Accuracy or lack off - what to expect

[Admiral_dk]

At the risk of teaching granny to suck eggs, have you been round all aspects of the saddle, cross slide and top slide and ensured all is tight?

That was how I improved it a bit in the first post (without telling how) - the nut wasn't as tight to the slide as it should have been.

Besides with my bad memory I might have forgotten the obvious .... so I'm not offended by any remarks on this subject ....

[simplyloco]

If your lathe has a vee bed then crabbing should not be an issue when the tool is under load. Recognising and overcoming backlash is part of a machinist's general competence. I would suspect slop in the cross slide - i.e. loose gibs, but without a photo or lathe ID your question cannot be answered to anybody's' satisfaction.
John

[Admiral_dk]

P.S. 1. As you get older and have sufficient finances behind you, you may not want to continue to be one of the "Make do and Mend" generation  .

Oh I wish - but that is not the way my economical situation has been going since I lost the job I had @ Aarhus Universitet in 2014. My current job just about covers my expenses and I will actually be better of on the public pension when I'm old enough to retire @ 67 in a little over 7 years time (I will have finished last payment on the house by then).

[Admiral_dk]

I have tried to measure the rise in height off the Cross Slide, over the Prism part of the Bed, when I try to push it to opposite sides in the front and back and I see 0.06mm. on the clock and absolutely zero if I do the same with the clock over the flat rear part of the Bed.

Thanks for the suggestions so far, but I have another appointment for the rest of the day and will be back in the shop tomorrow.

[steamer]

I have tried to measure the rise in height off the Cross Slide, over the Prism part of the Bed, when I try to push it to opposite sides in the front and back and I see 0.06mm. on the clock and absolutely zero if I do the same with the clock over the flat rear part of the Bed.

Thanks for the suggestions so far, but I have another appointment for the rest of the day and will be back in the shop tomorrow.

Show us a good picture of your tool as well....a tool that skips along and digs in sounds dull or of wrong geometry.

Dave

[Jasonb]

Or above ctr height.

Heavy cuts may be pulling it down to correct height.

[Roger B]

This is the back and forth play in my cross slide, ~0.3mm. I don't seem to have the problems you describe. With my old minilathe most of the difficulties were due to play in the headstock bearings. When I replaced them all was well 

<a href="https://www.youtube.com/watch?v=fEcYPNtgwF8" target="_blank">http://www.youtube.com/watch?v=fEcYPNtgwF8</a>

[Vixen]

Hello Per,

Do you experience the 'dig ins' during power feed or manual feed? It could be the carriage is lifting when power feed is applied, due to wear or loose gibs or whatever.

A razor sharp tool should give you no problems, if everything it tight and well adjusted. A razor shape tool will provide the minimum of force to deflect the carriage or slides.

There is a simple test to see if the cutting tool is sharp enough. Scrape the cutting tool across your thumb nail in the cutting direction. A 'sharp' cutting tool will cut a small curl of nail swarf. A 'dull' cutting tool will simply slide (skid) over your thumb nail.

Sharp tools give least problems.

Good luck

Mike

[steam guy willy]

Hi All chucking in my pennyworth here.. I always thought when I was younger that the "old boys" were successful with their turning because as the lathe started to wear out there increased weight would counteract this as they leant on the slides in the correct fashion ??!! when the new younger chap tried to use the lathe he would not know about this and wonder why his accuracy was all over the place !!!!  There are quite a few tricks from the old days the turners used ..one was to wrap a piece of stout rope around the chuck to stop chatter when parting off ?.

When I made a new crosslide nut for the Southbend lathe that had 3 mm back /front play I made the nut extra long with provision to take out any  future "play", a bit like the Pultra way of doing this. Pics...

willy

[john mills]

i would check that the saddle is contacting on the angular faces and not touching on other places some times with
lot of wear the saddle can be touching parts of the ways its not meant to stopping good contact and guide from the
angular faces that are meant to guide the saddle.  some of the problem sound like the shape of the tools not having sharp cutting edges with enough clearance to cut well.often older machines have lots of clearance in the
nuts and screws and don't give it much thought you just take up the slack traveling in one direction.it is easy on indusial CNC machines with good sized preloaded ball screws but old ones can still have problems after a few
 crashes which resultant damage.most of the time travelling in the one direction once the slack has been taken up
word ok.     
         John   

[Mcgyver]

What kind of Backlash do you find acceptable / expect ?

Hope I don't step on toes with this....but backlash really doesn't matter!

Why?  There has to be some, else the thread wouldn't turn.  Backlash is clearance between the thread flanks and even on a new feedscrew is going to be enough that it will mess you up if the backlash is taken up by say pushing on the crossfeed.  So, sans an antibacklash arrangement, we know no matter what we will have enough BL that exactly how much there is is immaterial.  5 thou or 50, it won't much matter.

Where backlash (imo) gained its status is in inspecting used machines were it can be an indication of overall wear.  Lots of backlash is bad not because it makes a difference in how far the tool is advanced, but because it suggests an well worn machine

Whats going on with yours?  The carriage might be crabbing as suggested, or the overall fit of the dovetail might be loose enough that vibration is moving it.  I was doing some internal grinding the last few weeks and with vibration of the high speed motor everything wanted to move.  Tightening the gib worked for me, but you don't it always really tight or it will wear more quickly.

You can check for crabbing by taking something known to be round and cylindrical, centreing it in the four jaw (centre in two planes!) to a 10th of a thou.  Then with an indicator on the tool post, sweep the work in a few spots.  If you get different readings, the motion of the carriage is not parallel to the lathe's axis (either crabbing or leveling).  Crabbing should be identifiable to apply a force the carriage by hand and seeing how the needle moves.  This little set up btw is really great for tweaking leveling on fussy work.

When I'm doing really fussy work, like striving for a 10th to fit an expensive spindle bearing say, I get rid of the compound and mount an indicator on the cross slide dovetail.  My indicator mount also has two cylinders with springs to press the crossfeed screw against the correctly flank of the nut.  You have to hold your tongue just so, but I can usually do better than .0002" with this set up.

Its not trivial getting a perfectly working machine...its why the good quality ones cost so much when new and masochist like me spend so much time reconditioning (scraping) them.  So don't feel bad or get frustrated if it takes a bit of chasing.  It is however very when they do perform well.  If none of this helps, the next thing I'd do get it apart and start check fits with some blue.  Some sort of spring arrangement loading the system against the nut might be an experiment to try - see if that fixed it.  If it does, imo its further evidence the crossfeed dovetails need attention.

Here's a couple shots of what I mentioned

everything is solid and scraped to fit, compound removed and replaced with a T slot plate



Cross feed indicator - digital is great for this, micron accuracy and easy to switch between imperial and the french system (haha couldn't resist), err metric



Here the sweep tool.  Its essentially a spindle with very accurate AC bearings.  I mostly use it in conjunction with reconditioning (second photo) but its easily mounted in the tool post for this sort of checking.  btw, in the second picture is a Holbrook B8, marvelous lathe and a current project

[steam guy willy]

one way of taking out the backlash is to tie a piece of string to  the cross slide and with a  heavy weight and dangle it over the back of the lathe ??!!! never tried it but just an idea ??

willy

[steamer]

one way of taking out the backlash is to tie a piece of string to  the cross slide and with a  heavy weight and dangle it over the back of the lathe ??!!! never tried it but just an idea ??

willy

Agreed!

[steamer]

What kind of Backlash do you find acceptable / expect ?

Hope I don't step on toes with this....but backlash really doesn't matter!

Why?  There has to be some, else the thread wouldn't turn.  Backlash is clearance between the thread flanks and even on a new feedscrew is going to be enough that it will mess you up if the backlash is taken up by say pushing on the crossfeed.  So, sans an antibacklash arrangement, we know no matter what we will have enough BL that exactly how much there is is immaterial.  5 thou or 50, it won't much matter.

Where backlash (imo) gained its status is in inspecting used machines were it can be an indication of overall wear.  Lots of backlash is bad not because it makes a difference in how far the tool is advanced, but because it suggests an well worn machine

Whats going on with yours?  The carriage might be crabbing as suggested, or the overall fit of the dovetail might be loose enough that vibration is moving it.  I was doing some internal grinding the last few weeks and with vibration of the high speed motor everything wanted to move.  Tightening the gib worked for me, but you don't it always really tight or it will wear more quickly.

You can check for crabbing by taking something known to be round and cylindrical, centreing it in the four jaw (centre in two planes!) to a 10th of a thou.  Then with an indicator on the tool post, sweep the work in a few spots.  If you get different readings, the motion of the carriage is not parallel to the lathe's axis (either crabbing or leveling).  Crabbing should be identifiable to apply a force the carriage by hand and seeing how the needle moves.  This little set up btw is really great for tweaking leveling on fussy work.

When I'm doing really fussy work, like striving for a 10th to fit an expensive spindle bearing say, I get rid of the compound and mount an indicator on the cross slide dovetail.  My indicator mount also has two cylinders with springs to press the crossfeed screw against the correctly flank of the nut.  You have to hold your tongue just so, but I can usually do better than .0002" with this set up.

Its not trivial getting a perfectly working machine...its why the good quality ones cost so much when new and masochist like me spend so much time reconditioning (scraping) them.  So don't feel bad or get frustrated if it takes a bit of chasing.  It is however very when they do perform well.  If none of this helps, the next thing I'd do get it apart and start check fits with some blue.  Some sort of spring arrangement loading the system against the nut might be an experiment to try - see if that fixed it.  If it does, imo its further evidence the crossfeed dovetails need attention.

Here's a couple shots of what I mentioned

everything is solid and scraped to fit, compound removed and replaced with a T slot plate



Cross feed indicator - digital is great for this, micron accuracy and easy to switch between imperial and the french system (haha couldn't resist), err metric



Here the sweep tool.  Its essentially a spindle with very accurate AC bearings.  I mostly use it in conjunction with reconditioning (second photo) but its easily mounted in the tool post for this sort of checking.  btw, in the second picture is a Holbrook B8, marvelous lathe and a current project

Nice set up!    Are you documenting this rebuild anywhere?....Love to follow along!....Seems we share a similar sickness....LOL

Dave

[steam guy willy]

To prevent crabbing on flat bed lathes my ford type would it be advisable to remove the middle part of the saddle slides front and back ?? and if so how much would one remove along the length from the middle ??

Willy