Lathe alignment - Emco Super11.

Got a test bar cut down for the funky 1/2 MT5 spindle and found the head out about 5 thou. in the X axis, y is pretty good.

From the drawings the head is fixed with 4 studs, would it be locked in on the ways? Or just bolted to the flat part?

Anything other that the head being out to cause this amount of misalignment?

Did you level the bed to remove any twist?

If it has 4 studs, you can loosen them and bump the spindle it. Most have a pin or spud in the middle that the head rotates on.

ewlsey said:

Did you level the bed to remove any twist?

If it has 4 studs, you can loosen them and bump the spindle it. Most have a pin or spud in the middle that the head rotates on.

Click to expand...

Yup, levelled the bed with a Starrett master level. Diagram doesn't show any middle pin, just the 4 studs...

Does the headstock fit over the V ways or is is just mounted to a flat surface? If it fits the V ways, you can't adjust it without scraping. If it's on a flat surface you can move it.

.005 is a lot.

ewlsey said:

Does the headstock fit over the V ways or is is just mounted to a flat surface? If it fits the V ways, you can't adjust it without scraping. If it's on a flat surface you can move it.

.005 is a lot.

Click to expand...

That's what I was wondering, the diagram in the manual doesn't show it very well. Muata been crashed or dropped somewhere in it's lifetime.

Guess before considering anything I would give it a good tug to see no quick movement to suggest a take-up needed. Then mark the test bar high and low, mark it to the lathe , give the test bar a 180* turn and test again.
Examine the test bar fit to the lathe perhaps blue it up..
Still, .0003 may be normal for that lathe.(?)

The five thow may be just loosening clean and re-fit, a bug or piece of junk under the head. so that needs to be cleaned checked and decide what to do.

It doesn't have a Y axis. No problem.

[Anything other that the head being out to cause this amount of misalignment?] the test bar not fitting properly.
Head being out for a number of reasons.

Testing with a ball test indicator to the top side the error can be actual height, and then more or less with going in or out on the radius of the test bar. Same with the horizontal, part of the error can be going up and down on the radius.
Yes run out is run no matter how you look at it.

and .005 is way too much to be just radius off center..

For top one can test as you did, then bring the cross forward to find the center and the highest place close to the head , then do the same at the far to find the high there. I like to test a lathe with my dial indicator and a flat nose point of about 1/8 inch, with that being square to the top or to the horizontal it takes out the error of running off center on a radius.

tdmidget said:

It doesn't have a Y axis. No problem.

Click to expand...

up and down from the bed?

OP, are we all on the same page with axis - on a lathe Z is spindle direction, X is the cross feed, is that what you reporting on?

where and how exactly are you getting the run out? Emco's are well made lathe, and it looks in good shape and .005 is a hell of a lot......leveled, no way it came with .005 runout....so there is either extreme twist, a botched repair/crash or error in how the measurement is being done.

Also, you don't really need a perfect fitting test bar, just something straight and the same diameter....indicating at different places along the bar and comparing the readings will tell where the spindle is pointing. to explain, look at it the other way.....if you got zero runout rotating the bar a different places, and .005 along its length.....then yeah the headstock is out. BUT if you just a reading along the Z and have also indicated by revolving, no way to tell if its spindle or work that is misaligned.....point is assume the test is out and compensate for it and you be perfectly checking spindle alignment

I really hate the use of xyz on manual machines. Long travel and cross make more sense

michiganbuck said:

I really hate the use of xyz on manual machines. Long travel and cross make more sense

Click to expand...

He's following the right hand rule so it's all good.

Where things really get confusing is rotary axes and parallel axis. Like a boring mill has X, Y, Z, B (rotary), and W (quill). A mill turn like an Integrex has X, Y, Z, C (also the lathe turning spindle), and B (swivel for milling spindle)

Thanks all for the suggestions, I got it!

The headstock is simply bolted down to the flats on the ways, not on the dovetail - plus there's no center pin. I re-levelled it to be sure and ended up putting a bit of 0.002 shim stock under one of the bolts and with some careful re-torquing got it very close. It looks like someone tried to "align" it at some point and missed the mark quite a bit.

Dunno how else you could really do this except with a test bar and with the funky 1/2 MT5 spindle I had to cut one down to fit.

BTW, there's the same 0.0003" runout in the spindle itself.

You were doing great until you mentioned the 0.002" loose shim. A machine of that quality and condition shouldn't need any shimming. Two wrongs don't make a right in engineering.

Regards Tyrone.

Tyrone Shoelaces said:

You were doing great until you mentioned the 0.002" loose shim. A machine of that quality and condition shouldn't need any shimming. Two wrongs don't make a right in engineering.

Regards Tyrone.

Click to expand...

I think he meant a shim under one of the feet (floor to machine) to help level the machine.
The headstock is on flat surfaces no, .005 towards/away from operator, should only need a bump over. No shims required.

Demon73 said:

I think he meant a shim under one of the feet (floor to machine) to help level the machine.
The headstock is on flat surfaces no, .005 towards/away from operator, should only need a bump over. No shims required.

Click to expand...

It's ambiguous at best. This was the quote.

The headstock is simply bolted down to the flats on the ways, not on the dovetail - plus there's no center pin. I re-levelled it to be sure and ended up putting a bit of 0.002 shim stock under one of the bolts and with some careful re-torquing got it very close.

Click to expand...

I can't tell from that, if he got the level "very close" or the alignment of the spindle.

Surely the lathe base to the concrete floor, levelling bolts are fully adjustable. Why would you need to shim to the concrete?

Every picture of one I can dredge up has levelling bolts in the base. http://i42.servimg.com/u/f42/15/42/79/68/super110.jpg

We have pictures of a Starrett 0.0005"/12" Level fairly dead nuts. Surely that's not getting done with shims under the feet. And how do you carefully torque them. There just jacks.

I wouldn't discount a shim under the head off what we have to work with.

Regards Phil.

(On Edit). I should watch the Video's first. At post #13 above, there is this video. Emco S11 - alignment 2. - YouTube

Pay attention to the 20 second -> mark. There's a 2 thou shim shovelled under one corner of the head stock.

Did either of you watch the video? He shows he shimmed the headstock Lol.

Terry.

YouTube served me up the 3rd part of your series. I don't believe that's here yet. >>> Super 11 - YouTube

You need to brush up on your test bar procedures. Run out and alignment are two different things. There will always be run out on a test bar. You measured 3 tenths at the nose, then set your zero at the low point, you mention it at the 40 second mark, you even said "not sure if its the best way to do that"

The spindle axis exist's but it's invisible and buried in the centre of that test mandrel. For the purpose of checking alignment, you eliminate run out by finding the mean average of it. So you had 3 tenth's. You should set your clock to plus / minus 1.5 tenths, so zero is in the middle of the run out.

When you moved out your 300mm, you were getting greater run out. That's to be expected. You mentioned 5 -6 -7 tenths. I'm seeing 4 to 12 tenths on your clock. A difference of 8 tenths. (Video seconds 48-55).

You have 8 tenths run out, but the mean of that is 4 tenths. If you averaged out your run out at the nose, and called that zero then averaged out the run out @ 300, +/- 4. 4 tenths,that's your axis mis-alignment. That was with your indicator on top of the bar. Dr. George S probably spec'ed 0.02mm per 300. 4 tenths is 10 micron. Your at half spec across the top of the bar.

I'd also note, Its pointing up, the only way the specs allow it. It can never point down. Pointing up allows for sag and deflection on the work piece.

Your clocks must work reverse of mine. Maybe that's the way it is down under, but every Mitutoyo clock I own spins C.W as it comes on the needle. @ 1:35 you show that 5" thou horizontal run out, but then said it's toward the operator. That 5 thou is OFF the clock. The mandrel needs to come towards you, to fix that. That's the test where you have the indicator on the side of the bar. If it's truly only sitting on flats, bump it around until its true to the saddle axis. If you get excited, you could tune it up to the tailstock at the same time.

If I where you, I'd piss that single shim off, and start again, even if it meant lifting the head, giving it a good stone, clean and start again.

Regards Phil.

When you have the headstock correctly aligned you will need to align your tailstock spindle/barrel to coincide with the headstock axis.

Regards Tyrone.

.oo5 is a lot of error for what looks like a near new lathe. .002 seems to correct the top but how is the front and how does it line to the tail. how does a true re-cut head center and a tail center turn a part? so the lathe may not be ready to run.

The head has to be aligned straight on the top top and at the front so a collect or chuck held part runs straight. then it has to be also center to the tail stock so it might also run a part between centers.

Head bearing first- make sure they are pre-load tight (simple tug a held round stock test). next, check the test bar is running zero checked in a V block to an indicator on the bench. Check the head run out and mark it for high and low. Then check the test bar fits a blue-up to the head. With finding error now with a flat nose dial indicator to top and front and to the tail, then a blue in check that the head is a proper fit to the bed.

Yes this is assuming the bed is good with not checking that first but one would have to pull the head to examine the bed is not twisted from being dropped or something crazy.

If the head does not fit well the torque of hold down may throw it off.
Does the book give torque specifications? If not look up the normal for that size bolts.

Actually you need to blue-in check the tail also because a mis fitting tail being over tight may put a twist in a bed.
Might want to do all the head checks with the tail loose.

.0003 error in the head seems a lot for a new or neasr new lathe.

QT: Tyrone :When you have the headstock correctly aligned you will need to align your tailstock spindle/barrel to coincide with the headstock axis.
*Oh. I See Tytone coverd this. Oh well might as well cover it again.