Making a dividing plate

[MRA]

hi all

Flushed with success at getting my home-brew milling machine to work, I fancy having a go at gear cutting. I've read up on an approximate method using a home-made hob by 'Helichron', referenced from this forum - very interesting and nicely explained. I have a long-term loan of a rotary table whose handle does 3 turns to the degree. So I next need to make something to attach one of my lathe chucks to the table, to come up with a big angle plate to sit it upright, and to make a dividing plate.

My question, which is a bit of a daft one, is to folks who've done a bit of this before - what kind of divisions are useful on the plate? I know that's just asking for the response 'how many teeth are you hoping for?!' - but it struck me that for a given worm ratio, certain numbers (like the eminently divisible 60, perhaps) might be more useful than others (43, say...). What do you think - are there rules of thumb?

I do have a goal in mind, which is a pair of spur gears for a knackered set of small bending rolls at work. But this is just the excuse, really.

 

[bazmak]

24 is the smallest useful one does 1,2,3,4,6,8,12 and 24
60 does 5,10,15,20 and 30
63 does 7,9 and 21
The more unusual nos do less and are usually only made for specific sizes
This is of course for direct deviding,with a worm and wheel and indirect deviding the nos usually range from 55 to 65 or thereabouts.Others will with much more knowledge than me will chip in

 

[MRA]

Thanks very much, that's a useful start for me. But I've already dropped a clanger:

...I have a long-term loan of a rotary table whose handle does 3 turns to the degree...

...by which I meant, of course (!), 3 degrees to the turn!

 

[MRA]

Well, as the late Ian Drury said, "there ain't half been some clever b*****ds".

I started to work this all out on a spreadsheet, which involved me looking at long decimals and trying to guess how many 17ths (for example) such decimals might correspond to. I lost heart somewhere in the 23rds.

But - here is a web page which will let you put in your worm ratio and tell you what dividing plate set of holes will give you what set of gear teeth. Then it will draw the plate for you, so you can print it out and use it as a template.

http://www.bilar.co.uk/engineering/index.html

This might be a bit iffy for a direct division, but my table is geared at 120:1 so any small error in the plate will be inside the slack and backlash of the operation in general. I hope!

 

[RM-MN]

Are you planning to make plastic gears or metal? Plastic gears will be easier to make and probably less critical of the perfection than metal.

 

[abby]

Plastic gears will probably be useless for operating bending mill rolls as the forces involved can be considerable.

 

[Hopper]

Tables of holes in division plates for worms with ratios of 60:1 and 40:1 are readily available in Machinery's Handbook, Harold Hall's book "Dividing" and GH Thomas's book "Workshop Techniques" and I am sure there are plenty of charts on the net. No need to reinvent the wheel.

 

[RM-MN]

Plastic gears will probably be useless for operating bending mill rolls as the forces involved can be considerable.

True, but plastic may be a great way to learn what not to do. So I screw up a Delrin gear and have to make another? It's a pretty cheap lesson.

 

[MRA]

Yes, I have some bits and bobs of nylon, polycarb, tufnol etc to practice with. The first pair of dividing plates (which obtain the large majority of sets of teeth to 100, using holes spaced between 12ths and 29ths) will start as paper glued behind 1/4" clear polycarb and drilled through the templates - I'll need a nice soft sprung finger to locate in them so as not to chew them up.

I think I might try to make the hob based on the DP/mod for Boxford change wheels. I could do with a couple more to extend the metric threads I can cut, and if I really get motivated I could cast an alloy or bronze blank for the 127/100 compound gear which converts the English lead screw, and sell my 'original' which cost me rather a lot!

cheers
Mark

 

[MachineTom]

To make plates you can begin with a 50 hole outer ring, which is two turns on the handle, drill a hole, now with a 50 ring done, do another of ring where you can make some prime multiples like a 39. to get a prime number of holes, easiest is to ask someone who has a DRO bolt circle program to do it for you. Or someone with a CNC would be even better.

You can use a mill with dials to make the prime numbers as well, it takes a while but it will be far better than a fax stuck to a disk and marked out in blue.

 

[gus]

To make plates you can begin with a 50 hole outer ring, which is two turns on the handle, drill a hole, now with a 50 ring done, do another of ring where you can make some prime multiples like a 39. to get a prime number of holes, easiest is to ask someone who has a DRO bolt circle program to do it for you. Or someone with a CNC would be even better.

You can use a mill with dials to make the prime numbers as well, it takes a while but it will be far better than a fax stuck to a disk and marked out in blue.

Tom.

Well said. While building the Howell V-2, I had gears with 34,21,14,13 Tooth to cut. Would be an awful lot of indexing plates to cover these holes. Choose to go direct indexing. Giving the choice of having a mate to CNC drill this combination Index plate would be best choice. I am a loner here in Singapore. So had no choice but to get a fren to do computer dividing and fax over. Using fax to prick punch all all holes was tedious. There was no room for missing the cross hair. Being a born loser I Spot Drill by hand till I get a good indentation. Went well but the 14 holes had one bad hole. Had to make another 14 plate.

 

[MRA]

...but it will be far better than a fax stuck to a disk and marked out in blue...

I think I see what you mean if one is going for direct indexing - but the rotary table I am working with is geared at 120:1! So even a 5 deg error on the plate, which would be really, really sloppy work, results in a table error of <.05 deg.

The downside is that I will be counting high numbers of turns, and I'm likely to drop one now and again. The table has protractor marks around the rim which will help as a check, and maybe I should rig up a cam and 'clicker' counter - I have some somewhere I was saving for an electrical coil-winding setup.