Metric transposing gears on early Heavy 10

[SLK001]

I get that, you'd have to work out the pitch taking into account the 3.5x (or something similary) factor of the carriage feed vs. halfnuts. I did that and found a few extra metric pitches that can be cut, but Peter and Allan's explanation of why one shouldn't makes complete sense to me.

If you can work out a correct pitch using the carriage feed, then go for it. The clutch isn't a "slip clutch" that will slip if it hits something, so if it is tightened properly, it should be just as rigid as your halfnuts. Just remember that once engaged, you don't disengage until the thread is complete.

Either way, threading to a shoulder could be a problem, as once started, you don't disengage the drive for metric (unless you have the elaborate metric thread dial). You'll have to be quick on the cross slide handle to pull the tool out of the cut.

 

[mattthemuppet]

I've done some metric threading to a shoulder on a different lathe using half nuts and a generous relief groove - turning off the motor and releasing the belt tension at the same time got the tool stopped within the relief groove. Wind out the cross slide and then reverse the motor and start all over again. A bit more tedious than normal threading but just another set of motions to go through.

I have read about metric threading with the half nuts where you open the half nuts, stop the motor, reverse to the mark on the threading dial where you opened the half nuts and then close them again, before reversing back to the start of the thread. Haven't had the call to try it yet, but that's another option.

 

[kitno455]

So the best easy answer I can find for .8 and .81 is a 43 tooth stud gear:

TPI MMPT %ERR
56 0.8127 0.328

Getting any closer is going to require something more drastic, like a compound gear.

 

[Roger Best]

So the best easy answer I can find for .8 and .81 is a 43 tooth stud gear:

TPI MMPT %ERR
56 0.8127 0.328

Getting any closer is going to require something more drastic, like a compound gear.

I think that is still useful Alan, thank you, half a thou' fit will allow some proprietary stuff to fit a couple of turns on securely.
If it is critical then make both bits or make them Imperial (English ) I say.

The full-works kit is for tool room work, making tooling that must be right. I don't have the skill to do that properly - yet.

 

[Roger Best]

I was going to tabulate the data in this thread into something smart for my wall, and obviously post it here.
Has anyone a preference? e,g pdf, A3/A4/US sizes etc? What works resolution size?

Sorry for my tardiness - too much time lost to illness, too much work.

 

[Peter.]

I'd like a sticker-sized table.

Roger, we gotta talk. I'm half way through making some gears based on the info in this thread and you're only a few miles away. Send me your phone number please.

 

[Roger Best]

pm sent

 

[Roger Best]

Hi all
I have tabulated the data, so generously provided thank you, against ISO standard sizes, its pretty interesting.

That 34-tooth gear is very useful, it covers M1, M1.2, M2.5, M3, M4 fine, M5 fine, M6 fine, M7, M8 fine, M10, M12 fine, M14, M16. So if you are making both parts who needs anything else?

Commonly available fixings are in bold, I know you can find all sorts of things with wacky threads but I just stuck to nuts and bolts for that category. I even found a M5x0.9 die which avoids the 0.8 pitch problem. Amazing.

I can make this look pretty but I could only upload a txt file, the pdf was 20 times as big as the forum allows. I may have to get creative.

View attachment South Bend metric gears.txt

 

[Roger Best]

This is how it should look

 

[Roger Best]

View attachment South Bend metric gears.txt

trying again. And failing.

 

[kitno455]

Coarse, not Course

allan

 

[Roger Best]

Coarse, not Course

allan

NNNOOOOOOOOOOO

ta

 

[Roger Best]

I am having fun with maths deciding which gears are necessary.

Pete is having fun cutting them.

How accurate does a pitch have to be to fit a nut? We can drop some gears if we use the standard gear for some of the sizes.

 

[Roger Best]

I have found something useful - a table of "permissible deviation limits for linear measures" from ISO 2768-1 published by a company called PI. It downloaded from google and appeared in my downloads without me noticing, and I lost the link, sorry.

Sizes ranges for the linear dimension have tolerance classes and values are tabulated in +/- mm. The numbers look reasonable if translated to thou' so I will do some calculations on the permutations we have and work out what the tolerance class would be.

Obviously any fit needs to allow for other errors too.