Pultra model P

[JasonB]

On JasonB Said:

[…]

1% of 25thou pitch is 0.00025″ and 5.5/25 of that is 0.000055″ that is half a tenth of a thou or 0.0014 mm which is 1 1/2 microns. Wish I was able to measure that let alone work to it.

[…]

A nice ‘reality check’ there, Jason

It also says a lot about the original capability of the Pultra

… see my reference to the comparable Levin

MichaelG.

.

Edit: __ for fairness, let’s also note Derbyshire:

http://www.fwderbyshire.com/accuracyfirst.html

[duncan webster 1Participant @duncanwebster1]

As a Pultra doesn’t do screw cutting, a bit of pitch error probably doesn’t matter, but why not make it as good as you can? The original Pultra was a quality machine.

I noticed the pitch error when I made a really long M6 thread to work some kind of indicator for a friend. It was noticeably different to  standard bolt over 50mm or so, but at 1% that would be 1/2 a pitch, so you’d easily see it.

Jason’s suggestion is good if you can find a drawing.

[JasonBModerator @jasonb]

I was thinking of just using the main leadscrew so you have plenty to turn and adapt at the end (assuming direction is correct.

bit about the accuracy of their screws.

Yes the effect of pitch error is dependant on the length over which it is measured if only the amount a tool needs to move to make the final cut after checking the work then not really worth worrying about.

[Michael GilliganParticipant @michaelgilligan61133]

Looks like a ‘must buy’ bargain !

.

 

.

MichaelG.

[david bennett 8Participant @davidbennett8]

in error

dave8

[david bennett 8Participant @davidbennett8]

Mark, re tapping percentage, start with the tightest thread, then work looser if necessary. I doubt if there’s any info on this for leadscrew fits. But please let us know what you find out.

dave8

[mark smith 20Participant @marksmith20]

I had a bit of time today and tried a couple of combinations.

I first drilled a piece of sae660 bronze for a nut   with a drill to give an approx. 90% thread  also a piece of steel bar with same.

Then threaded two pieces of silver steel .

First one with bronze nut i opened up the split die ,made a thread followed by another pass with the split relaxed a bit. This thread is pretty pointy in profile with not much land at the peak. It turned outto run smoothly but had a tiny amount of movement in the hand when grabbing the nut and trying to wiggle it about .

The second one with steel nut ,only made one thread pass with die opened up.

This thread left more land at the peak of the profile and has a slight appearance of acme thread. This one was slightly tight but after running the nut back and forth a couple of times using the jog function on my southbend 9A it slackened off enough to allow the nut to move freely without any back and forth movement  or looseness. Seems pretty ideal . I prefer the slightly less pointy ,lessdeeply cut look of the thread.

Which profile would wear faster ,the pointy  or the one with more land at the peak???

I intend to use the silver steel with bronze nuts on the final pieces.

 

 

 

[david bennett 8Participant @davidbennett8]

The pointy thread sounds like the poorer thread form. I would go with the other one.

dave8

[mark smith 20Participant @marksmith20]

Dave 8 , i assume the more pointy is because its more true to the UNS thread form or more fully cut. The flat on the peaks is supposed to be 0.125 of the pitch whereas the flat in the trough is supposed to be 0.25 of the pitch.

The less pointy (one with steel nut on) appears to have more or less equal flats. Same thread just not fully cut.

[david bennett 8Participant @davidbennett8]

The wiggle movement on the pointy thread sounds bad. Have you compared the finished shaft diameter to the specifications? This could indicate if and where the thread is in error

dave8

[mark smith 20Participant @marksmith20]

Heres the funny thing , the precision ground silver steel measures 6.34mm according to my calipers . The first one i did with the die  opened up and then relaxed a bit ,( two passes of the die, the pointy thread)  has a finished diameter of 6.32 mm  ,it wiggles very slightly.

The second one with only one pass with the die opened up with the grub screw measures a full tenth of a mm less ,around 6.23mm??  This one feels better with no noticable movement or wiggle.

Does a die deform the metal  some what like a rolled thread??  I  would have expected the diameters of the finished threads to be the other way around???

Im pretty novice at thread tech.

[JasonBModerator @jasonb]

It is just the basic fact that if you used thread wires to measure your two threads cut with the same die you would get two different readings.

The pointed crest one will have a smaller reading as the die was closed up, the “flat topped” one will have a larger reading as the die was opened up

It is the flank to flank contact that matters and the larger reading will mean the flanks of the male thread are closer to those of the female

Yes you can get some extrusion

[mark smith 20Participant @marksmith20]

Jason, i think i follow you. What i dont understand is if in both cases there is still some flat at the top of the crest shouldnt  both outside diameters still be the same  as the unthreaded bar??? Or does thois only happen with single point threading on the lathe.

Heres a picture of the UNS thread form  plus tapping drill sizes i gotfrom the guhring calculator for 1/4″x40 .

[mark smith 20Participant @marksmith20]

Also better photo of the test threads

[mark smith 20]

On mark smith 20 Said:

…
Does a die deform the metal  some what like a rolled thread??

 

Yes.  A die isn’t a good way of making a long precision thread because cutting tends to twist the job, whilst the limited number of supporting teeth in the die allow it to wander off axis and/or at an angle.

A good compromise is to make life easy for the die by pre-cutting a partial 40tpi thread in the rod so the die only has to correct the form.   The lathe produces a precision 40tpi helix for the die to follow, but is less satisfactory getting the form right.  The die is good at getting the form right, especially if it only has to take light cuts.  Lathe and die complement each other.

In industry well-made lead-screws aren’t cut on a lathe or with a die.  I believe the preferred method was to roll a blank and then finish by grinding for accurate form, pitch and polish.   Rolled threads are stronger and harder than cut threads, and  grinding provides high accuracy.  Requires specialised machinery, which is why it’s better to buy lead-screws rather than DIY.

How good does a lead-screw need to be?   Rule of thumb says the screw should be 10x more accurate than the lathe is aiming for.  But, as we know, a skilled operator can get good results out of a badly worn lathe by compensating for it’s errors.  Similarly, we would like a lead-screw to resist wear and tear for donkey’s years, and a soft home-made article won’t stand up to persistent hard work.  However, a poorly made lead-screw could be perfectly acceptable for several years in a lightly loaded home workshop owned by a chap skilled in getting the best out of his slightly wonky lathe.   If Mark is happy with the result, that’s fine, even if it would give Dr Schlesinger apoplexy!

🙂

Dave

[david bennett 8Participant @davidbennett8]

Did they start off the same diameters? You cannot expect a perfect die-cut thread over this length. It would be usual to turn down the diameter by about 0.002″-0.003″ on !/4″ diam. before threading and maybe run a fine flat file over the threads afterwards. Don’t forget, you are just seeking a good, useable thread. It sounds like you have one.

dave8

[mark smith 20Participant @marksmith20]

Dave and Dave8 ,thanks for your help and advice. I dont need extreme accuracy with my work ,just something a bit less sloppy than the leadscrews i have on now. If these work ok ,then if they wear quicker or whatever i can easily remake them.

The thread on the pultra screw is only 100 mm long. I made this simple talstock holder to hold the 13/16 size  die. Mounted on a used blank end MT2 arbor i found in a drawer. I drill out the centreof the arbor a bit  for clearance for a 1/4″ bar. So the 100mm thread length is pretty well supported .

I just threaded a small portion of bar with the die opened up abit more than initially and got this in the photo, looks a bit less pointy than the first attempt and more true to the UNS thread form. It wasa little tighter initially but running the nut back and forth ,it now seems pretty perfect in regarded to no play and ease of movement.

Dave i do have a full gearbox on the Southbend so cutting 40 tpi isnt a problem but ive always had bad experience cutting on thin stock  . But may give it a try sometime for an initial pass before using the die. The half nuts on my southbend  lathe are almost like new.

Dave 8 , yes same diameter silver steel ground bar.

 

 

[JasonBModerator @jasonb]

The other issue with using a die and a tap for that matter is they are generally made to cut threads to suit typical fixings so are a bit loose on tolerance which is not what you want for a backlash free lathe screw & nut. No doubt Pultra adjusted the tolerances to something closer to ideal to suit their use possibly with specially ground taps or single point cutting and adjustment of the thread rollers or whatever was used to make the screws.

If you look at this chart ( imperial is similar) the * marks the typical 6H and 6g tolerance that a tap and die will cut but you want closer to the exact size line down the middle. You can’t do much about the tap cutting a little over so you are having to compensate buy spreading the die to what could probably be beyond the line and that puts the part of the die that should be rounding the crests beyond the OD of the material.

The overall diameter is not the critical part it is how close the < and > of the thread form fit. This is why most leadscrew nuts adjust backlash by spreading the two halves of the nut apart lengthways as you will get the flanks running closer together without the risk of a crest rubbing in a valley.

[david bennett 8Participant @davidbennett8]

Mark, using a travelling rest on your Southbend should enable threading thin stock.

dave8

[mark smith 20Participant @marksmith20]

I have  those fancy telescoping southbend travelling and steady rests but as they were in mint condition and quite expensive with big resale value i dont like using them for threading , i used the travelling ones whilst making replacement cross and compounds screws for my southbend and they were mangling up the jaws . Also the way the travelling steady mounts on the rear of the cross slide is not very practical for short lengths of bar.

[david bennett 8Participant @davidbennett8]

I presume you were also using tailstock support?

dave8

[mark smith 20Participant @marksmith20]

yes

[david bennett 8Participant @davidbennett8]

Ir sounds like your threading tecnique on the Southbend needs a re-think. A thread like the Pultra’s should be a doddle. I have pre-cut threads like this on a Unimat 3 just hand turning the headstock. I would look at your cutting tool and speed. i,e sharp and slow.

dave8

[mark smith 20Participant @marksmith20]

Dave8 , i havent done an awful lot of threading and have done any for around 7 years,so yes my technique needs improving.

Anyway im making 3 complete sets of sae 660 bronze nuts for the pultra,just for the fun of it. 6 front nuts, 3 longer spigot shaped ones for the cross slide and 3 shorter ones for the the compound slide.

All going well so far except whilst aiming for an easy push fit for the 3 longer spigot nuts i ended up a few thousands under size ,so they are currently sitting getting nickel plated to hopefully better the fit . I use my own recipe for the plating solution but its based on the Watts recipe.

 

[mark smith 20Participant @marksmith20]

progress so far plating worked ok. Any tips on accurately cross drilling the 6 at the top of the photo?

[Author]

Posts