Myford Super 7 screw cutting gears (metric)

[Allan BParticipant @allanb]

I know there is a lot of maths going on here in the background, would one of you very nice knowledgeable gentlemen mind sharing some of your workings out so I can try and get my head round it?

[John Stevenson 1Participant @johnstevenson1]

No 

Edited By John Stevenson on 20/03/2017 20:45:49

[John Stevenson 1Participant @johnstevenson1]

Allan,

Just joking.

In my case the maths were done years ago and TBH I'd probably struggle to sort out what I did nowadays.

Because these questions always came up I wrote a spreadsheet and upon this I superimposed a drawing of the Myford gearbox legend.

This is the standard plate with the 24 driver.

Top row of number per row is the standard imperial tpi from 8 to 56. The lower row of numbers which make no sense are the pitch in mm.

However if we swap the 24 for any other number, things change dramatically.

Here we have changed the 24 to one of the magic number, in this case 34 and the top boxes are now gibberish but the lower boxes have the number we need for metric threads.

So the maths you need are embedded and hidden under that image of the legend plate. This spreadsheet was written about 1998 or so hence my fuzzy memory of just what it does.

[Allan BParticipant @allanb]

That makes sense John, I can work out the gear ratios, the bit I was interested in was converting the gear ratio, even a compound gear ratio into the TPI of the thread, I can do it with motorcycle gearboxes, but that is converting horse power and talk, but not into actual movements in a linear motion

[Simon Williams 3Participant @simonwilliams3]

Evening all!

Well, bless my soul, that's a mighty pretty summary of where we've got to! Thank you John.

Do you have a spread sheet printout with the tumbler gear set to 33T?

Allan – request for explanation noted, really John's spreadsheet says it all, provided we don't worry too much about what's hiding inside the gearbox. It's only really the relation between the spindle and the carriage travel we're concerned with, and that's what the label on to top of the Q/C gearbox details. What John has done (bodger that he is) is to alter the input gear ratio driving the gearbox, which alters the amount that the carriage moves each revolution of the spindle.

It happens that the alteration chosen magically makes the carriage movement fit most of the useful numbers of the metric series of threads, hence being now able to cut metric threads – or at least a damn fine approximation thereto – with an 8 TPI leadscrew.

It's just a confusion that the imperial threads are traditionally measured in Threads Per Inch, and metric threads are expressed in the pitch between successive crests, measured in millimetres. They are actually the same thing expressed in two ways.

I've gone and confused the otherwise confusing logic of this by finding that the gearbox fitted to my lathe happens to run at half the speed of the one John's calculations were designed for, so all this stuff about 17 tooth gears is me trying to bodge my way round this. I suggest you ignore that for the time being.

The arithmetic is actually fairly simple – or at least I think it is though I'm waiting on Brian to tell me I've misunderstood something. The usual gearbox is fed from a 24 tooth gear. How the motion gets to the leadscrew doesn't matter, but the lever positions tell us the overall result, expressed in TPI. John's spreadsheet replicates this, AND adds the information of the thread pitch in mm. Remember it's the same information, just expressed in a different way.

What we've then done – not my idea, I take no credit for this, is to change the gear driving the gearbox, so that for the same speed of the headstock we have changed the speed of the leadscrew, hence the carriage travels at a different speed, and the TPI numbers (and thus pitches) alter. There's nothing magic about this except for the canny choice of the new driving gear, the gearbox itself hasn't changed and so the output variable – the speed at which the carriage travels along the bed – varies in strict ratio of the new driving gear to the old driving gear.

The original Myford set-up drove the gearbox with a 24T gear. We've decided that's not good enough, and now decided to drive it with a 34 tooth gear. So all the thread pitches (gearbox outputs) increase by 34/24, which is what John's spread sheets show, and the TPI's of course reduce by the same proportion. The clever bit is that someone realised that by choosing 34 as the new number many of the new pitches happen to fall neatly into the numbers you would want to cut metric threads. Eh voila! Metric threads on an imperial lathe.

This cunning plan was developed further by realising than some other useful (metric) pitches fell into place if said input gear was now altered to be 33 teeth.

So now for an apology, because I've come into this late, and presumed to write this up, but it's not my idea and I've got a sneaky feeling that the clever chap who did think of it is somewhere in the mix of this thread. Gent's I apologise unreservedly for my arrogance in taking this upon myself, I only hope I've got it right!

Hope this helps, best rgds to all

Simon

[John Stevenson 1Participant @johnstevenson1]

33T gear fitted.

[Roderick JenkinsParticipant @roderickjenkins93242]

Posted by Simon Williams 3 on 20/03/2017 22:28:38:

So now for an apology, because I've come into this late, and presumed to write this up, but it's not my idea and I've got a sneaky feeling that the clever chap who did think of it is somewhere in the mix of this thread. Gent's I apologise unreservedly for my arrogance in taking this upon myself, I only hope I've got it right!

The earliest reference to this method I can find is from The Rev. David Hoskin in Vol 171, issue 3955 of Model Engineer. JS has done much to publicise the technique and make the gears available and I, for one, am very grateful to him.

Rod

[John Stevenson 1Participant @johnstevenson1]

Post to explain how corrected gears can be cut using standard cutters to a reasonable accuracy**

Could be long and may need to be in two parts because of the length of it, so best go get a coffee and throw the cat out.

Bit of background first and some of this may be simple but as I don't know who's reading this I need to explain the whole.

Making corrected gears is something that is done in industry all the while. Strip and measure any modern automobile gearbox and I'll bet none of the gears are standard, all modified for strength, quietness etc.

In production work it's easy to modify corrected gears as any manufacturing method that uses the GENERATING method like hobbing, shaping etc will form the corrected involute automatically.

Unfortunately the home shop method uses FORMING which is limited to the shape of the forming cutter.

As every gear as it increases in tooth number has a different shape then we need a different cutter for each shape, rather a tall order to have say 100 cutters and then you need a hundred in every pitch or DP ] you want to use.

So very early on Brown and Sharpe decided to 'band' these as the differences are very slight.

The bands were numbered 1 to 8 so only eight cutters were needed per set.

Cutter No 8 cuts from 12T to 13T

Cutter No 7 cuts from 14T to 16T

Cutter No 6 cuts from 17T to 20T

Cutter No 5 cuts from 21T to 25T

Cutter No 4 cuts from 26T to 34T

Cutter No 3 cuts from 35T to 54T

Cutter No 2 cuts from 55T to 134T

Cutter No 1 cuts from 135T to a rack.

As you can see the number of teeth in the lower range is limited by the differences in shape whereas as you get bigger the difference isn't so great as regards shape.

A Number 8 12T to 13T is very curved whereas a number 1 from 135 to a rack is virtually straight sided as a rack should be.

A note here to explain that the band of cutters is most accurate at the lowest number of teeth listed.

Now lets take Simon's 29T special gear.

Standard 30 T on the left, 29T on the right, both standard at 20DP and 14.5 pressure angle. There is virtually no difference in shape as these would have been cut with the same FORMED cutter but these shapes are GENERATED.

The important bit is the 30T has an OD of 1.600" and the 29 has an OD of 1.550" but Simon needs it to have an OD the same as the 30T so it meshes correctly.

So we correct the gear by moving the relationship of the Addendum and Dedendum [ top bit of the gear above the pitch line and bottom bit of the gear below the pitch line ] and we get this.

Now both OD's are the same but now look at the shape of the tooth. Because 29 teeth have been spread out where 30 should be they now have fatter stronger teeth but unfortunately you can't use the standard No 4 cutter or you just finish up with the same shape at the 30 T.

End of this post. I need a coffee and I reckon the cat is still outside.

More later.

[John Stevenson 1Participant @johnstevenson1]

Couple of corrections first

Accuracy** the ** was meant to be a note to say working tolerances in a home shop and ignoring the armchair machinists and angels on pin heads.

Chinese cutters are often Banded 1 to 8 the same but in reverse order so be careful when reading descriptions.

Now how to do the corrections. As usual, many ways and it depends on equipment to hand, skill levels and the price of fish.

The most accurate** way would be to save the drawing of the new gear as a DXF, blow it up and transpose two circles onto the involute flanks and then the relevant date of diameters and distance between can be taken direct from the drawing and a button cutter made than in turn is used to make a form cutter.

Next way would be to cut one tooth space out as a template and then go though a massive stack of gear cutters to find one that fits the gap best. Only applies IF you have a massive stack of gear cutters.

However most people only have a few, usually pst of the 8 band as they feel 12T and racks are beyond thier pay grade.

So we will take a look at this method. Again it can be calculated but if you find it's something that might come in handy from time to time then it's worth spending 1/2 a night just drawing one tooth in in the 8 bands.

Doesn't matter what DP as they can be scaled.

The red is Simon's corrected gear, the green is one tooth of the next band down, the No 5 and has been sunk into the blank 3 thou deeper than standard so the main flanks of the gear match the corrected gear.

Because the band is one lower there is more undercut but as gears work more on the addendum that the dedendum for what it is need for , it will be perfectly fine.

It is also possible to be this mathematically but TBH the hands on paste and drag is far quicker and gives the same results.

That's it for tonight as both me and hound need our cheese and biscuits. He's a very discerning hound, best stilton for him, no mousetrap.

[Simon Williams 3Participant @simonwilliams3]

John – thank you for an impressive piece of work. I stand in awe of the expertise.

I'm hoping tomorrow to make the 16 1/2 tooth gear I promised you. Watch this space.

Rgds to you and the hound (hope he hasn't been chasing my cat!)

Simon

[Allan BParticipant @allanb]

John, that information is wonderful, gear cutting is my next learning project after I get my head round screws so I might print this lot out if you don’t mind so I can keep it for reference.

Allan

[Brian WoodParticipant @brianwood45127]

Simon,

​I have another heavy day and there is much to read here and catch up with now anyway

I will cover your points later, I can also see the page I dropped in my earlier predictions and will put that right to your satisfaction.

​More later

Regards
​Brian

[Simon Williams 3Participant @simonwilliams3]

Fruit of my day's labour ( the bronze bit centre stage):

Is a 29T gear pretending to be a 30T gear, with an integral 16 gear attached, meshed with the gear train driving the QC g/box on my trusty S7.

 

 

xxx

The proof of the pudding is in a 1.75 mm (approx.!) thread.  Q/C gearbox is set for 20TPI, I've only made a scratch pass just to prove the train.

 

 

With special thanks to John Stevenson for the geometry lesson!

Rgds to all

Simon

Edited By Simon Williams 3 on 21/03/2017 19:56:38

[Allan BParticipant @allanb]

Looking good, my new gears arrived today with a big thank you to John for supplying them (I wish I had the skill set to make my own, or the kit for that matter lol) but I have been like a kid at Christmas who isn’t allowed to play with a new toy all day as not had the time to escape to try them out.

Allan

[John Stevenson 1Participant @johnstevenson1]

Posted by Simon Williams 3 on 21/03/2017 19:50:10:

Simon

.

Weird shaped tumbler gears ??

And not Tufnol gears ?

[Simon Williams 3Participant @simonwilliams3]

Evening John, (JS) – You've got me bang to rights guv.

Though that's not the whole of the story. What I haven't let you see is the hand filed teeth on the 17T pinion of the first transfer gear (just bottom of picture.

I bought the gearbox second hand from you know where-bay, fitted it ( the comments about a spacer behind the casting worry me, but I'll save that for another day) and found that it jammed in one ratio. Should have stripped it and checked first, but was far too excited with my new second hand purchase to be that clinical. Can't remember the details, but suffice it to say that the single tufnol gear that was fitted to the tumbler is still in the drawer with the rest of the set of changewheels from the days when I set it up long hand, and it's got three teeth chewed off it.

At the time I had a little Alpine simple dividing head, so the ally gear you see here was cut with a single point flycutter, probably of a crude and dubious shape, and has seen maybe 20 years service since the scrunch up. I'm fairly sure this was my first ever adventure into making a gear, so begging yer pardon if it's not up to snuff.

Now that I know a bit more about cutting gears, t'would be the work of a few moments to make another. Must get roundtuit.

To the best of my recollection the other tumbler gear is as original – or at least I haven't changed it. This lathe came out of Stroud Technical College in about 1976, so it's probably had a hard life even before I got hold of it.

Thinking about it, if I was going to have only one tufnol gear it would be better if it was the other one – which is in service all the time. Too late to worry about that now!

Thanks as ever

Simon

PS  The chewy chewy of the transfer gears wasn't me.  They were like that when I bought the gearbox.  20/20 hindsight says that should have told me I needed to strip the gearbox and investigate, but I didn't follow the clues.

Edited By Simon Williams 3 on 21/03/2017 22:40:09

[John Stevenson 1Participant @johnstevenson1]

To be technically correct the tumbler gears are 20 DP but 20 degree PA and not 14.5 like the rest.

Only the spindle gear and tumblers are this PA, done I presume to make the tooth form stronger and easier to produce than a corrected gear ?

[Simon Williams 3Participant @simonwilliams3]

Well, that's a low blow, what!

So I guess it's easier to buy the spares than a whole new set of cutters. Bother!

But that means that the transfer gears and my magic combo should all be 20 deg PA as well. Blow that for a game of soldiers. The change gears meshed with the tumblers, what's going on?

Looks like I'd do better to replace the tumbler gears and the one on the spindle. No rest for the wicked.

[John Stevenson 1Participant @johnstevenson1]

Posted by Simon Williams 3 on 21/03/2017 23:28:39:

Well, that's a low blow, what!

So I guess it's easier to buy the spares than a whole new set of cutters. Bother!

But that means that the transfer gears and my magic combo should all be 20 deg PA as well. Blow that for a game of soldiers. The change gears meshed with the tumblers, what's going on?

Looks like I'd do better to replace the tumbler gears and the one on the spindle. No rest for the wicked.

Sorry Simon, my bad in that not thinking I did the 29T gear at 14 1/2 degree PA.

As you have found out they will mesh, a lot of overthought is put into gears which are actually quite crude, hence your alloy one.

Where PA really counts is where speed is concerned and in turn noise.

Some years ago I sent a batch of 20 T 20DP gears in for some racing bikes to be cut and they cut them at 14.5 and not the 20 degrees on the drawing.

Not easy to see and i went ahead and built these engines up and they fitted with no problems given all gears need some clearance.

However on starting these engines up they howled really badly and it took a while to locate just what the problem was. A new set of gears were cut at 20 degrres and problem solved.

Later tonight I'll do the same exercise with 20 degree and a o 4 cutter in 14.5 to see what the difference is. Im betting not a lot.

[Simon Williams 3Participant @simonwilliams3]

John –

Thanks for that, though please don't concern yourself on my behalf. I've got what I wanted, and compared with my alloy fake 30T gear the tooth shape of the one you've calculated for me can't be far out. Indeed, I'd venture that it's as near as I can make it with the resources I have available, particularly since (if you look very closely) there's something a bit odd with the width of the crests of the some of the gear teeth as shown in my photo's. Given that I've checked the size of each blank carefully, that tells me that either I've used the wrong cutter betimes (which I was very careful of) or the cutters perhaps aren't as accurately profiled as one might assume. Look at the difference in crest width between the 16 and 17T gears for a case in point, they are each cut with their respective cutters yet the crest width is quite different. What the hell, it works!

I checked that the cutter numbers were not reversed, as you suggested sometimes they can be. I'm pretty certain that's not the explanation, but given my ambitions are largely agricultural I'm very pleased with where we've got to. And it's been an interesting insight into what can be done with a bit of ingenuity along the way. If the lathe is a bit noisy I won't even notice – my impression is actually that it's quieter!

Having said all that, I (like you?) am intrigued to quantify the extent of the error implied in this mix of 14.5 and 20 deg pressure angles. I bet it's within the realistic manufacturing tolerances I can expect to achieve.

All the best Simon

[Brian WoodParticipant @brianwood45127]

Allan, Simon and John,

​You will perhaps have noticed I have been 'dark' for the last few days. Our son in law, who was skiing in Austria in January despite having lung cancer [ he was a non smoker] took a turn for the worse very recently and added two strokes to the battle and died on Monday morning. He was only 53 and up until then one of the fittest people I knew.

​I will get back on topic and put the maths together for your interest, but right at present we are rather pre-occupied as I am sure you will appreciate.

Regards
​Brian

[Simon Williams 3Participant @simonwilliams3]

Brian – my very sincerest best wishes for you and family. Please don't give this any thought for the time being.

In due course I'm sure we'd all like to hear your take on things here, but it'll keep until you've got the brain space to suit.

With best regards

Simon.

PS Just read your book, courtesy of the eager elves at Amazon. Thank you for this expertise, it's all there isn't it!

[Allan BParticipant @allanb]

Brian, my sincere condolences, and I echo Simon’s statement of please don’t worry yourself about this issue until you are ready, it isn’t going anywhere.

Best Regards

Allan

[Simon Williams 3Participant @simonwilliams3]

Oops, sorry for the double post. That appears to be something to do with having the cat sit on the keyboard (she wants her tea). I wonder what short cut she found! Shades of the Infinite Monkey cage!

Simon

[Simon Williams 3Participant @simonwilliams3]

Good evening again all,

Here's my summary of the best choices I've found for the old style of gearbox on the S7

Myford Lathe Cutting Metric Pitches

(Old Style Q/C Gearbox)

Pitch (mm)	Mandrel Wheel	TPI Setting	Actual Pitch
0.5	30/17FF	8	0.499
0.7	30/17	52	0.692
0.8	12	32	0.794
1.0	30/17	36	1.000
1.25	29/16	28	1.251
1.5	30/17	24	1.499
1.75	29/16	20	1.752
2.0	30/17	18	1.999
2.25	30/17	16	2.249
2.5	29/16	14	2.502
3.0	30/17	12	2.999
3.5	29/16	10	3.503
4.0	30/17	9	3.998

The FF appended to the 0.5 pitch selection indicates to engage the fine feed drive for this ratio, giving an additional 9:1 reduction.

Best rgds

Simon

[Author]

Posts