PDA

View Full Version : any tips for "turning" a lathe carriage that is facing convex? (scraping)



1200rpm
06-07-2013, 05:03 AM
my lathe project has the worst fit carriage I have ever seen - the cross slide(which is good) is facing convex by about .002-.003" in 6 inches.

this is on my Yamazen Lancer project(the first wave of Taiwanese imports- mid 70`s)
it was apparently that way when new.

the saddle is too big me to mill it, so I am hand scraping the v way of the saddle.

I am about 5 cycles in and have seen no appreciable change yet.

any tips for getting it "turned"?

thanks!

loply
06-07-2013, 06:24 AM
3" is a lot to scrape by hand, luckily the surface area is quite small.

If you're real careful you might find you could use a file to bring a thou or two off it and save you some time.

Are you sure it's that the cross slide ways being out of perpendicularity to the bed, rather than the spindle being out of parallelism with the bed? You would want to be 100% sure about that before you scrape it!

1200rpm
06-07-2013, 07:04 AM
you are absolutely correct!

when i first checked the cross slide to spindle i freaked out- .010" in six inches convex!!

then it dawned on me the headstock is adjustable.:)

that was my incentive to ditch the vari-speed sheave and hook up the direct drive VFD conversion - so i could get it turning straight before i did anything else - it will now turn a 2" aluminum bar to within .0005" over about 10 inches - more than good enough for my needs, now i just have to get it facing flat.

this machine is the reason early imports got such a bad rep! very poorly fitted, but with decent quality castings.

the good news is the bed is hardened and shows no appreciable wear.

if i can just get it facing flat it will be a great home shop machine.

big job
06-07-2013, 07:21 AM
Yes I read alot abought this and its common new out of the box, you have to
go all over everything seems those over the pond none are turn key and have
to be tweaked now and then over time. .0005 is close enough for govt. work,
but then when facing some lathes needs the carriage locked and then some
lathes dont care------ nature of the beast.

saltmine
06-07-2013, 10:53 AM
That was one of the first things I checked on my 7 X 12 lathe when I first got it. It's kind of hard to make anything on a lathe that cuts tapers and faces convex (or concave)

I guess I was lucky. The spindle and tailstock were nearly perfectly aligned (.001" over 10 inches) The lathe bed and carriage were perpendicular and square (.0005" in 6")
I did have the tailstock centerline a couple of thousandths high, but a good cleaning and a couple passes with a file cured that. The machine has held its accuracy for quite a while, even though I have had to adjust the gibs on all of it's sliding surfaces a few times. The only major change I've made was replacing the plastic change gears and drive gears with steel gears. I did find out something interesting during the time I've owned and used the machine. Normally, I used hydraulic oil to lubricate everything. But, last year, after changing the oil on my motorcycle, I had some full synthetic oil left over. I put some in my oil can and set about oiling all of the sliding surfaces....Bad idea. The first problem that arose was the cam-lock on the tailstock wouldn't hold, and the tailstock would slowly back away from whatever it was that I was turning. This didn't register, for some reason. I soon found my carriage saddle backing off, all by itself while facing a part. This led me to fabricate a carriage stop...which soon began to slither down the ways. Eventually, my feeble mind connected the dots. I carefully washed the synthetic oil off of the machine with acetone, and re-oiled it with non-detergent hydraulic oil. Problem solved. Uh, until the carriage drive rack almost fell off...

dian
06-07-2013, 12:58 PM
"I soon found my carriage saddle backing off, all by itself while facing a part."

that sounds weird.

winchman
06-07-2013, 01:50 PM
"....convex by about .002-.003" in 6 inches." So, the surface is spherical (or curved), and the center is closer to the tailstock by .002-.003"?

Seems to me the surface would be conical ( a very shallow cone in this case) if the spindle was not perpendicular to the cross-slide travel.

beckley23
06-07-2013, 02:01 PM
The first thing you want to do is to make sure the scraped cross slide has been properly fitted to the scraped cross slide ways.

Looking at the saddle from the top, you have to turn it counterclockwise, which means that you have to scrape the left inside V slide end and the right outside V slide end. This will get confusing when you turn the saddle upside down, so make sure to mark the ends that you want to concentrate on. As you progress you will start scraping further down to the other ends. Don't forget to scrape the flat slide at the same time, just be careful how much you take off. All the while, you will have to run accuracy checks to keep the saddle level front to back and left to right. It's a tall order, and one of the most difficult alignments to obtain, but if done with care will be easy to obtain.

The entire sequence is shown starting at post #147 in this topic.
http://www.practicalmachinist.com/vb/monarch-lathes/wreck-update-146913/
Harry

winchman
06-07-2013, 02:02 PM
How about removing the jaws from the chuck, and placing a magnetic level horizontally across the center. Put your dial indicator on the cross-slide with the tip against the level as you run the cross-slide back and forth. Rotate the chuck 180 degrees, and repeat. This should tell you if the cross-slide travel is perpendicular to the spindle.

1200rpm
06-07-2013, 03:02 PM
The first thing you want to do is to make sure the scraped cross slide has been properly fitted to the scraped cross slide ways.

Looking at the saddle from the top, you have to turn it counterclockwise, which means that you have to scrape the left inside V slide end and the right outside V slide end. This will get confusing when you turn the saddle upside down, so make sure to mark the ends that you want to concentrate on. As you progress you will start scraping further down to the other ends. Don't forget to scrape the flat slide at the same time, just be careful how much you take off. All the while, you will have to run accuracy checks to keep the saddle level front to back and left to right. It's a tall order, and one of the most difficult alignments to obtain, but if done with care will be easy to obtain.

The entire sequence is shown starting at post #147 in this topic.
http://www.practicalmachinist.com/vb/monarch-lathes/wreck-update-146913/
Harry


this is exactly what I`m trying to do, and is also how I was thinking I should do it... I guess it`s just going to take a while- it`s about 14" along the v way and it is sooo tempting to try to mill off the bulk, but it exceeds the travel of my mill and would take 2 separate set-ups - very easy to mess up, so I`ll stick to scraping it.

better get some more hi-spot! :)

thanks!

Norman Bain
06-07-2013, 05:32 PM
12oorpm:
then it dawned on me the headstock is adjustable.
that was my incentive to ditch the vari-speed sheave and hook up the direct drive VFD conversion - so i could get it turning straight before i did anything else - it will now turn a 2" aluminum bar to within .0005" over about 10 inches - more than good enough for my needs, now i just have to get it facing flat.


Take care when performing the "turn bar over length". It DOES get you the right (repeatable) diameter for that size of bar; however it does not address the possibility that the headstock is tilted high or low at the spindle end.

At the end of the day the headstock alignment to ways setup stands alone (must be done by itself) and is the FIRST thing that should be done when aligning lathe. Once the headstock is aligned side-to-side and up-n-down the tailstock can be altered (aligned) to match.

Now the above said; I started doing my head in envisaging what would be the result of (change to) the concave/convex cut of the headstock being high or low at the spindle end. My conclusion (and I do stand to be corrected) is that high or low will cause the spindle end to be needed to be twisted INTO the cutting tool in order for the cutting tool to travel across the work cutting at constant diameter. Hence high or low is going to result in greater CONCAVE.

12rpm has a CONVEX issue so ... flame me.

Cheers,
Norman

Jaakko Fagerlund
06-07-2013, 11:41 PM
How about removing the jaws from the chuck, and placing a magnetic level horizontally across the center. Put your dial indicator on the cross-slide with the tip against the level as you run the cross-slide back and forth. Rotate the chuck 180 degrees, and repeat. This should tell you if the cross-slide travel is perpendicular to the spindle.
Easier is to just put a mark on the chucks face, zero a dial indicator to it, then rotate the chuck around and move the DI with the cross slide on the mark again. It is now showing you the error in perpendicularity, granted that your bearings are not destroyed.

Richard P Wilson
06-08-2013, 04:39 AM
You do use the saddle lock to lock the saddle to the bed when facing don't you? Not just leave the saddle free relying on friction to stop it moving, or just use the half nuts? I'm not familiar with your particular lathe, but some manufacturers put the saddle lock near the end of the saddle, so if the adjustment of the saddle is a bit slack, applying the saddle lock can cause the saddle to twist slightly on the bed. You can check this by putting the faceplate on, and running a dti across using the crosslide, first with the saddle locked and then with it unlocked. If there is any difference in the readings, then the saddle gibs need adjusting. If you know that the faceplate is flat, then this is also a good check for squareness of the crosslide.

Richard

1200rpm
06-08-2013, 07:04 AM
i have not actually checked a part after facing it - I bought this machine with the understanding that it would need some re-building and other than getting the headstock trued up to the bed by actually turning a piece my checks for the cross slide have been by 2 methods- a parallel held in the chuck in the zero to zero condition, and by a square on pins clamped to the cross slide guiding way - both ways show the same thing.

from what I can gather it was a common thing on the early imports.

dian
06-08-2013, 11:09 AM
so far i have not seen a chuck, that has a face running really true.

"My conclusion (and I do stand to be corrected) is that high or low will cause the spindle end to be needed to be twisted INTO the cutting tool in order for the cutting tool to travel across the work cutting at constant diameter. Hence high or low is going to result in greater CONCAVE."

i dont see how a spinde axis not aligned vertically can lead to convex/concave facing. anybody?

winchman
06-08-2013, 11:38 AM
I just checked the 3-jaw chuck on my late-'70s Taiwan-made JET 1024. The runout at the edge of the chuck face is less than 0.0005". With the dial indicator on the cross-slide, the TIR as the cross-slide is moved across the chuck face is also less than 0.0005". I must have one of the good ones.

If the spindle axis is parallel to the ways, an error in height has little effect. The trig involved is the same for having worn ways near the headstock, which has very little effect.

Having the spindle axis point upward or downward with respect to the ways won't have any noticeable effect on facing cuts as long as the tool (mounted on the cross-slide) is at the height of the spindle axis where the cut is being made. If the tool is too high or too low, the facing cut will be slightly conical with the center higher or lower than the edges.

The critical alignment is having the spindle axis parallel to the ways when looking down on the machine. Trig does nothing to help you there.

beckley23
06-08-2013, 02:04 PM
In the link I provided above, post #99 is the headstock alignment check, This is done in both the horizontal and vertical. Post #175, 6th picture, shows the cross slide alignment check.
Harry

winchman
06-08-2013, 03:39 PM
Rats. I just realized that I left out some zeros in my post above. It should read:

The runout at the edge of the chuck face is less than 0.0005". With the dial indicator on the cross-slide, the TIR as the cross-slide is moved across the chuck face is also less than 0.0005".

J Tiers
06-08-2013, 04:41 PM
If you DO actually need to scrape something like that to correct an angle error, there is a technique to use that I have used with good success.

identify the end that needs more taken off. Mark a spot every inch or so back from that, on a surface to the side of it.

Scrape the last inch at that end. next scrape the last TWO inches. Then the last THREE inches, and so on. A "pass" is a full set of those, ending with the full length. if you have lots to take off, scrape heavy at the end. Check with indicator, don't get upset about flatness just yet (it's OK to check with a reference and modify technique to correct problems, though).

Once your indicator says you are about there, make only full length passes, and use the blued reference for flatness. Continue to scrape "heavy" in the areas that need more work per the indicator.

You can also take a couple more of the "progressive passes", this time scraping lightly each time.

Richard P Wilson
06-09-2013, 04:28 AM
i have not actually checked a part after facing it - I bought this machine with the understanding that it would need some re-building and other than getting the headstock trued up to the bed by actually turning a piece my checks for the cross slide have been by 2 methods- a parallel held in the chuck in the zero to zero condition, and by a square on pins clamped to the cross slide guiding way - both ways show the same thing.

from what I can gather it was a common thing on the early imports.

Hang on a minute. in the first post you said it was facing convex, and now you tell us you haven't actually tried facing a part at all? Even if a vendor swore on to me on a pack of bibles that a lathe had a certain defect, I'd want to check it out thoroughly myself and be 110% certain that it did have that defect before launching out on a lot of work, with the potential of ruining the lathe if the vendor was mistaken.
Try facing something off before you go any further.

Richard

1200rpm
06-09-2013, 04:55 AM
i took no one`s word of anything- my indicators and eyes tell me what needs to be done - facing a part would be pointless- it would rock like granny`s chair- fine for making tops- not so good for making anything useful.
it is progressing(although slowly) just as beckly23 suggested.

the saddle casting probably weighs about 80 lbs or more and I am at least getting a good workout.

1200rpm
06-14-2013, 09:57 AM
whew!!
it took about 30 cycles but it finally turned the corner going fron convex to slightly concave!

luckily, the apron side of the saddle was high to begin with so i don`t anticipate any issues with the leadscrew/feed rod from the drop in the saddle caused by the scraping.
it was really nice to see the gib/apron mounting surfaces come into parallel with the bed at the same time.
i can now lay a precision(.0005 in 10") level on the cross slide(it`s flat) and it is within 1 division in all directions.
i suspect that the saddle was only rough machined and never actually fitted.
by far, this was my toughest scraping job yet!

thanks!

beckley23
06-14-2013, 02:37 PM
It'll be easier the next time.

BTW, 80 LBS is a featherweight.
http://img.photobucket.com/albums/v152/beckley23/se198.jpg
Harry

1200rpm
06-14-2013, 09:53 PM
holy crap, i don`t even want to think about something like that!
my neck and shoulders ache just looking at it!

beckley23
06-14-2013, 10:15 PM
The scraping was the easy part, the slides have Multifil 426, a Turcite type material, epoxied to them. Getting there was a bit more difficult.
http://www.practicalmachinist.com/vb/monarch-lathes/another-new-toy-163406/
Harry