Lathe Refurbish Help...

[jxxxoxxxe]

I wanted to ask for help here before trying any of the actual machinist boards...

I'll try to keep it brief...



I got a free Victor 1640G lathe a year or 2 ago, i believe its from the late 70's. Got it setup, wired in, and bought some tooling...

The best finish I could produce out of it is below on the left...

Replaced all three spindle bearings and could make the finish on the right of that part. The bearings were badly pitted...1st problem solved...


Next I started trying to thread some things, but kept splitting the threads when using the threading dial. Also wasnt getting the correct TPI the machine was set for...Someone had installed a 20 tooth gear on the lead screw when it was supposed to be a 50 tooth. Got one ordered and put on and made this in practice....

16 tpi, checks out with a gauge...2nd problem solved...


Now to the current problem...Any work I've done on the lathe to this point has been mostly practice, or something that required no real specifications or tolerances...Its time for that to change as I have a project to complete in the next 6 months...The chuck didnt seem to be centered at all. I believe it uses a D1-6 cam mount.

With an indicator on the side of the chuck I got .034" of movement/runout...

I thought I'd take the chuck off and basically start all over from even checking the bearings tightness to the concentrically of the cam plate.

this measurement ran to about .006" out

this measurement was also a ways out....

I believe the only adjustment of the bearings is here in the center where the orange support runs though...Whats the best way to set the bearing tightness?

Rather than me wasting time trying to figure it out on my own. Where/how should I proceed from here? I've got a manual for it, but its a reproduction and very bad...

Should I hope for .001 or less movement on the mounted plate? When mounted shouldn't I want the same from the chuck?

Is there a certain way the chuck should mount to the plate, i.e. point x to point x? Or is it trial and error to find which cam suits which hole the best?

Is there a certain technique to tighten the cams? skip every other? tighten them in 1/4 turns?


Any ideas are helpful. I am all self taught on this, so maybe I am just missing something. I am ready to be making more chips....

 

[2oolhound]

Are they timken bearing? Otherwise they will be preset to tolerances. Were they an interference fit for installation? If you run the lathe for 20 minutes it should heat up and may make a difference. Also you will get more accurate readings if the dial guage is perpendicular to centre and 90' to the surface. It's angled on the photo so the reading will be less accurate.

I'm no expert but hopefully some of the whiz kids will chime in soon as there are a few of them around here. How about a photo of the whole lathe? It's an interesting machine.

 

[454ragtop]

What class of bearing did you use? Should be super precision, expensive bearings that require extremely careful assembly. Never done it myself, but I'm told there are witness marks on the races that need to be aligned. Best to ask someone with experience on the Practical Machinist board, this isn't a home shop grade machine, so you should be welcome there.
HTH, Jim

 

[GeneralDisorder]

If you are certain the bearings are correct and installed correctly then you will almost certainly have to true up the chuck backing plate. Never worked on a cam-lock unit like that but its standard practice on a threaded spindle machine to true up a new backing plate or an old one after spindle repair work. Basically you just recut the face and the C flange for the chuck. Once that is done the chuck will usually be true once its mounted.

Someone may have trued it up with bad bearings and now you have to correct it.

GD

 

[jxxxoxxxe]

I'll try to get a better measurement tomorrow....

I bought the bearings from the manufacturer. They were NSK, Koyo, and a Nachi branded...a little over $300 total. I didnt notice any witness marks when replacing them...I think I still have the old bearings so I will check them for marks...





EDIT: after just going out and looking at the old bearings, I saw no witness marks....


I do think maybe generaldisorder is onto something though....after looking closely at the chuck, it appears that maybe someone did try to true it, as the lines on it somewhat resemble the lines on my bad bearing example part from above...

Also you can tell in these 2 pics that someone has attempted to bevel the front edge of this piece rather crudely. Its completely uneven and I could probably be more consistent with a electric grinder...

Before I go just trying to true up the plate I definitely want to hear any more ideas...could that really be it? Is it possible to mess it up more by trying to true things up. I'd hate to have to buy a new chuck, although I could go for a 4 jaw....

 

[rsanter]

First off, do not measure the outer runout where the cam locks are, that is not a precision or centered surface, measure at the ring or step just in from that OD. Measuring at the face is correct.
Also do not measure the OD of the chuck as again it is not the precision surface but so measure at the chuck face

You may need to grind the chuck jaws

Bob

 

[Bolster]

Interesting discussion. Subscribed, thanks to OP and contributors.

 

[DocsMachine]

I hate to be a potential bearer of bad news, but I suspect the spindle is bent. What may have happened at some point in the past, is somebody had a "crash"- ran the carriage or tool into a chuck, had the tool "dig in" and rip the workpiece out of the chuck, whatever.

As another poster noted, the center "stub" is the key surface- that's actually supposed to be a short precision taper that indexes to a matching short taper socket in the backplate of the chuck.

Somebody has, as you noted, tried to repair the damage by turning the damaged bits back to shape- but it clearly wasn't done well, and the center stub needs to be precision ground rather than turned anyway.

At this point, it looks like pretty much everything is damaged. There's really no "half-assing" it; you either need to fix it right, or just slap it back together as-is and just use it for rough work.

If it were me, I'd turn a tad more off the stub of the spindle- just a little- and press on an oversized ring of new metal, probably 4140, and probably heated, pressed and left to shrink fit.

Then gage the center of the chuck backplate. It's probably out of round and oversized to boot. The metal's thin to the locking pins, so you can't go far, but it needs to be turned- or better yet, ground- to a new, reasonably precise taper. This will likely have to be done on another lathe.

Remove (and probably replace) the locking pins and grind the face true as well.

Then, back at the bent lathe, use a toolpost grinder to true the face of the spindle, the face of the stub, and then using the backplate as a gage, very very carefully regrind the new metal ring to match the taper of the backplate. You'll want it to seat firmly on the taper, and juuuuust barely make contact with the face.

Finally, once the chuck's replaced and properly seated, regrind the chuck jaws so they're decently concentric.

The more careful you are in doing all that, the more accurate and repeatable it'll be when you're done. As I said, it'll be tricky, but it can be done. And it won't take much cash, just time and care.

Doc.

 

[454ragtop]

Probably should see if the morse taper inside the spindle is concentric and work from there.
Jim

 

[A_Pmech]

Hi Joe,

The spindle nose has obviously been "modified" while the machine had bad spindle bearings as evidenced by the tool marks on the spindle pilot and the shape of the pilot nose contact marks. It also appears they cut back the spindle face judging by a couple of the tool marks. The chuck is no longer seating on the pilot and spindle face correctly and that's a big source of your runout. The runout of the spindle pilot and face are also adding to the problem as they were machined with bad spindle bearings.

One way to solve the problem is to re-machine the spindle mount to bring it back to standard dimensions. However, when you re-machine it you'll have to take a cut on the mount face to increase the pilot nose diameter to standard which will alter the datum length of the retention stud even more than it already has been. In other words, you'll have to shorten the retention studs on every chuck you own, or countersink them deeper into their backing plates. That's a minor problem though.

A tool post grinder and good care will be a necessity for final finishing. Grind back the spindle face, mount your chuck and take a series of feeler gage measurements, then calculate how much the taper needs to be ground down. Use a tenths indicating DTI to measure your cross slide travel and keep kissing the taper until the taper and face of a new chuck backing plate print well with blue.

I'd rather do the work in the machine than send the spindle out, but you might have to send the work out if you don't feel comfortable doing it yourself. If done in the machine the spindle is machined in it's own bearings which will result in a much more accurate spindle nose, especially if the spindle has been damaged.

P.S:

Remember that surface finish plays a large part in indicator readings. The poor surface finish on the spindle parts will cause swings in the indicator beyond the actual reading. Having the indicator at an angle relative to the spindle also causes a "cosine error". The indicator reading will be less than the actual displacement when the indicator is not perpendicular to the axis of rotation.

 

[454ragtop]

Since the bearings were relatively cheap, might inquire about a spindle, never know.
Jim

 

[jxxxoxxxe]

little update....here's 2 short videos of a measurement on the face of the spindle, and one just on the inside edge of the taper...

<iframe width="420" height="315" src="http://www.youtube.com/embed/mHHj32RWYbg" frameborder="0" allowfullscreen></iframe>

<iframe width="420" height="315" src="http://www.youtube.com/embed/FwIudpVMCDg" frameborder="0" allowfullscreen></iframe>


I also put an indicator on top of the spindle mount and attempted to pry up on it, but got no movement at all...guess the bearings are fine...


This is looking bad to me...almost .020" of runout inside the taper seems awful. For it to be that bad inside, doesnt it pretty much have to be bent?

Any other ideas to confirm it being bent besides taking it out of the machine and putting it in V blocks...?

I guess I'll call and price a spindle....this free thing isnt going very well...lol

 

[930dreamer]

Someone told me, nothing is ever free. I hope you can fix it. Any pics of the whole lathe?

 

[jxxxoxxxe]

Someone told me, nothing is ever free. I hope you can fix it. Any pics of the whole lathe?

Sure is looking that way...i hope i can fix it too...i've been reading a ton, and it seems like it usually takes something majorly catastrophic or stupid to bend a spindle. There are no signs I see on the ways, compound or cross slide that indicate a bad crash...that's not to say no one ever lifted it by the chuck though...from many threads I've read elsewhere, a bend may be rare...

I'll get a full pic of it up this weekend. The place I got it from bought it used many years ago. I want to take it out of my shop for a decent cleaning. It's got that dirty machine shop film all over....I just need the weather around here to cooperate...

 

[jxxxoxxxe]

Finally got around to getting it ready to be cleaned up. It helped that it was almost 70* here today. Here's a pic of it sitting at the edge of the shop...don't mind the mess, had to move and stack a lot to get this out as it sits in the back of the shop...

should be able to get it pretty clean tomorrow...






Update on the spindle....

I called Victor, a new spindle is $1500....thats probably out...


I don't have a tool post grinder, nor am I quite sure I could complete the job being such a novice, although I may have to try...Is there a reason a grinder must be used? Is it just for the finish, or is the spindle hardened or something...


After a week of research there does not seem to be anyone in the state of Oklahoma that grinds spindles. I got a recommendation about a place in Michigan that travels, but they dont do lathes. I also found a place in Ohio called Locher that I think grinds them, or a place in New Hampshire called Spindle Parts & Service...I am waiting to hear back from both...anyone ever had it done?

 

[DocsMachine]

I don't have a tool post grinder, nor am I quite sure I could complete the job being such a novice, although I may have to try...Is there a reason a grinder must be used? Is it just for the finish, or is the spindle hardened or something...

-A little of both. The spindle face, and especially the central locating taper, should be reasonably hard, for wear purposes. And also, the grinder will give you a much smoother finish- also important for a precision surface.

Doc.

 

[rsanter]

There is an easy way to fix this and put nit dead on. But there is a sacrifice

Get a D mount lathe backing plate and a plain back chuck
Install the backing plate and face it true to the machine.
Face a small counterbore that tightly matches the OD of the chuck.
Drill the backing plate for the chuck and mount the chuck.
Install a piece of true round material in the chuck and indicate it to be sure you are now tuning true. If it is running true then you turn the remaining flange ( the backing plate will be bigger OD than the chuck) to match the od of the chuck.
Be sure to punch a mark on the spindle flange and the chuck mount so you alwaysut it back in the same spot.
My big lathe I know has been crashed a couple of times and the spindle I know is bent. This is common on old lathes. If you crach the machine then you reface the backing plate and reinstall the chuck again

You loose the use of the collet hole unless you use emergency collets where you bore the collet to match your need.
Or you do whatbi do and I have a 5c collet holder that I can grab in the chuck

Bob

 

[KenS]

Keith Fenner is a marine shafting specialist, but also versatile machinist who has a very popular YouTube Channel. He also seems to be very helpful. It may be worth a call to him to see if he could rebuild your shaft, or perhaps turn a new one for you at less than factory cost. His current project is a giant woodchipper shaft, but in several of his videos he works on shafts the size of the one in your lathe.

If you watch his videos you will see cases where he adds missing material to shafts-- such as that which has been ground off yours-- using brazing or stick, mig or tig welding depending on the application. He can flame straighten and then re-machine the shaft to factory specs.

IMO your lathe is easily worth the $1,500 investment in a new shaft--as long as the ways, gearing and other components are in good condition-- especially since you got it for free. But reworking the existing shaft may be an alternative for consideration.

Keith Fenner
Turn Wright Machine Works
791 Main Street
West Dennis, MA 02670‎
(508) 394-0724

turnwrightmachineworks.com

 

[Nelson58]

Keith Fenner is awesome, and he has a job shop in New England. You also might try Miller Machine works, and see if they can fashion you a new spindle. It would be worth the money.

Nelson

 

[jxxxoxxxe]

For those that are following along...LoL...


I pulled the entire spindle out today...only took about an hour...

I've found a place, http://www.centerline-inc.com/ that is located in Ponca City, Oklahoma. They've been in the spindle repair business for 20+ years, albeit mostly fixing things much more expensive than this.... I've talked to one of the owners a few times now and I am going to run it up to them next week and let them evaluate it, and possibly repair it...I hope it just needs to be reground...


The only thing that looks bad to me is 2 wear spots on the spindle...


This is at the center bearing where the preload is adjusted....

This is at the headstock bearing....

You cant feel anything in either spot, but it has obviously worn through the initial plating a little...

The actual bearing and races I installed last year still look good....

 

[930dreamer]

Keep us updated on the spindle repair.

 

[jxxxoxxxe]

Update...first off, I appreciate everyone's advice and help...

talked to a few spindle repair places, and took the spindle to one repair shop, and they all said the same thing....can't be repaired for less than I could get a new one...

So....I thought, its already basically trashed, I'm gonna try something...

I decided to weld the nose to add material so I could machine it back to spec. I didnt really have anything to lose... It was already .100" under the needed 4.188" large end diameter that I needed...I didnt really want to have to cut the face back to gain material to make the nose taper diameter come back to spec...



Nasty welds...I actually had to weld it up twice to gain enough material back to machine..

So to figure out the correct taper angles for my compound slide, it was quite easy. The D1-6 taper is 9/16" (.5625") long.

Hello trigonometry... So over that .5625" with the 7* 7' 30" angle, I would need .0703" of movement on the indicator from my straight edge...Got the compound set where I got a repeatable .0703" movement over the specified length, and I was good to go...



You can actually see some taper now...not near as flat as earlier pictures..

Blued nose taper...

Transfer to backing plate when pulled tight with the cams

I still need to machine the front edge to pretty it up...


I think I'm happy for now. The blue transfer seems good, and it took a hit or 2 with the rubber mallet to unseat the taper.

So now, how should it hold up? hopefully the welded area stayed somewhat close to the original hardness?

 

[454ragtop]

Looks good, bet it holds up just fine to moderate use. Be interesting to see what you get for chuck runout now.
Good luck, Jim

 

[jxxxoxxxe]

Looks good, bet it holds up just fine to moderate use. Be interesting to see what you get for chuck runout now.
Good luck, Jim

on the taper, I obviously get none...I took the back plate off the chuck to use as a gauge for bluing and to make it more manageable. I took a quick measurement on a spot on the inside of the back plate that centers up on the bore of the chuck. I was getting about .003" there. I'm going to try each cam hole line up to see where I get the best reading.

With the chuck installed, where will be the best place to check run out?

 

[DocsMachine]

Nice work so far!

BUT... you're putting way too much blue on the mating parts. You don't want a thick coating, like paint- quite the opposite, you want as thin a layer as you can get.

Try this instead: Clean off all the blue (and of course any dust or grit) and color the spindle nose with a Sharpie or ElMarko. Then install the chuck again, and remove. Look for where the marker has rubbed off- the rubbed spots tell you where the contact is.

Marker ink is very thin- I've heard anywhere from 0.0002 to 0.0005"- and easier to use than prussian blue. You should also be able to see some of the transferred ink on the inside taper of the backing plate.

Doc.

 

[jxxxoxxxe]

Nice work so far!

BUT... you're putting way too much blue on the mating parts. You don't want a thick coating, like paint- quite the opposite, you want as thin a layer as you can get.

Try this instead: Clean off all the blue (and of course any dust or grit) and color the spindle nose with a Sharpie or ElMarko. Then install the chuck again, and remove. Look for where the marker has rubbed off- the rubbed spots tell you where the contact is.

Marker ink is very thin- I've heard anywhere from 0.0002 to 0.0005"- and easier to use than prussian blue. You should also be able to see some of the transferred ink on the inside taper of the backing plate.

Doc.

Good point doc... I'll try it again tonight with a sharpie... Just for reference, what's the best way to apply the Blue correctly? I was spreading it with my finger...

 

[Trey T]

Awesome thread. More photo, the better.

I hope to get into machining in the future.

 

[Steinmetz]

Good point doc... I'll try it again tonight with a sharpie... Just for reference, what's the best way to apply the Blue correctly? I was spreading it with my finger...

TP. Or facial tissue.

 

[Steinmetz]

BTW, Excellent effort by the OP. I really enjoyed following his journey.

 

[DocsMachine]

Good point doc... I'll try it again tonight with a sharpie... Just for reference, what's the best way to apply the Blue correctly? I was spreading it with my finger...

-The finger's fine, it just needs to be a thinner coat.

For a situation like this, what you want to do is clean both surfaces thoroughly- and that also means, if necessary, lightly stoning down any burrs or dings in the "master", the backing plate.

Then smear a very light trace of blue inside the backing plate. It should be as streak-free as you can get it (but don't obsess ) and translucent. Enough color you can see it, but not so much it gets opaque.

Then mount the backing plate as you did, separate, and see where the blue has transferred to the part you're testing, in this case the spindle.

If the tapers match, you'll get a full surface of contact. If it's too steep or shallow, you'll get a "ring" at the peak or the base of the taper.

And the thinner your marking compound, the more accurate you can match the taper. That's why I suggested the marker ink- not only is it easier to use, in my opinion, but pretty much guaranteed to be very thin and consistent.

However, marker works differently than bluing. With the marker, you're looking for where the ink has been removed by the contact between the parts. With the blue, you're looking for where the color has been transferred by the contact.

Same idea, same essential result, different method. Either way, though, contact is relatively obvious. Also, don't be afraid to blue/mark and check multiple times, cleaning and reapplying between tries. You want it to be right, so it's worth checking twice.

Doc.

 

[A_Pmech]

PLEASE tell me you grounded the spindle nose and didn't clamp the ground to some other part of the machine.

The repair work looks very good though, especially for an new turner. (Use less blue though.)

 

[DocsMachine]

PLEASE tell me you grounded the spindle nose and didn't clamp the ground to some other part of the machine.

-He clamped the ground to a drill chuck mounted in the tailstock.

Doc.

 

[bullnerd]

Nice lathe,I like the large dials.

I would have done what Doc mentioned early on with a new "ring" shrunk on.

What you did looks pretty good just be careful putting the chuck on and off.

nice lathe!

Wait ..is that a VCT floor?

 

[jxxxoxxxe]

-The finger's fine, it just needs to be a thinner coat.


Doc.

I want to try the marker idea anyway...I blued the nose taper first, then it transferred to the backing plate. I wonder if it matters which one is blued...

BTW, Excellent effort by the OP. I really enjoyed following his journey.

Thanks, hopefully more than just myself learned something...

Nice lathe,I like the large dials.

I would have done what Doc mentioned early on with a new "ring" shrunk on.

What you did looks pretty good just be careful putting the chuck on and off.

nice lathe!

Wait ..is that a VCT floor?

I thought about doing it that way, but the nose wouldve need to be machine flat, then material found and machined, then heated and pressed, then machine to the correct tolerance anyway. Welding it just seemed easier to me, although it might not be the correct way to repair it...

And that is VCT that is due for a serious cleaning...its been down and uncleaned or waxed for over 4 years now...

PLEASE tell me you grounded the spindle nose and didn't clamp the ground to some other part of the machine.

The repair work looks very good though, especially for an new turner. (Use less blue though.)

Come on, give me a little credit...LoL...If I told ya I grounded it directly to one of those new bearings I put in there, would it make you cringe...?









I'm going to get the chuck put back on sometime this week. From everything I've read a 3 jaw will never run true anyway, hopefully it'll be close enough to turn some internal threads and make threaded caps that fit correctly...

 

[DocsMachine]

I wonder if it matters which one is blued...

-Not really, but keep in mind you're looking for contact points on the workpiece. If you blue the work, you have to see where it marked the master, locate those positions on the work, and then change the workpiece as needed.

However. if you blue the master, you look for the points it contacted the work, and can see where the changes need to be made directly.

Clear as mud, right?

Generally speaking, blue the part you're not going to be changing- the "master"- and look for where it contacts (and thus transfers blue to) the part you will be changing- the workpiece.

But since marker ink is dry and doesn't transfer in any significant manner, in that case you want to mark the workpiece, and look for where the master removed it by making contact (and thus rubbing it off.)

Really, you can blue or mark either piece, but as described above, you remove the extra step of mentally transferring the contact point from one to the other- you don't have to look at the contact point on the master and figure out where it was on the work.

From everything I've read a 3 jaw will never run true anyway[...]

-Generally speaking, a typical 3-jaw won't run true- meaning the centerline of the workpiece isn't necessarily the centerline of the spindle rotation. That's just the nature of the beast- no matter how accurately the chuck is made, it has to have some tolerances in order for the scroll and the jaws to move, and they don't always return to exactly the same spot.

But what you DO want is that it turns true to the bed. In other words, the chuck might not be holding the work precisely centered on the axis of rotation of the spindle, but it DOES need to turn it parallel to the bed.

If the spindle is "bent", it will hold the chuck at an angle to the bed- say you chuck up a two-foot-long bar; it's centered to the spindle axis at the chuck jaws, but at the far end, points several inches to the left or right of the tailstock.

It's easy to deal with a chuck that holds things slightly eccentric to the spindle axis, but it's a real pain to deal with one that holds the work at an angle.

Doc.

 

[darkk]

We're having a similar problem with our 1903 Pratt & Whitney 10x29 toolmakers lathe. The runout is about 0.015" at about 4" out. I have disassembled both our 3 and 4 jaw chucks and scrubbed them pretty clean but it didn't help much, they are a little beat. Our backing plate runout is about 0.003" out. We probably won't find any replacement parts for this either...glad we bought it cheap!

 

[bullnerd]

"I thought about doing it that way, but the nose wouldve need to be machine flat, then material found and machined, then heated and pressed, then machine to the correct tolerance anyway. Welding it just seemed easier to me, although it might not be the correct way to repair it..."

Yeah ,hard to make parts on your lathe for your lathe if the chuck dont fit! Nice job BTW.


And that is VCT that is due for a serious cleaning...its been down and uncleaned or waxed for over 4 years now..."

How do you like it? Any pics of your shop here anywhere?

 

[jxxxoxxxe]

well I ended up mounting the chuck today...I also dressed up the spindle sides and put a chamfer on front of the nose taper to make it look more like the original probably did...

on the mating surface of the back plate I was getting about .0015" of movement on the indicator..hopefully thats good enough...

I then made this little part to hold a bearing, and threaded it for a 1/2 x 13 nut. the threads looked a little visually rough. I dont know the reason behind it though...may couldve been because i built it from a piece of 2" rebar...lol

bullnerd, here is what it looked like when it was first done...its held up pretty good. I've had to replace 3 or 4 tiles, but its really easy...

 

[jxxxoxxxe]

Any thoughts on grinding the jaws? Would it be beneficial? Or only worthwhile for the diameter I ground them at? I need to get a test bar...

 

[Nelson58]

Love the floor...I'm afraid of white because of trying to keep it clean...

Nelson