Cutting clock wheels using a myford dividing head

[John Manning 5Participant @johnmanning5]

problem using dividing head to cut common tooth number

[John Manning 5Participant @johnmanning5]

Hi all you clock makers. I have restarted my project to build a clock and find that the great wheel is 144tooth. No problem says I, set up dividing head and away we go!. Not quite as easy. The Myford dividing head is a 60 tooth worm wheel and 144 teeth calls for 5/12 rotation of the index handle per tooth. There are no index plates with hole counts that are divisible by 12, either in my workshop or available from Myford. Have I missed something here because several clock designs use 144teeth for the great wheel.

There are several alternative solutions that I have,

A. make a dividing plate with suitable hole counts.

B. use the simple dividing head in compound mode.

C.Use the Myford head with a 100tooth plate and a vernier to achieve 1000 divisions per rev of the handle. (this method is a bit irksome!!).

Some books refer to "Equipment mentioned many times before" which is very helpful!

Any ideas or solutions greatly appreciated.

JohnM

[Michael GilliganParticipant @michaelgilligan61133]

John,

A 12 hole plate is very easy to make [or you could even use a scrap clock dial]

MichaelG.

[AjohnwParticipant @ajohnw51620]

I don't think that B&S cater for 12. Madestone engineering have a set of 4 plates which include 24 holes. That's one of the reasons i bought them.

They are smaller than myford plates and I'm surprised that they didn't include 12.

John

–

[John Stevenson 1Participant @johnstevenson1]

Some years ago when ower Gert was alive and selling on Ebay and Metoolsonline, we used to make loads of division plates.

Generic ones that mimicked the B&s ones but had extra sets of holes in like 63, 50 and 127.

Clones of the Myford ones but again with extra holes, 50 and 127 in this case.

 

But after talks with Alan Timmins who gave us a list of all the different division needed on clocks we worked out in Excel what hole numbers would be needed to produce these divisions using 40:1 or 60: dividing heads.

These were called our watchmaker plates, 4 in number and very accurately made and finished.

144 could be got on these plates at 60:1 ratio by using 10/24 and at 40:1 ratio by using 5/18

 

Since Gert died, Ebay and the Metoolonline site have been closed but hopefully early next year myself and the Delectable Debs hope to have the Metoolsonline site back up and running again to showcase what we made before and also some of Debs work as from about 14 she has been closely associated with the Scale 7 Loco Society making specialist products on her own lathes, mills and CNC.

 

To this end we are slowly stocking up on things like dividing plates, gear gauges, gears, drawbars plus a load more bits so hopefully we won't get down to low stock levels.

 

Typo.

Edited By John Stevenson on 06/10/2016 21:50:11

[John HaineParticipant @johnhaine32865]

John (i.e. original poster), I have a set of 4 Myford division plates now surplus as I've motorised my DH – I'll check tomorrow morning to see if I have one with 12xN teeth, you could borrow or buy.

[John Stevenson 1Participant @johnstevenson1]

John H,

Save you the bother. Nothing on the 4 Myford plates have 12 x N on them.

 

And I have made enough of them

 

Edited By John Stevenson on 06/10/2016 22:40:04

[AjohnwParticipant @ajohnw51620]

Don't know how you are making them John but the ones I mentioned may well have been made on a CNC press.

That sets the min size of the holes with a certain metal thickness so in some respects the plates could have been usefully bigger. Ok for me though as my dividing head is pretty small.

If the OP does a "manual" one I would suggest making a plate rather than the part. The gain on larger hole counts is ever increasingly higher division numbers. The Madestone ones looked to be aimed that way. The B&S ones take a different route as they are aimed at obtaining "sensible" ratio's for gearing. The Cincinati ones are rare and do cater for 12.

My either Dore or Westbury plates includes 60 holes. They went their own way on hole counts. Not sure if anyone ever sold these ready made but one of the casting suppliers sold the castings for them. Seems to be an odd choice to me. 40,45,50,55,60,65 and 70 but a lot less holes than the usual set of plates. There is only one. Small too.

John

–

Edited By Ajohnw on 07/10/2016 00:27:57

[John HaineParticipant @johnhaine32865]

I bothered anyway but should have known JS was right! What an omission. However as 5 turns on any plate advances the mandrel 1/12 of a rev, making a 12 tooth plate is trivial, then you're away….

[John Manning 5Participant @johnmanning5]

Well thanks folks, loads of info in there. Route A, make a 12 hole division plate is best. I am surprised that the new Myford don't include a suitable additional plate in their range. At least I don't appear to have lost anything "round the edges" so to speak. Another point is that the tables with the Myford head stop at 100 divisions and don't include some of the prime numbers which makes one wonder whether the unit is going to be used out side it's design limits?

Next up, some work on the raising block and then next week-end (Midlands model engineering show) obtain some 6" dia plate.

Again, Many thanks for your comments.

JohnM

[John HaineParticipant @johnhaine32865]

You don't need 6 inch diameter if you're only making one circle. I think I'd use a 2 inch circle of 1/8 ali, it won't have to last for ever.

[John HaineParticipant @johnhaine32865]

Or even a square…

[Andy AshParticipant @andyash24902]

I guess you will have to do some sums to find out if you can tolerate the differential error, but…..

If you use the Number 1 Myford plate, I think you can get to 144 by using 32 holes on the 77 circle.

If you do this, then every 12 divisions you will need to use 33 holes on the 77 circle.

I suspect no-one will notice the difference, and you will save having to make a 12 hole plate.

I suggest the usual caveats about checking my sums, which can sometimes be amiss.

PS

I can compute the maximum angular error at every 12th division, but it is more meaningful if it is understood as a linear dimension at the periphery of a gear wheel. To calculate the maximum accrued error in thousandths of an inch at the periphery I need to know the diameter of the wheel.

Irrespective it would be approximately 1/32 of a tooth. It accrues as 1/384 of a tooth at each tooth and would be corrected every 12 teeth by 1/32 of a tooth. At the end of a complete revolution there are exactly 144 teeth.

So for a 4" wheel that seems like 2 and 3/4 thousandths worst case every 12 teeth.

Edited By Andy Ash on 07/10/2016 21:10:29

[John HaineParticipant @johnhaine32865]

A creative solution Andy. This is a bit like finding metric thread approximations using imperial gears, maybe a spreadsheet is needed to show all the combinations like this for minimum error.

[BazyleParticipant @bazyle]

If you are really lazy you can make use of the fact that the plate has 3 fixing holes and can be rotated to multiply each existing plate by 3 (except multiples of 3). So you only need a 4 hole plate and probably have something that is suitable.

[Phil PParticipant @philp]

Unless I have missed something obvious, surely you will be able use what you already have to get 12 divisions and add a 12 hole circle to one of your other existing plates, then use that one to make your clock wheel.

Phil

[AjohnwParticipant @ajohnw51620]

Personally if I made a plate to cater for 12 I would make it nice and simple – one turn on any other plate = 60 holes in case the increased ever finer divisions proved useful. If I didn't fancy drilling holes I would still go for 24 rather than 12.

If the head has a quick index it might be worth checking what that has on it.

Something could be lashed up to use the worm itself for an index.

John

–

[John Manning 5Participant @johnmanning5]

Hi, Still lots of ideas, many thanks again. I am making a plate out of a chunck of Perspex that has been in the "may be useful" heap for about 30 years. First job has been to make a mandrill to hold it in order to drill the 36 hole circle and a 40 hole circle. May as well drill two circles that complement each other. There is no spare space on the other plates and being Myford originals I am unsure of the material as it could be hardened or gauge plate. I shall have to check the other wheels before I commit to drilling as it would be a bit silly to find that I can't make another wheel when I am specially making a plate. At the moment the grass is too wet to cut so it's unfortunately back to the man cave!!

Many thanks

John M

[John Manning 5Participant @johnmanning5]

Hi, evening everyone.

Plate rough cut and arbour turned. I inherited the lathe and accessories from my late father who made very useful notes about the machine. I have just picked up a note that is labelled "plates". He writes

"three screw holes of 5/32" 4BA clearance on 15/16" PCD. NOT EQUIDISTANT.(his capitals)

3 fixing holes clockwise (looking at front) from "zero" position, 119.5 deg. 118.0 deg 122.5 deg"

Apart from the fact that I have either; A, failed to find, or B, given away any angle measurement device of this kind of accuracy. This seems an amazing bit of info to investigate. He was making the 100 hole plate and vernier device probably at this time. I wonder if John Stevenson can shed a light on this. 2 degrees error on fixing holes does not seem much and may reflect the manufacturing method more than a design feature.

I shall just spot through the existing plate to the arbour.

best wishes

John M

[Andy AshParticipant @andyash24902]

Posted by John Haine on 08/10/2016 09:56:10:

A creative solution Andy. This is a bit like finding metric thread approximations using imperial gears, maybe a spreadsheet is needed to show all the combinations like this for minimum error.

There was a thread I posted on a while back where I did exactly such a spreadsheet.

**LINK**

The spreadsheet shows the exact ones in green. The "nearly" ones are in yellow. The nearly ones are prefixed with a minus sign (which can be ignored) purely so that they can be highlighted yellow. You can change the acceptable error (along with other things) to make more or less approximate yellow matches show up.

I notice that the OP is making a new plate. That is really the right way to go, but there are other options.

[John HaineParticipant @johnhaine32865]

See

**LINK**

[John HaineParticipant @johnhaine32865]

Andy, excellent!

[Author]

Posts