Cutting metric threads on an imperial lathe

[Karl_T]

I'm sure you are right.

I am plenty happy with the leads possible with just the 37 and 47 tooth setup . Only two are very far off at all 1.739 for the 1.75 or 0.6% and the 0.612 for 0.600 for 2% (I'd reject this one). All the others are closer

 

[mcmental]

hi there, sorry for dragging an old thread up but i have a harrison l2 and i have previously managed to cut a 1.5mm thread on it but im scratching my head trying to do a 1.25mm, i have the magical 127 gear!! can anyone shed some light maybe please? my leadscrew is 4tpi, heres a few pictures of the charts on the lathe

 

[Karl_T]

Just an educated guess. Put your metric set in, then set the lathe to cut 8 tpi and see what you get

 

[MrWhoopee]

Just an educated guess. Put your metric set in, then set the lathe to cut 8 tpi and see what you get

On my SB Heavy 10L (8 tpi leadscrew) with transposing gears, that will get a 2.5 pitch. Setting the QCGB to 16 tpi gets 1.25 mm pitch. I believe those same settings will work on the OP's lathe.

 

[Downunder Bob]

My lathe is a full imperial machine. Imperial lead screw, norton box etc, but it also came with some change gears as well including 127 and 120 T gears so it can cut true metric threads it also has a metric thread chart, so is ready to go.

Because Australia changed from imperial to metric fairly recently, only about 50 years ago many new machines sold here are set up this way.

 

[oldscouser]

Hello mcmental.

You need to work out the required speed reduction between the lathe spindle and the leadscrew. For a 4tpi leadscrew and a 1.25mm pitch thread the reduction is:


Reduction = 25.4 / (1.25 x 4) = 5.08 exactly

Looking at your thread chart I see that position A of the gearbox gives a reduction of 2. I suggest you try to use a 40T gear on the top shaft and a 80T gear on the bottom shaft and a compound 100/127 on the intermediate. (the 40T top gear drives the 100T gear and the 127T of the compound gear drives the 80T bottom gear)

This will give an overall reduction of 5.08:

O’all reduction = gear box red’n x 80/40 x 127/100 = 5.08

If the gears do not fit in the envelope you will need a different combination.

 

[ian999]

The best way to cut metric threads on an imperial machine, or any other threads is to throw away your change gears entirely!
Make yourself an Electronic Lead Screw (really a gearbox) where a microcontroller does the work of controlling leadscrew speeds.
See the thread "Asian Lathes & Mills/Electronic Lead Screw". My comments and pics are at post #457.
At times the project was very frustrating but now it gives me lots of freedom.

 

[rjs44032]

There is another decent combination 36/34

Error per thread for the given combinations at 2.5mm:

36/34 - 1.16 micron

47/37 - 0.53 micron

80/63 - 0.31 micron

While the last two options are closer approximations, the 36/34 is the most economical. These gears are standard and available from most vendors for fraction of the cost of the other 2 options. Additionally, the 18/17 is even less expensive than those. However, the 36/34 will be easier to compound with bushing. Moreover, the error is still less than 0.00005" per thread

One other observation I have learned doing this research is that the error for all approximate transposing combinations increases as the thread size increases. Conversely, the error decreases as the thread size decreases. All these are still acceptable for the ISO thread specs up to 7mm.

Final observation is that the combination does not have to be set up on the same axis. For example, you could place 37 on the stud and 47 on the gearbox and run the division gears as the compound.

Best Regards,
Bob

 

[evan-e-cent]

The Myford lathe is included in my online computer program for calculating gear trains: RideTheGearTrain.com.
This is a hobby for me and I am in the process of re-writing some of the code so let me know if you come across any bugs! Thanks.