Calculate thread pitch and feed rates...

[John Hill]

My lathe has the usual table of feed rates and thread pitches but it by no means covers the whole range of available combinations, not that that is a real issue as there are more than enough listed to meet my needs.  My problem is the way they are listed as it seems to take ages for me to scan all over the tables to find what I am looking for.

My Chinese lathe has two three-position levers for changing the spindle speed and a single speed belt drive from the motor,  thats easy enough, 9 speeds and the associated table is not too difficult to handle.

The lower gear box that drives the feed screw and another drive rod has four levers (big knobs actually),  one of these levers has just two positions 'M' and 'S' which as far as I can tell select drive to the feed screw or the drive rod.

One other lever has just two positions 'I' and 'II' which I think is only effective when the drive rod is selected.

The feed screw cuts metric threads (as far as I know) and the drive rod (when the feed screw is not being driven) provides power to the saddle for both 'X' and 'Y' feeds.

That leaves just two levers which each have five positions '1'..'5' and 'A'..'E'.

By my reckoning thats 25 combinations on those last two which means 25 metric threads and considering that the 'I','II' lever is effective when using power feeds thats 50 'X' and 50 'Y' feed rates and of course 50 thread pitches in addition to the 'native' metric threads.

All this is before changing gears (cogs).

Has anyone ever tried to write software that would list all these combinations and sort them into some convenient order?  If they havent I might just give it a go...

[Darren]

Good grief, with that lot I'd be confused too.

Didn't it come with a  "manual" 

Sounds like it needs a hardback.....

[bogstandard]

Darren,

The manual that came with mine is perfect bum fodder for the outside khazi.

John's lathe is basically the same as mine, but I have 18 spindle speeds instead of 9, so is a little more complicated, but not all that beneficial, as the range that I use only uses half of them.







I personally, once I had the code broken, have not found any problem getting what I want. But if someone could make it a little easier to find the correct combination, there would be no complaints from me.

Bogs

[John Hill]

John, did you notice any errors in your lathe tables?

For instance, when I compared the 24 to 48 tooth feed rates with the 48 to 24 rates I expected one would be exactly four times the other but on the table they dont always seem to be.  They are mostly very close and can be explained by decimal rounding but some seem to be more than that.

Of course these tiny rounding errors are of no practical consequence but they will somewhat complicate the calculation of intermediate speeds and ratios.

We have a little project going in Saudi Arabia but once I bat the ball back over the net to their side, which will not take much out of the day,  I expect there will be time to think about how to do this.

[bogstandard]

David,

To me, the feeds rates don't need to be that exact as all they are used for is for finish. As long as they are near enough, that is close enough for me.

It is the accuracy of the threading charts that worry me, and I have yet to start to cut any. But when I do, I will do a sample of every thread that can be done, and go over them with a microscope to see just how far out they really are.
I still have the suppliers in my sights if anything is too far from what they state.

John

[John Hill]

Hi John

My first pass at the tables:-

Longitudinal feed rates in MM with cog ratio 48:24
0.162   D2   SI
0.188   D4   SI
0.211   B5   SI
0.219   B1   SI
0.226   A4   SI
0.264   D5   SI
0.264   C1   SI
0.273   D1   SI
0.306   B3   SI
0.317   A5   SI
0.323   D2   SII
0.329   A1   SI
0.332   C4   SI
0.345   B2   SI
0.369   C3   SI
0.374   B4   SI
0.376   D4   SII
0.382   D3   SI
0.396   E5   SI
0.411   E1   SI
0.415   C2   SI
0.422   B5   SII
0.438   B1   SII
0.451   C5   SI
0.452   A4   SII
0.459   A3   SI
0.461   E4   SI
0.517   A2   SI
0.527   D5   SII
0.528   C1   SII
0.546   D1   SII
0.574   E3   SI
0.612   B3   SII
0.633   A5   SII
0.646   E2   SI
0.657   A1   SII
0.664   C4   SII
0.689   B2   SII
0.738   C3   SII
0.747   B4   SII
0.764   D3   SII
0.791   E5   SII
0.821   E1   SII
0.830   C2   SII
0.902   C5   SII
0.918   A3   SII
0.922   E4   SII
1.033   A2   SII
1.147   E3   SII
1.291   E2   SII

I assume a practical thing would be to choose a dozen or so and put them on a reference card, please note I had to do a bit of interpolation and with rounding errors etc I expect some of these that appear very close are actually the same.

[John Hill]

Now I am in a very grumpy mood as due to some error Excel changed all my calculated values to "#REF" when I saved the workshet...

[John Hill]

John, re accuracy of the thread charts.  I assume that is something that can only be determined by measuring the finished work?

[bogstandard]

Thats right John.

Cut 'em to 'by the book' size, then, if possible, try to find a commercial nut, or tap one myself (if I have a tap that size) and see if it fits ok. If it won't fit, that means you will have to make both the nut and bolt by single point threading, which defeats the object, as normally I do single point cutting to make a bolt to match say an old tapped hole. If you can get within a small part of a % on final pitch or TPI, you stand a good chance of it fitting anyway.

But usually some of these cutting charts give such a large percentage error when the thread is cut, they will never fit the correct sized internal thread. That is when you need to start looking for non standard change gear sizes to get to where you need to be.


John

[John Hill]

Well John I suppose we can at least accept the screw cutting thread is accurate?  I can think of ways of measuring the ratio between spindle and feed screw and if I get really interested I could fit a couple of micro switches or Hall effect sensors, one on the spindle and one on the feed screw and with a bit of software I could count the ratios and Excel could do the rest for us.

If we think the feed screw is unreliable I could measure carriage travel against turns of the feed screw but that would require accurately measuring carriage travel over, say, 30cm.

It wouldnt really be such a big job as once I have the first 25 ratios (A1..B5) the rest are software tasks including all the change gears you would care to imagine.

Regarding the usefullness of the feed ratios, I agree with what you say but what annoyed me last evening was setting the parting tool going with the power feed then scratching my head and squinting at the chart trying to find which is the next fastert, combination. I think my initial table is worth it just for that especially if the close ratios are culled.

Meanwhile, I am looking for a nice little brazing torch as everything I have is really too old and they wont fill that type of cylinder anymore.  I need a bit of brazing for my gyroscope project.

[John Hill]

John,  I did some more research today and much to my suprise I learned that inch is exactly 25.4mm,  for some reason I always assumed that was an approximation.

Anyway, as we have a 127 tooth gear and 127 is the lowest integer mulitple of 25.4 I believe the imperial threads should be as accurate as the metric threads.  Of course the issue may be that instead of, say, 20 TPI  the nearest is something else, but at least I should be able to calculate them!

Tonight is Coronation Street so thats only one hour in the play shed this evening, heck!  It is Coro St tomorrow night too, at this rate I will never get anything done!

[John Hill]

John, Mr Bagstandand Sir!  What size change gears do you have for your lathe?  I am ready to plug those into the program.

[bogstandard]

Hi John,

These are my bits.

Paired gear - 127/120

Other gears - 22, 24, 26, 44, 48, 52

Leadscrew - 3mm pitch

It will be very interesting to see how things turn out.
I have also attached scans of my gear layouts for the lathe, to see if they match up with yours.

John

[John Hill]

Bogstandard, this is the image you get when you run that program I sent you..(hopefully).




The first time you run it the list boxes will be empty so you need to load in all the values,  your ever reliable Artful Bodger has already done that but you still need to click on the 'Load' bar,  then wait, and wait, and wait, you will have enough time for a cuppa while it calculates all the combinations possible with your lathe if you had all the change gears you have, plus all the change gears I have, plus all the change gears that are issued with the same lathe in the imperial version.  There a about 2.71 squillion combinations!  So just wait, OK?

It will make a file on your PC so it does not have to calculate them again and when you press the Load button after that you will have only time to blow your nose while it loads them.


You can do two things with Thread Finder, you can put in a desired thread pitch in the top left boxes, only one at a time of course, click the 'Find' button and it will find that thread in the big list which you will be able to scroll up and down.  There are quite a few possibles for most pitches and my idea is that you choose the one that matches the change gears you have on hand, clever eh?

The other thing you can do is calculate the pitches available from a particular setup of change gears, just enter the values in and click the 'Calculate' button, you will get 50 lines in the smaller list which is the 50 pitches available from the combinations of the three change levers.

Notes:

D  is the driving gear close to the spindle

X1 is the intermediate gear that D meshes with.

X2 is the gear paired with X1, or the same gear which is usual when doing metric threads.
    (So if using just one intermediate gear, or just one of a paired gear, enter X1 and X2 as the same)

Z is the gear which X2 meshes with, i.e. the gear on the change box.

Note that metric 1.25 is shown on the lathe table but we cant get it with our gears, the closest is 1.24 but if you really, really wanted 1.25 you could see from Thread Finder that there are a few ways you could get it, for example:


 20.3200"     1.2500mm    A1   D=60   Z=48   X1=120   X2=120   M=1

..so cut yourself a 60 tooth gear (or hit the supplier up for one) and you are there!


1.25 is just one I decided to check on, there could be others like this.

Have fun,  I have not attempted to make the program dork proof so if you ask it dumb questions by putting in daft values it may well spit the dummy.

John

[Darren]

Hi John H, that looks like you've worked hard 

If one could change the leadscrew pitch, could it work for any lathe and gear set?

[bogstandard]

That was quick John.

Now I have the destructions, I will give it a dose of looking at in the morning.

At the moment it definitely looks promising.

I have already  the suppliers, and they don't do any extra change gears for this lathe. So it looks like I need to purchase a few gear cutters.

Bogs

[Twinsquirrel]

John, looks like a great little prog, very complex to program I would imagine... 

 

nice work

David

[bogstandard]

John,

Just thought I would let it do a few intitial setup calcs.
After about 10 seconds it comes up with an 'out of memory' error.

I actually have 4gb in this old steam powered jobbie.


Bogs

[John Hill]

Hi John H, that looks like you've worked hard 

If one could change the leadscrew pitch, could it work for any lathe and gear set?

Hi Darren, well I would have to know quite a few details, fortunately The Bogster and I have the same (near as dammit) lathe so it was easy for his.

[John Hill]

John,

Just thought I would let it do a few intitial setup calcs.
After about 10 seconds it comes up with an 'out of memory' error.

I actually have 4gb in this old steam powered jobbie.


Bogs

Aw well, theres the problem, the calculated table is about 8mb, did you mean 4megs or 4gigs?  Are you talking about memory or disk size?

[John Hill]

That was quick John.

Now I have the destructions, I will give it a dose of looking at in the morning.

At the moment it definitely looks promising.

I have already  the suppliers, and they don't do any extra change gears for this lathe. So it looks like I need to purchase a few gear cutters.

Bogs

The imperial version gets a whole 'nother set of change gears, can you lean on them for a set of those? Maybe a US supplier?

[bogstandard]

4 gigs memory, 700 gig hard drive (spread across 4, size of drive C, 120 gig), so I don't see that as being a problem.

It would most probably be a lot cheaper knocking up extra gears myself. Is yours an Imp machine?

Bogs

[John Hill]

Yep, should be no problem with that configuration, what stage were you at when it karked?  Did it load the values into the big window?  What was the very last thing you did?

No, I have the same change gears as you but I added in the ones listed in the American Grizzly manual for the same machine.

If that program works for you at least you should be able to decide just what gears to make.

[John Hill]

One more thing, did it make a file called 'All Ratios.dat'?

[bogstandard]

I have the program on my desktop.

Run it.

Click 'load'

About 10 secs later it comes up with the memory error.

No folder is generated.

John