Is this a bad idea…?

Advert

New Replies

Report Bug

Help/FAQs

Is this a bad idea…?

This topic has 10 replies, 5 voices, and was last updated 12 March 2019 at 22:10 by Hopper.

Viewing 11 posts - 1 through 11 (of 11 total)

Author

Posts
12 March 2019 at 12:55 #399887
Paul Relf-Davies
Participant
@paulrelf-davies37806
Hi all,

I have a small screw-cutting lathe (similar-ish to an ML7, without the gearbox) and I've been pondering how to improve accuracy, repeat-ability & simplify screw-cutting. Specifically, how to automatically stop the saddle at a the same point on multiple passes.

I did a fair bit of research and have found numerous references & articles talking about de-coupling the change gears from the tumbler with a single-tooth dog clutch.

As far as I can see, the primary facility that these dog-clutch assemblies offer is the ability to dis-engage & re-engage the timing gears while the motor is running. The point (assume) being that the dog tooth in the clutch assembly is stronger than the gear teeth, and is therefore un/less likely to break when the gears are re-engage, while the motor is running…?

So…to my (potentially bad?) idea…

My requirement is to be able to automatically dis-engage the timing gears only. I'm happy to re-set things manually. As such, I wonder if I even need to add a dog-clutch assembly?….Could I just add a trigger to disengage the existing tumbler mechanism at a repeatable point each pass?

I guess the bottom line is this: while I appreciate that there is a risk of damage if the tumbler gears are engaged with the lathe spindle turning, is there a similar risk when dis-engaging them with the lathe spindle turning?

BTW – the reason for this train of thought is that the shaft that the tumbler output gear & the timing gear chain input gear is 1) very short – about 3 gear-widths long and 2) acts as the pivot for the tumbler-actuation lever.

I'm sure I could design something to fit, but it occurred to me that there might be a simpler approach

Any thoughts, much appreciated!

cheers

Paul
Advert

12 March 2019 at 12:55 #19268
Paul Relf-Davies
Participant
@paulrelf-davies37806
12 March 2019 at 13:04 #399889
JasonB
Moderator
@jasonb
The reason the dog clutch works well is that the actual gear train stays in mesh so that when drive to it is re established the tool picks up in the same position. As soon as you start taking the gear train out of mesh you will get multi start threads and a chewed up bit of bar.

Electronic stops also work well if you have braking such as a VFD and the carrage can trip a switch.
12 March 2019 at 13:08 #399891
peak4
Participant
@peak4
Sorry if I've misunderstood you, but by disengaging the geartrain between the spindle and the leadscrew, wouldn't you lose the register to pick up on the next screwcutting pass??

A friend sent me this link the other day as a simple alternative to a carriage stop.

Bill

p.s. looks like Jason and I were typing at the same time.

Edited By peak4 on 12/03/2019 13:09:15
12 March 2019 at 13:35 #399901
Brian G
Participant
@briang
If I understand correctly, the dog clutch doesn't have to be on the spindle to allow re-engagement in the thread, but it does have to have a 1:1 relationship with the spindle and a constant relationship to the leadscrew and one of these wouldn't happen if the tumble reverse is used to disconnect the drive. In this design which takes advantage of the output gear from the tumble reverse having the same tooth count as the bull gear, the clutch is mounted after the tumble reverse which has to stay engaged to maintain its relationship to the spindle. http://www.gadgetbuilder.com/Dog_Clutch.html .

Brian

EDIT: .com not ,com <FACEPALM>

Edited By Brian G on 12/03/2019 13:36:06
12 March 2019 at 13:36 #399904
Hopper
Participant
@hopper
You should google "Graham Meek Screwcutting Clutch" (or dog clutch). It's been featured on here and in MEW magazine and elsewhere.

In a nutshell, no don't disengage gears under drive. Overloads the tips of the last two engaging teeth. You will only get away with that so many times.

The dog clutch both absorbs the shock of engagement without damage and ensures the gear train re-engages in the one and only correct position.

But really, if you are not doing a lot of specialized screwcuttng like cutting metric on an imperial lathe etc, you would be better off learning to use the halfnuts and the thread chaser dial in the conventional manner. It just takes a bit of practice.

If on the other hand you want to really get into the nitty gritty of it, read Martin Cleeve's book "Screwcutting in the Lathe". He includes the theory and practice of making your own dog clutch for screwcutting, as well as going waaaay into the esoterica of the art.
12 March 2019 at 14:36 #399919
Paul Relf-Davies
Participant
@paulrelf-davies37806
Thanks Hopper.. I have read all I can fin on Graham Meek's (elegant) design…. but I think it would need a complete re-design to have ant hope of fitting to my lathe….

It was actually Martin Cleeve's March '77 article in ME that got me thinking on this in the first place…

What I'd really like to see are more detailed pictures of the dog clutch on the ML7 that were in that article. They seem to show the clutch operating mechanism on the outside of the gear pair, rather than between the gears, which would seem to give more 'wiggle' to, but I imagine could limit the change gear combination that you could use…?

P.
12 March 2019 at 20:00 #399992
Paul Relf-Davies
Participant
@paulrelf-davies37806
So..I've gone back and re-measured things.

The shaft the tumbler output 'twin' runs on (and the spindle holding the 2 gears) is an 1 1/8" long. I think there is enough clearance beyond that (before there is a risk of interfering with any change gear combo) to increase the spacing between the 2 gears from 3/8" to 1/2". I think that would give enough space to fit a clutch of some sort.

It certainly should prove an interesting little project.

P.
12 March 2019 at 20:06 #399996
JasonB
Moderator
@jasonb
The gear train starts with the gear on the spindle, as soon as one of the tumbler gears goes out of mesh with that you have about a 1 in 20 chance of getting it to engage in the right place
12 March 2019 at 21:44 #400019
Paul Relf-Davies
Participant
@paulrelf-davies37806
Indeed. The plan will be for the gears to be in mesh all the time, but as has been mentioned, engaged/disengaged using a single-tooth dog clutch.

The extra 1/8" I found still allows for the gears to be meshed all the time.

I'll post some pictures when I'm done.

P.
12 March 2019 at 22:10 #400023
Hopper
Participant
@hopper
Please do. What sort of lathe is it?
Author

Posts