Dividing head onto an ML4

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.


Well-Known Member
Nov 30, 2016
Reaction score
Hi all,

Been a little while, hope your all keeping well. After finally getting around to making a bolt for the stud on my change gears which snapped. (Photo below for anyone interested in me bolt making).

I looked over this forgotten part of the lathe, and thoughts started stirring in my head about a dividing head. (I plan to make some gears in the future). After some research H
I came across Harold Halls website that showed a simple dividing head that just used a gear mounted on the spindle, much as I have now, and a key that locates into the teeth of the gear and also a seperate key that locates into the top of the teeth, which using a 60tooth gear provides quite a few combinations of divisions.

I have some spare bolt holes and places that I could attach a dividing head over the headstock, so firstly what pitch worm gear would I need to engage the myford gears, i believe they are 1MOD but that be a figment of my imagination, once I have this info I can look to setup the change gears to cut the thread for the worm wheel, and Jerry rig some device to attach the lot to the headstock of the lathe. Ultimately I'd like to make one that can be universal so it can mount on the bed and have a spindle and chuck for dividing, and drilling holes ect.

Thanks in advance for your input



Ok with that information and some maths, i think I can get a worm gear, if someone could explain Harold Halls gear ratios to me please.

From his website he explains how to work out the gear ratios for any given thread using an 8TPI leadscrew.
As thus (anyone wants to correct inaccurate working out please feel free)
Pi x D = C
So 3.1416x1"=3.1416

So 3.1416/20=0.1573

No of teeth × CP = repeat
So 40 x 0.1573 = 6.292

The closest ratio I can find in his table is for a repeat of 6.2995, which gives an error of 0.0003" which by the time it has passed through the 40:1 gear ratio is negligible. This worm wheel should work with my other gears also.
The gear train is described as below.

He has them set out as this (R1) 35 (R2) 65 (R3) 70 and (N1) 38 (N2) 55 (N3) 60. I am guessing these are in pairs but can someone explain what R and N mean as I'm somewhat confused, and cant find reference to it on his site.
The hardest part now is accurately grinding a 55° tool for cutting the teeth.

I see that you have been deafened by the response:rolleyes:

All my attempts at such things over the years simply bears no relationship to what you are trying to fathom out.
FWIW, I went along the David Lammas way of things originally and recall that Martin Cleeve 'thinned' the screw thread to accommodate the non skewed 60 TP gear. It's all in Engineering in Miniature and I suppose that Blackgates still has bits for the job. Again, Sparey wrote it up in the Amateurs Lathe book ( which you should have made an effort to obtain)

If you wanted to go the whole hog, bells and whistles and shouts of 'Eureka', Bill Bennett's compilation of George Thomas's books ie Workshop Techniques has it ALL and with the further use of your wallet spanner, how to cut not just one screw thread but TWO. Clears throat- which is wot I have -so there!

Again, one has the toss up of buying all 9 gears with intervals of 5 teeth or has a set or sets of the range of 20 DP cutters to do the everything including the rack for a tool and cutter grinder-- to make 59 degrees( if you must)

One cheerful thought is that collecting seemingly endless gears for the Myford is almost as profitable as acquiring 'bitcoins'

Fun- innit?

Last edited:
Dunno about HH's calculations. But as Norman Goldstar said, GH Thomas's book Workshop Techniques has all the exact info, right down to the change gear train, to machine a worm to mesh with Myford lathe gears. He includes full drawings for the Radford-style indexing attachment that engages with the bull gear rather than gears on the end of the spindle. He also includes details on how to grind up the 29 degree tool bit needed to screwcut the worm.

That said, a lot can be done by using change gears to index the spindle, using compound gearing and a plunger that engages with the gear teeth. I think it is outlined in Sparey's book. I do this for things like engraving graduations on handwheel dials etc.

But if its making gears that you want to do, you might be better off to use the lathe spindle to run your cutter and hold and index the gear blank in a standalone dividing head bolted to the cross slide via a vertical slide. Either HH's dividing head or GH Thomas's would do the job. Then get yourself a copy of Ivan Law's "Gears and Gear Cutting" for some light reading.

Otherwise, you end up using the lathe spindle as an improvised dividing head to index the blank and then have to improvise some kind of cutter spindle to mount on the cross slide, so you end up wth double improvisation, not so ideal.
Not only I would 'like' Hopper's remarks but thoroughly endorse his advice.

It is probably salutary not to attempt to 'design the wheel' yet again. Over the years, the late Martin Cleeve wrote extensively on his experiences with the ML4 and then the half Myford ML7 which he eventually pressed into being his 'bread winner' when he was made redundant.
Somewhere in the ' Almost Seven Ages of Man' where I am presently, Cleeve modified a lathe tool to cut the double flanks of his gears in lieu of either purchase of gears or gear cutters.

Might I remind you- for the umpteenth time, that the wheel is invented and that should give you time to see how others have tackled the problems for which you seek answers
I have just seen what Harold hall means by R and N, the drive"n" gear and the drive"r" gear.
So above combination of gears would be 35/38, (spindle to first stud), 65/55 (first stud to second stud), and 70/60 second stud to leadscrew gear).
Penny drop moment just happened!!!
However I'm not sure I have a 70tooth without checking so the next combo I could go for is 20/21, 30/25, 50/45, which I am sure I have all of those gears. This would enable me to turn a worm that should mesh nicely with the existing gears.
However thoughts on a seperate dividing head are quite valid so I'll work on that one.
This would enable me to turn a worm that should mesh nicely with the existing gears.

Again, I strongly doubt that a conventional
skew gear cut on a ML4 will mesh properly with ordinary straight Myford gears.

I recall Alan Timmins and Martin Cleese and the kit- and I made one up- was as I said. Blackgates Engineering actually altered the 60 tooth wheel to mesh.

Engineering in Miniature- if my brain is still ticking over
I can certainly see your point Norm, but I've got to cut my teeth on the lathe somehow lol.
I take your words on good advice and if I can cut a worm, then i can cut a gear by creating a second worm and making it into a simple gear hob. So if it won't mesh at 90° then I'll file the backs of the worm and try methods two.
If you cut ONE worm, you can use a one hole dividing plate but if you make two--- then -----------------well time to whisper differential indexing!

Hi just wondering if someone could check this working out for me as I'm crap with gear ratios, (if there's any corrections needed please can you show working out as to where I went wrong).

So i want to cut a 1" worm gear, (which essentially is a screw thread) with a TPI of 6.292, with the change gears available to me which are 20,20,25,30,40,50,55,60,65,65,95.

Using Harold Halls charts which I assume would be identical to Mr Cleeves charts, the closest above this figure I can get with available gears is 6.303 TPI. There's three different combinations for this TPI.

25/20 55/50 60/65
30/50 55/25 40/65
30/55 20/65

The above is the different gear ratios so spindle gear 30, driving a 55T gear, coupled to a 20T gear driving a 65T gear on the leadscrew for example.

Also is the error of 0.011 TPI acceptable tolerance for cutting a worm gear, or what would my diameter of worm gear need to be to rule out this error? I imagine we are talking only a small change to increase the diameter of the worm gear by a few thou? Ie for 0.01" increase in dia equals a change of 0.001TPI?

It's late and I've been looking at this most of the night without a eureka moment! Including looking at the faint whisper by Norm above. ;)
Another interesting whisper are the words of Professor Dennis H Chaddock when he cut the spiral thread of 1.000 " on the 1.250" vertical pillar of his Quorn.

Too much information to be written here== and it's a long, LONG time since I went that way. You have the reference and that should suffice for you to avoid perhaps costly mistakes. Recalling Chaddoch, there is an interesting snippet in Cleeve's Screwcutting in the Lathe of Chaddock's deliberations on gearing. Again, too much for me to have to repeat for the umpteenth time.

And then there are the many articles on something called the 'Big Dig in'. For my part- and my past experiences--- modesty forbids.

I think I can surmise what the "big dig in" pertains to, and my modesty don't really matter for now. As I had one of those in the last week, my own fault, not concentrating enough, cutting Ali the cross slide was set to around 20thou, on a piece that wasn't fully dialled in, consequently the lathe cut 40thou dia, plus the amount of wobble that existed. It very nearly tore the Ali out the 3jaw, (it was a 2" lump 1.25" from the chuck jaws), jammed the machine up a good 'un. Thankfully I have my belt set so that is can turn blue chips easily on steel, however if a dig in like this occurs, the belt just rotates on the pulley. Worked well on this occasion and nothing damaged but my pride and the chunk of Ali.

I take it my calculations don't seem correct then? I will look into the references you've given.
How would changing the diameter of the worm thread make any difference to the pitch? Whatever you set your change gear train to will be the pitch, regardless of whether you cut your thread 1" diameter or 6" diameter,surely? The helix angle will change with diameter but that makes no difference to pitch.

GHT quotes a geartrain to cut a .1575" linear pitch worm thread to mesh with a a 20DP Myford gear.
He uses 60--55/65--idler--50 to obtain a ratio with an error of 1 in 84,000. Sounds better than the 1 in 100 or so you quoted (.011).

GHT's train is what I used when I made my versatile dividing head using a Myford 60T changewheel as the improvised wormwheel gear. Just followed Cleeve's lead and thinned down the thread to allow for the worm shaft axis to sit a 90 degrees to the wormwheel gear shaft axis. It seems to work perfectly well. Used it to index and drill the 814 holes in the three indexing plates and they all came out dead-on.

(Of course, the VDH included the second, smaller worm and wheel made on the lathe to facilitate the needed differential indexing for drilling the 814 holes. You can read all about that in GHT's two books.)
It can get worse. I was presented with a Super7 top slide that had been torn out of the boring table.
As you will guess- this was a damned sight tougher than your ML4. Cleeve, if you finally do get to read his Screwcutting in the Lathe, demonstrates his rather different boring table on his later ML7.
The keener observer will perceive that it has no hole for a top slide as in all the ML10's, ML7's and Super7's. In fact Myford made up a special for him-- out of STEEL following his 'dig in'
I bought a Super 7B ( from Darlington) with a warped boring table.

You, might I suggest, start by making the fixed steady and spindle support a la Cleeve and finally correct the worn spindle bearings on your lathe-- before trying to to do much more.

Of course, you CAN ignore the advice, waffle on and start a new thread of how to repair self inflicted injuries.

'Nuff for today'

Norm I did look into replacement bearings, and I believe the only cost effective solution for my poor bearings is to melt off the old and cast in the new so to speak, already i am conspiring with a friend to make myself a gas bottle furnace, powered by Coke (not the drink) and a 12V air conditioning unit from a car. This and a few other bits of tooling should have me on my way to recasting the white metal bearing surface, if I can find the material in small enough quantities to enable it to be cost effective. Which currently I can't.
Hopper I said it had been a long night. Thanks for the correction.
There are several ways to skin a cat and so with the ML4.
The first, and this was how Cleeve, was to use sintered phosphor bronze bearings.
Rambling on a bit, he also used the same to do the suggested( by me) steadies.
In your inimitable and somewhat cack handed way, you are going for miles around in almost ever decreasing circles. If you really want 'bearing material' melt down existing scrap shell bearings out of vehicles and make ingots. If, and I am guessing, that you already have tinned bearings you could remittal the existing ones and line bore them out.

My guess, and I did have 'rather' high marks in cost accountancy, the overall cost of the latter will outweigh the simple bunging out old bearings and bunging in the properly made ones.

One reasonably close source of bearings is this place in Station Road, Bill Quay, -- ?????? Gateshead Bearings. I came out with three needle roller sets for the Super7 original clutch for all of 15 quid the 3.

I've lost the b*****s, and no one is perfect!

Just a thought though. If you line bore a bearing using a shagged spindle and tailstock poppet- and the bed is straight, you get a perfectly parallel bore. I bored TWO dead parallel bores in three castings of a dead 1.000" and a 1.003" for the bed rods on a Quorn- and then cut one of the bores with a micrometer end-- on a old Pools Major ex WW2.

Kept my fingers out of my perineum for a long while

Sounds like gear cutting is all a bit academic at the moment if bearings are that shot. I like the sintered bronze bushing idea. If you want to melt white metal, a steel or cast iron pan on a gas stove will do the job. Much easier than digging your own coal out of the hillside and making a coke oven etc.
Sounds like gear cutting is all a bit academic at the moment if bearings are that shot. I like the sintered bronze bushing idea. If you want to melt white metal, a steel or cast iron pan on a gas stove will do the job. Much easier than digging your own coal out of the hillside and making a coke oven etc.

Bearings aside for a moment, i would like to learn to cast Ali, and even brass, though I'm told brass is a pain because it burns off the zinc, so more needs adding. So a little melt pot is on the cards to start some simple Ali castings.

Now bearings, i dont think the bearings are too bad just user error, they do need replacing granted. However they are adjusted to zero play in vertical movement, and nipped up to less than 1thou horizontal movement.

The tool has quite a bit of rake on the top, (makes nice tight swirly swarf). The large dia wasn't dialed in very well, so had anything upto 10-12thou wobble at its furthest most end. When the large (for my myford) 40 dia cut was applied and the tool hit the extra 10thou of wobble it dug in with the top rake and the pulled the soft Ali out the jaws a little (1/2" at most) switched machine off and released the tool, rechucked the Ali and set about cutting just a 20thou dia cut instead.
Unless the ML4 should be capable of taking something like a 40-60thou cut on dia? It was also only chucked in the 3jaw so less holding force as opposed to using the 4jaw.

Latest posts