece not repeat, or
how far off is the centerdrilled hole from the
true center of the part?"

And the response is ususually, well, about 15 or
20 thousanths.

Then I ask, how old is the three jaw scroll chuck
and the answer is, I don't know, it came with
the machine.

Then I tell them to perform the two-collar test
on a piece of stock held *firmly* in any kind
of tooling in the headstock.

But the key here is that MrFluffy does not give
any real numbers about how far out of 'alignment'
his lathe is. Basically to determine what's
wrong, we're going to need some real numbers to
start.

1) how old is the chuck.

2) how many thousanths of motion is seen when
a one inch piece of stock is chucked up, and
the indicator touched off an inch from the
chuck, and the one foot long stock is pushed
or pulled with a few pounds at the end?

3) Turn a diameter on a one inch piece of stock.
Then slack the chuck and spin the part 90 degrees
and turn another diameter. How far off are the
two surfaces, in thousanths?

Jim

Asquith regarding the lathes.co.uk, yes Ive studied his L5 pages quite a bit, from his info mine is a second generation L5 from late 40's to early 50's in age. My tailstock also has those screw heads, so I guess thats how it is done.
The trouble with the advice about not taking it off its stand is, its already been disturbed from the stand a few times that I am aware of. When I first sighted it in the previous owners workshop it was already seperated, and in my ignorance I have split it apart since to move workshop with it in this way myself also, it being just in the load range of my engine hoist to carry in through narrow doors in this way.
For the machinists spirit gauge, I will have to purchase one for the job, since my workshop is in a very rural location and I dont know anyone else locally who can help (and there is a language barrier too) and I probally would make good use of one anyway. I have just been checking them out on a certain auction site and theyre within budget..
I will try to buy a standard too so at least I know I have one thing that will be straight and true.

Jim, I understand the chuck is always the first port of call with misalignment problems, but I don think this is the case here. If I put in my hardened solid centre into the headstock taper it bypasses all the chuck and backing plate, even the register itself. As an aside I did the chuck deflection test on the best of my 3 jaw's and it registered 2 thousandths of an inch.
I took a photo of the alignment though Ive posted it as a link as it is 400k in size.
centres misalignment

For the resolution, I will order the spirit gauge when the seller gets back to me concerning its accuracy, and work through the posts here in sequence. Thanks for all suggestions.

Like everyone says, make sure the ways are level and parralell (?) I bought a "truing bar" from Miller Machine that draws into the headstock. I mount a dial indicator in my tool post and sweep it, both on the side and top. When I am happy, then I take my test cut to make sure. Good luck

OK, you got rid of the three jaw chuck, that's
good. Now take the tailstock off the machine.

Yep, you cannot learn a *thing* about the
headstock or bed alignment by starting with
the tailstock.

So take it right off the machine.

Really. Leave it off for the following test:

Put a good chuck on the machine, one that
will hold a piece of stock securely. Does not
matter that the stock runs out a bunch, as long
as it does not shift.

Put a two or three inch diameter piece of
stock in the chuck, so that it extends out
from the chuck by ten inches or so.

Next turn two collars on the stock. One
near the chuck, one at the extreme end of the
stock. Make each collar about a half inch wide,
or thereabouts.

Basically just rough out the stock between
the collars, maybe 1/8 deep or so.

Then take a cleanup cut across the top of the
collars so the surfaces are fully machined.

Next, select a HSS toolbit with postive rake,
sharpened as sharp as you can do it. Don't
round the edge, don't put any radius on the
front.

Use that tool to take a one or two thou cut
across the top of the collar surfaces,
*without* touching the crossfeed dial at all
between the cuts.

Mike the collar diameters and report the
results back here.

Jim

Is it possible to adjust the headstock on a lathe? I have never seen such a thing, which just goes to show how few lathes I have looked at...

If it is possible, then I am not surprised that the designer put the bolts inside the headstock - to keep your cotton-picking fingers off them!

I would have thought that even a headstock that is bolted down on a flat surface (as opposed to locked into the vee ways) would at least be doweled in place?? We are talking about foundational truth here.

I suppose if Deckel made a lathe, it might have a headstock that could be tilted and angled...

Peter S, I admire your knowledge, so please don't take me wrong or get ill at me.
If Deckel made a lathe, it would be maybe something along the lines of the Hembrug A1


Stuff that's supposed to move does, and stuff that ain't supposed to move does not.
Dobber

I could adjust the spindle on my Colchester Lathe so I'm sure a Harrison is similar.The adjusting screws are under the rear of the headstock. Do it the way Jim Rozen and Tyrone describes.As long as the chuck holds securely it needn't run dead true to align headstock.Martin

DD,

Thanks for the Hembrug info, thats a new one for me, looks very nice! "Alignment garuanteed".
I am also most impressed that the company has been around since 1679!!

Mr Fluffy, sorry to get off the topic.

If I read you right, the head has been off several times? I would take it off again before going through the pain of aligning it. Check for any bumps or bruises on the underside of the head as well as the bed, and stone anything you find. Then make sure that it is impeccably clean and remount it, being careful not to bump the iron and introduce any more bumps and bruises.

May seem obvious, but I've seen enough mishandled machine parts to almost bet you got some chips and/or burrs on the contact surfaces.

Peter S, I think the reason they are just bolted on is to minimize damage in event of a smash up, I've had to re-align a lot of headstocks over the years usually because the operator have run into the chuck or had the tool rip the job out of the chuck.
Some do dowel the headstocks on with taper dowels so you can redowel after a smash up. D.S.G. were notorious for hiding the odd bolt behind labelled covers whilst leaving all the others in plain view, O.K. until the labels fall off and get lost.
Thats the thing about engineering, you can be in it all your life and learn something new every day.
Regards Tyrone.

Just a small popint, it ia possible for the headstock to be out of line in the vertical plane, ie the spindle is not paralel to the bed when viewed from the front, do check this before doing anything dramatic. The l5 tailstock has horizontal jacking screws for settting the off set so that is easy to correct, but again take care that the spindle is parallel to the bed in both planes, to find out turn between centres a test bar with the tailstock spindle wound both in and out.
finally I seem to think that the underside of the l5 head has dovetails machined into it so alignment should be automatic, if sois there a bit of crud under there ?

Ok, I just bit the bullet and dismounted the head, and discovered the 4 clamping bolts holding it to the bed were loose, probally about finger tight.
It has some beach marks where it looks like it may have shifted in the past and there are a couple of small scars ill dress up before reassembling it.
For the alignment, it could be very very gently dovetailed, it presents two flat surfaces down to the mounts on the head along the axis of the spindle, it has a "keel" like segment that sticks out down into the bed and on the side of this segment are two screws (although there are 4 holes) one at the front and one at the rear, currently both on the same side. These two screws are positioned such that if you wound them in or out and remounted the head, they would shift the alignment of it to the bed. Would this be the adjustment? It seems a very coarse arrangement. Perhaps I need to try and take a photo to explain more. Ill get to that in the am when I'm not covered in old lathe gearbox oil...