Inserted cross slide feed nuts

[Michael GilliganParticipant @michaelgilligan61133]

Good result, Andy

… it all seems obvious [albeit perhaps not very well made] now.

MichaelG.

[not done it yetParticipant @notdoneityet]

…. For future reference  …..

Agreed. Pressing or pulling is definitely potentially far gentler than hammering.

Edited By not done it yet on 24/06/2020 15:25:30

[old martParticipant @oldmart]

Good result, now you have to decide what to do next, there is very little chance of finding a new one. If one is made, it would be safer to make it a just gentle push in fit and use Loctite, the cast iron is so thin on one side that there is a risk of cracking if the nut is tight.

[Andy CarlsonParticipant @andycarlson18141]

Yes indeed… thoughts are turning to making a new nut.

Materials… I'm sure I've seen plenty of discussion on various forums, not all of them entirely relevant to an 80+ year old lathe. From memory the front running options seem to be brass (on the basis that it will wear in preference to the screw), leaded gunmetal or colphos 90.

Replacing the screw is also on the cards at some point. If I keep a square thread form then I can do one job at a time. If I was in the US then I'd probably buy some 3/8 x 12TPI LH ACME threaded rod and a matching tap all of which seem readily obtainable over there but not so easy/cost effective to get over here?

At the moment the plan is to use 3/8 silver steel to make a square thread profile tap (or is it three taps?) to match the existing feed screw. Then I can replace the feed screw as and when the mood takes me.

I will likely be doing this job several times over so making taps to suit should be worth the effort… if I can do it.

[Brian MorehenParticipant @brianmorehen85290]

Hi Andy,

Great news Success now all the fun and work starts now

Good luck Brian

[old martParticipant @oldmart]

You might have some luck googling Kingston Engineering.

[Andy CarlsonParticipant @andycarlson18141]

Posted by old mart on 24/06/2020 18:20:35:

You might have some luck googling Kingston Engineering.

I checked HPC but theirs start at 0.5 inch. Kingston too. No doubt I could ask about a custom one but presumably that would be pricey.

Mc Master Carr does a 6ft length of 3/8 x 12 LH for 14 dollars in the US. HPC want 63 quids for a 1/2 inch one half as long in the UK.

I havent so far found any taps for a sensible price on this side of the pond either. Maybe there are some but I havent found them.

[BazyleParticipant @bazyle]

Locktite will make future removal difficult if the rapairer doesn't know it was used. You might consider boring out the rest of the length so the new nut can be longer – last longer. Maybe add a second screw to secure it.

[old martParticipant @oldmart]

I used a metric trapezoidal leadscrew and nuts on an imperial machine. The pitch is 3mm, that is close to 0.118", so a crosslide scale with 118 divisions was made for the lathe. The worn out original was 1/2 X8 ACME.

Tracey tools has ACME taps, rh and lh, but the 3/8 is 10tpi, and new leadscrews would be required. You will probably find that both nuts and matching threaded rod are the only way to go.

[Andy CarlsonParticipant @andycarlson18141]

Yes I really want to stick with 12 TPI in spite of it being a bit silly.

The cross slide (the proper one, not the spare) has a nicely made brass resettable dial using the usual 80 graduation fudge. I'd be really loathe to lose that or make it so that the scale was meaningless.

[old martParticipant @oldmart]

How do you intend to produce a new nut in 3/8 x 12?

[Andy CarlsonParticipant @andycarlson18141]

Posted by old mart on 24/06/2020 21:41:42:

How do you intend to produce a new nut in 3/8 x 12?

I think DiogenesII was correct yesterday. The hole is not concentric so the insert needs to be pressed into the casting and the hole and thread made in situ. So whatever thread I choose will need to be tapped I think.

Can I make a square thread tap (or three) myself?

No idea but I will give it a go. Worst case I will press the thing out again and go with a different option.

[old martParticipant @oldmart]

To find if the thread in the nut is concentric with the od, all you have to do is screw the nut on the leadscrew and see if it runs eccentrically. I would bet on it being in the same centreline, these big threads have a way of producing optical illusions. What is the OD of the nut?

[Andy CarlsonParticipant @andycarlson18141]

Posted by old mart on 25/06/2020 16:37:09:

To find if the thread in the nut is concentric with the od, all you have to do is screw the nut on the leadscrew and see if it runs eccentrically. I would bet on it being in the same centreline, these big threads have a way of producing optical illusions. What is the OD of the nut?

Good point.

So I went and checked… but I think it really is not concentric.

OD measured at between 0.505 and 0.511 depending on where I measured it (and that was avoiding the ends and the area around the grub screw hole).

[old martParticipant @oldmart]

I assume that you took it off because it is so worn that the backlash is unacceptable. When the thread is worn, the wear may well be biased to one side, if there are sideways forces, or even gravity. I realise just from seeing the dreadful design (only an idiot would bore a casting so close to the edge) that you worry about the concentricity, but would it matter if the replacement was made properly?

Don't worry too much about my ranting about the quality, I have seen the guts of a Tom Senior mill, and the quality of that particular example leaves much to be desired in design and the dreadful casting quality.

Edited By old mart on 25/06/2020 20:33:57

Edited By old mart on 25/06/2020 20:34:28

[Andy CarlsonParticipant @andycarlson18141]

Yes I was thinking that it might be worth just making a concentric one using single point threading… although even then I will need to make a tool to do a square thread inside a 7.4mm (I think) hole. The front plate on the cross slide is hardly high tech – just a flat steel plate with the hole for the feed screw and a couple of 1/4 BSW fixing screws. The 'aim' judging from the old nut seems to be about 0.5mm left of centre.

Whether an internal threading tool is a sensible idea is debatable. I've seen opinions expressed elsewhere that 3/8 is below the size at which one whould switch to tapping.

I agree the design is pretty odd. I don't think Faircut had really got the hang of the whole lathe thing when they made this but one can't be too critical – mine has lasted 80+ years and is still usable… plus they have provided me with a way to refurb the cross slide nuts – a job that would be a more difficult proposition on a lathe with the threads cut straight into the cast iron.

The nut in the spare is on the verge of a complete thread breakdown which is actually quite useful because I now have no qualms about messing with it. The proper one is not so bad but there is still significant backlash in my cross slide and if I can make it better then I would very much like to do so… after I use the spare for practice.

[old martParticipant @oldmart]

It did cross my mind that a twin start thread of two 24tpi either Whitworth or 60 degree form would be easier to produce a in the limited space. Of course, a new leadscrew would also have to be made at the same time. Just making a new nut will highlight the uneven wear on your leadscrew.

Doing the maths, I was surprised to find that for 12tpi, the dials would have to have 83.3333 divisions per turn, that is total madness if thousandths of an inch were needed.

[Andy CarlsonParticipant @andycarlson18141]

Yes 12 TPI is an odd number when you put a dial on it but I dont think that the Faircuts were the only lathes to use 12TPI.

The usual fudge is a dial with 80 graduations… which is what my lathe has. It seems pretty well made (and resettable) but some of the graduations are a bit wonky so maybe not a factory item. Anyway, whatever its origins, I like it.

I say the 'usual fudge' but in practice most 12 TPI lathes probably had no dial at all. If one is resorting to chalk marks on the handwheels or whatever then 12 TPI is probably not a big issue.

[BazyleParticipant @bazyle]

If you splash out for a bigger bit of brass and make the nut a loose fit and secured by holes in a flange projecting to the front you only lose 1/10 inch on cross slide travel and maybe make a modified front plate or spacer to get that back. When you do make a new screw make it longer and remake the front plate etc to incorporate taper roller thrust bearings.

[Howard LewisParticipant @howardlewis46836]

If you are prepared to depart from absolute prototype, for a closer approximation to a thou per division, then the dial needs 83 divisions. This makes each division 0.001004 inch.

4 millionths of an inch is probably an acceptable error, in an non temperature and humidity controlled environment!

The Myford ML1,2, 3 and 4 also used 12 tpi leadscrews on Cross and Topslide, again with 80 divisions on the dials, giving an error of 0.0004" per division..

Howard

[Andy CarlsonParticipant @andycarlson18141]

I want to keep the existing dial so I will live with the 80 division oddity (if I can manage to make a new 12TPI nut anyway)

TBH it is not a big issue for me. If I want to know how much cut I am putting on it is perfectly fine for that. If I want to turn down to a size then I get the mike out anyway. Quite often the job is metric in any case.

@Bazyle: the order for some LG2 has gone in already so the first attempt will be a fairly like for like replacement. If the new one lasts 80 years then I won't be very worried when it wears out.

[David George 1Participant @davidgeorge1]

I made a new nut for my M type lathe and made a finish tap from silver steel. I roughed out the thread by screw cutting and used the tap to Finnish it.

Have a go.

David

[old martParticipant @oldmart]

If 80 divisions is required, then make a leadscrew and nut of 12.5 tpi, it will be perfect. To do that, you need a lathe which will do 25 tpi, and double the leadscrew speed.

[Nigel Graham 2Participant @nigelgraham2]

For a 1/8-inch lead-screw, I work out 12.5 tpi being feasible with a 32T change-wheel. Pair it with a 50T wheel. (Or 64 / 50 with 1/2 compounding.)

Note you might need keep the half-nuts engaged the whole time unless the lathe has a thread-dial indicator.

I have a small Denbigh horizontal mill, not yet back in service, and one of the screws on that is 6TPI. I forget which but I think it is the long-travel lead-screw!

[Andy CarlsonParticipant @andycarlson18141]

@David: thanks – good to see someone who has actually done it. I was thinking along similar lines. Making one tap seems less hassle than making three.

Did you screw cut a square thread or a pointy one?

Looks like you heat treated your tap – did you just harden it or did you harden and temper?

12.5 TPI… an interesting idea thanks but I probably lack the gears to do that. The Faircut has 2 off 20T and then goes up in 5T steps to 65T. I suspect this is the full original set. It would also mean making the screw and nut at the same time.

FYI the Faircut has a permanently engaged leadscrew..

Cutting a feed screw length thread brings some additional challenges that I'd rather save until another occasion.

For now I'm going to go with 12 TPI and see if I can actually make a nut that works. At least the gearing is straightforward.

[Author]

Posts