Any hope for metric threads without change gears

Shotgun

Registered
Registered
Joined
Aug 13, 2020
Messages
1,342
Just about every video I've watched describing how to cut metric threads on an imperial lathe involves changing up gears on a banjo that has a couple long slots meant to handle the various sizes of gears.

What do you do if you don't have a slot on your banjo? My Sebastian H model does have a banjo, but it is only for changing the leadscrew direction. The banjo is a solid plate, Each set of the other gears ride on a post, bolted directly into that plate. The only slot, is the short one to allow forward and reverse. I could get very close to the 1.5mm thread I need, if I could swap the 64 tooth sliding gear for a 60. But, the distance between gear centers is locked, so the switch would mean that the gear teeth couldn't reach each other.

Am I missing something?

DSC05028.JPG
 
Sounds to me that your lathe was not made to accommodate metric threading. None were until perhaps 30 or 40 years ago.
 
You need a transposing gear to fit in there.
The optimum is a 127/100, but there are other combos that work.
 
It does not look promising to say the least, My 19" Regal Leblond has a full adjustable banjo and was simple to retrofit with transposing gears (1944 vintage). Yours is more complicated with the non adjustable center distances and the two range gearing on the feed box, it does not look that there is sufficient room to do transposing gearing. With mine it was necessary to fabricate a new gear case/guard to accommodate the larger gears.
 
There are several transposing gear combinations that will provide an "almost" but usable reduction. The 127 tooth is as close as you can get, call it absolute. . . What you are after to start is a stack, on the same stud, of two gears in the threading train. A 52 and a 44 tooth get pretty close. You simply find a gear that is a filler in the train. Input to one and output from the other. I can't tell exactly where to fit it, but the gear that feeds the large gear in the lower right looks like it would do. All that is needed is a stud long enough for 2 gears stacked. The banjo may need to be extended or retracted, but there is no need for another stud.

.
 
I can't tell exactly where to fit it, but the gear that feeds the large gear in the lower right looks like it would do. All that is needed is a stud long enough for 2 gears stacked. The banjo may need to be extended or retracted, but there is no need for another stud.

.
Interesting.
I could remove the reduction gearing that is stacked up on the left. Just get them out of the way. That leaves me with the spindle, and idler, and the sliding gear (which is the input to the QCGB). Then I just need to find a gear to mate with the 64T idler. I need to keep that one to maintain the set distance between it and the sliding gear. The one mating with the 64T just has to change the ratio enough that one of the QCGB combinations hit 1.5mm with under 1% error.

Easy peasy.
 
On my SB9, the transposing gear goes between the stud and screw, no idler.
 
While my lathe will cut metric threads by changing the gears, I have found that Imperial TPI works in most applications. I'm too lazy to set up the gears for one off parts. I just keep cutting until the nut will screw on without binding. The thread form is of course not perfect, and sometimes I make the root slightly wider by adjusting the compound.

These are the TPI that I use

The formula is (1/M)*25.4
M ThreadTPIClosest% Error
0.35​
72.57143​
72​
0.787402​
0.45​
56.44444​
56​
0.787402​
0.5​
50.8​
52​
2.362205​
0.6​
42.33333​
44​
3.937008​
0.7​
36.28571​
36​
0.787402​
0.8​
31.75​
32​
0.787402​
0.85​
29.88235​
28​
6.299213​
1​
25.4​
26​
2.362205​
1.25​
20.32​
20​
1.574803​
1.5​
16.93333​
16​
5.511811​
1.75​
14.51429​
14​
3.543307​
2​
12.7​
13​
2.362205​
3​
8.466667​
8​
5.511811​
 
Thanks, @JimDawson
I'm going for the 1.5mm, as I'm wanting to make an ER32 chuck. The Sebastian is an oddly threaded spindle (2.125"x8tpi), so I've no hope of buying one off the shelf at a price I can afford. Have you tried 1.5mm? Will 5.5% error be close enough? I've yet to find anything that points to how close is close enough.

OTOH, if I could get my hands on a 68T gear to mate with the 64T idler gear, it would drop my leadscrew rate by just under 6%. Then I could hit the 1.5mm almost on the money.
 
3D printed gears work fine.
My transposing gear is printed.
 
Back
Top