Methodologies for one-off gears in an antique home shop

[Just a Sparky]

This is just a brain teaser that's been stuck in my head off and on for a while now. I figure if I can get it down into writing here and discuss it openly rather than keeping it bottled up as a jumbled collection of half-formed ideas in my head then maybe I'll finally be able to stop thinking about it.

I'm trying to sort out the pros and cons of various different approaches of cutting the occasional one-off gear on old manual machines.

• One can index their lathe spindle somehow and grind an approximated single-point involute tool to be used in the tool post akin to a shaper. Or use a milling attachment and a spindle perpendicular to the work to perform a gear milling operation.
• Or one can use a horizontal mill in conjunction with an indexing head and either a real set of involute cutters or again, an approximated single-point tool.
• One could also use a shaper as described in the lathe solution above... with the need once again to purchase an indexing head.

Specific to my situation are the small work envelopes available on my mill and shaper. A small Burke No. 4 with a single slot table and roughly 7" x 2" of travel + a 7" shaper. Finding suitable dividing heads for such small machines is not easy nor cheap.

This makes my 13" South Bend an attractive option in my mind for this sort of work - with a plentiful work envelope and a QCTP with height-adjustable tool holders for zeroing homemade involute form cutters. Through the use of a milling attachment plus a hydraulic motor and my 3/4HP power pack it wouldn't be terribly difficult to build a small horizontal milling spindle for it. This would open up the option of using true store-bought involute cutters in addition to cheap fabri-cobbled single point cutters. It would also minimize the amount of wear on my machine and the labor involved. E.g. 52 passes for a 52 root gear rather than 52 * N strokes per root. The only conundrum from there is figuring out how to index the spindle without turning the whole project into a Rube-Goldberg nightmare.

What are your guys' thoughts? Are compact, affordable indexing heads available to fit my single-slot mill and/or shaper that I'm not aware of? Should I look into electronic indexing with a stepper motor and some timing pulleys? Is there a better, easier way of doing things that I haven't thought of? Bearing in mind that this is all aimed at cutting a hypothetical once-in-a-while oddball gear that can't be purchased and adapted to suit. What's the best angle to approach this problem?

 

[magneticanomaly]

I do very limited indexing on my 10" SB, using the gears already on the spindle, IIRC 78 teeth on the large back-gear, 38 on the small, and 24T on the spindle gear that drives the gear train. I have a couple of "fingers" that I bolt to the headstock casting, and i can swing them into the teeth of any of these gears to do any sub-multiple of these tooth numbers. Lots of potential inaccuracy because of lack of perfect centering and lack of rigidity of my "finger", nad of course not a lot of choices of index number.

I have tried "shaping" on this lathe, and it works rather poorly. Usualy there is not room to set the tool up so that it drags instead of digging. Lots of wear on the carriage drive.

I have cut a few gears with a standard indexing head on my old Dow horizontal Mill and formed cutter. Slow but works fine as long as you do not miss a move with the indexing head.

 

[Joe Michaels]

A home-made "index" for gear cutting can be made. If you can find a gear with the correct number of teeth, even if a different pitch, this forms the 'dividing plate' for a home-made 'index'. A mandrel is machined to take the gear blank upon which you want to cut teeth. The mandrel passes thru a bearing block that is clamped to the table of your Burke mill. This bearing block is nothing fancy. It is a block of steel or aluminum bored to a very close fit with the journal on the mandrel. I'd put a saw cut in one side of the bearing block and drill/tap for a pinch bolt. This will lock the mandrel at each division. The mandrel is machined to mount the 'index' gear on one end, and the blank on the other. A short stub on each end of the mandrel is turned to the next lower size of thread (example: if your gear bore is 0.750", then I'd turn and thread the stub end of the mandrel at 5/8"). This lets you use commonly available nuts. The index gear and blank are locked securely to the mandrel by way of these nuts as well as shaft keys. In a pinch, if you cannot mill a keyway, you can drill and ream thru the hubs of the blank and the 'index gear' for fitted pins (small diameter drill rod works well for this).

A locking 'dog' is filed from steel or aluminum so that it fits snugly in between two teeth of the 'index gear'. This locking dog is held on a bracket off the bearing block, or on a separate mounting clamped to the table of the mill. The locking dog is loosened and pulled back out of the index gear teeth and the pinch bolt is loosened on the bearing block to index to the next tooth to be milled on the blank. How you secure the locking dog is a matter of what you have laying around and what works on your mill.

This same idea could be mounted on the cross slide or compound of a lathe and a gear cutter run on an arbor in the headstock.

Creating a cutter with the correct profile to cut involute gear teeth- whether my 'shaping' or for use on a fly cutter- is a bit tricky. If you can get hold of some O-1 drill rod, and are handy with files, you can hand file a cutter from the drill rod. This is then hardened and tempered to a light straw color. After that, it is finished by hand stoning with small oil stones of various shaped. If you want to use high speed steel, such as a lathe tool bit blank, it is a trickier proposition in a home shop. Dressing a radius freehand on the corner of a bench grinder wheel or using a small die grinder or "Dremel" tool would be a starting point. Freehand grinding of a tool bit to the rough profile and then endless rounds of oil stoning would be the way to go about it. Milling cutters for cutting 'involute' spur gear teeth are available as new on eBay. These are Chinese made milling cutters, usually having a 1" arbor hole. I've bought a few of them over the past few years for cutting some spur gears. For my purposes these cutters did the job.

 

[Just a Sparky]

Since my OP I've been thinking about using a stepper motor to index an arbor using timing belt pulleys with 6:1 reduction. I'm thinking either XL or MXL timing belt pulleys should offer adequate precision without taking up too much space. With a 200 step motor running 1/8 micro-steps, that should yield a resolution of 1/32 degrees on paper. Torque shouldn't be an issue since a brass dog driven by a wing screw can be added to lock rotation, much like the ones on my South Bend compound. It would be an arrangement that is fairly compact, having an overall height-above-table of 2" with MXL pulleys or 4" with XL pulleys and centers at half of that. A power supply, microcontroller, LCD/7-segment and stepper driver would ostensibly be all the major guts I need. Much less expensive than a true dividing head and much less risk of human error when advancing between teeth.

 

[fciron]

With regards to Joe Michaels' suggestion, you don't even need to find an appropriate gear, in a pinch you could lay out and drill your own index plate and make a locking pin to fit it.

That is a solution that could also be attached to the lathe spindle if you need an excuse to make that live tooling for your lathe.

 

[plastikdreams]

This is just a brain teaser that's been stuck in my head off and on for a while now. I figure if I can get it down into writing here and discuss it openly rather than keeping it bottled up as a jumbled collection of half-formed ideas in my head then maybe I'll finally be able to stop thinking about it.

I'm trying to sort out the pros and cons of various different approaches of cutting the occasional one-off gear on old manual machines.

• One can index their lathe spindle somehow and grind an approximated single-point involute tool to be used in the tool post akin to a shaper. Or use a milling attachment and a spindle perpendicular to the work to perform a gear milling operation.
• Or one can use a horizontal mill in conjunction with an indexing head and either a real set of involute cutters or again, an approximated single-point tool.
• One could also use a shaper as described in the lathe solution above... with the need once again to purchase an indexing head.

Specific to my situation are the small work envelopes available on my mill and shaper. A small Burke No. 4 with a single slot table and roughly 7" x 2" of travel + a 7" shaper. Finding suitable dividing heads for such small machines is not easy nor cheap.

This makes my 13" South Bend an attractive option in my mind for this sort of work - with a plentiful work envelope and a QCTP with height-adjustable tool holders for zeroing homemade involute form cutters. Through the use of a milling attachment plus a hydraulic motor and my 3/4HP power pack it wouldn't be terribly difficult to build a small horizontal milling spindle for it. This would open up the option of using true store-bought involute cutters in addition to cheap fabri-cobbled single point cutters. It would also minimize the amount of wear on my machine and the labor involved. E.g. 52 passes for a 52 root gear rather than 52 * N strokes per root. The only conundrum from there is figuring out how to index the spindle without turning the whole project into a Rube-Goldberg nightmare.

What are your guys' thoughts? Are compact, affordable indexing heads available to fit my single-slot mill and/or shaper that I'm not aware of? Should I look into electronic indexing with a stepper motor and some timing pulleys? Is there a better, easier way of doing things that I haven't thought of? Bearing in mind that this is all aimed at cutting a hypothetical once-in-a-while oddball gear that can't be purchased and adapted to suit. What's the best angle to approach this problem?

While I can appreciate the interest and enthusiasm, it sounds like a complete waste of time, money, and energy. Draw it up and send it out, the results will be much more better

 

[Just a Sparky]

the results will be much more better

Yeah, but I go down to my garage and it's projects to escape from other people and their grief - not to sit in front of my computer and deal with approval drawings, quotes, invoices, slow E-mail correspondence, etc. If I can't simply order it from McMaster and have it delivered to my door the day after next then I'd rather just settle in and do it myself one way or another.

For a commercial job I would wholeheartedly agree. But I just like to tinker. It gives me purpose and the freedom to create.

 

[Conrad Hoffman]

Any MSO (mill shaped object), a cheap BS-0 dividing head and a gear cutter is the way to go at home.

 

[EmGo]

I just like to tinker. It gives me purpose and the freedom to create.

I would be a prima donna and hold out for the smallest K&T universal horizontal I could find, and the dividing head that goes with them. Then at least you could cut helicals as well as spurs.

 

[magneticanomaly]

There has been talk here lately about gears with missing teeth. A variation on Joe MIchaels' technique can be used to restore missing teeth, where the damaged gear does its own indexing.
You set the relative positions of the dog that fits between teeth , and the cutter, so that cutter and dog wil both be precisely in tooth spaces at the same time. Dog needs to be far enough away from the area needing to be re-cut, so that it will lock in a good tooth-space for every new toooth-space that needs to be cut.
I fixed a partially-stripped valve-actuating sector from a fire truck this way not too long ago. Had no need to know how many teeth the full gear would have had, just found a cutter that fitted the space well, silver-soldered a chunk of brass in the stripped space, put a bore-sized bolt through a block of aluminum, filed a tooth-space-like notch in the Al, at aprox same pitch-line radius as the teeth, in which I ldogged down a dowel-pin to act as the indexing dog.

 

[JST]

One of the mills here is a single slot, although the work envelope is probably bigger than what you describe.

I found an L-W Chuck Co 6" swing (3" center height) dividing head that works very well with it. Found a tailstock for it also, it did not come with one.

The hold-down slots on the L-W are in-line with the key, so it works well in one slot. A lot of DHs are like that, so it ought to be very do-able.

 

[janc]

A Geared Dividing Head for Lathe, this is part of the Master Lathe Converter assembly,for grinding, milling, keyseating, broaching, as well as numerous end milling and angular drilling operations, most of which was shown in the below post, part of "Shop,Engineer, Motorized, Aviation Battalion, Machine, Couse, Type B" TM5=9127

Masters Lathe Converter Model B F.S.

 

[Just a Sparky]

So I've been thinking some more.

I've got a pile of freebie 6-wire NEMA 23 stepper motors that have been collecting dust for the past decade. I could very easily mount a 100:1 planetary gearbox on the output of one, then mount the gearheaded assembly directly in line with a spindle threaded to accept a miniature 4-jaw chuck. I would arrive at a very cost-effective and extremely compact solution with a resolution of .018 degrees without micro-stepping.

Backlash would be the only issue. But with 100:1 reduction, 50 oz-in of nameplate torque becomes 26 ft-lbs on paper. Plenty to overcome the resistance of a sprag bearing on the drive end of the spindle.

It looks like someone has already done something similar with published code for an Arduino: Snailworks StepIndex

 

[plastikdreams]

Yeah, but I go down to my garage and it's projects to escape from other people and their grief - not to sit in front of my computer and deal with approval drawings, quotes, invoices, slow E-mail correspondence, etc. If I can't simply order it from McMaster and have it delivered to my door the day after next then I'd rather just settle in and do it myself one way or another.

For a commercial job I would wholeheartedly agree. But I just like to tinker. It gives me purpose and the freedom to create.

Hey if that's your thing then have at it!

 

[Joe H]

Sparky,

Here is a Burke #4 set up for gear cutting using gear hobbs. Maybe a little complicated for one off's.

 

[MrStretch]

Easy, just buy a Schaublin 70 with the dividing head built into the headstock and milling spindle and z slide.

 

[TomBoctou]

I'd vote for whichever method you're most comfortable with and have, or can get, the tools to do.

A method you didn't mention is casting one. I've tried ramming up gears in sand (Petrobond), but have never been able to get the sand to come out of the teeth intact, plus you'd have the shrink to deal with. This can work for some parts, though, assuming the shrink isn't an issue and you have an original to work from.

People talk about using 3D printed gears for light duty machine tool applications. I've never tried it, but it doesn't sound like a path to success to me, at least not for the plastic filaments that are available to most of us.

I recently had a job to replace a broken part from a fancy-pants 1960's aluminium designer coffee table. I only had part of the broken piece, but it was symmetrical, so I modeled it in CAD - there are some good CAD packages these days that are intended for skill sets ranging from child to amateur/artist. I printed it in PLA and did an investment casting from that, but the PLA didn't burn out clean enough. Polymaker makes a filament called 'Polycast' that's intended for this - it burns out as cleanly as wax and prints just like PLA. The print is smoothed with alcohol before investment and gives a pretty good finish on the end part. I probably got $10/hour for my time from this job, but most of that was the learning curve, so the second part was a lot faster. Making money isn't the OP's stated goal anyway.

Some CAD packages have tools specifically for making gears, so you don't have to draw the gear manually. I haven't looked to see if the free packages (Blender, Fusion 360) can do this.

This method worked for me since I already had a 3D printer (they can be had new for, what, $200 these days?) and a somewhat ghetto casting setup.

I tried cutting a gear on a 7" Brand-That-Must-Not-Be-Mentioned shaper once. I did a half-assed job of grinding a HSS cutter by hand, but found that the blank had too much flex, requiring really light cuts, and the indexing head (not especially flimsy) didn't seem happy either. I gave up after a few hours, I don't think I got as many as five teeth cut in that time.

 

[ramsay1]

Cutting Gears In Home Shop

I would be a prima donna and hold out for the smallest K&T universal horizontal I could find, and the dividing head that goes with them. Then at least you could cut helicals as well as spurs.

The smallest K&T with a dividing head and lead attachment will work but even with over 40,000 different leads, one cannot get the exact lead nor can one produce a true involute....... One can, however, produce a gear that will turn the gearless boat anchor of a vertical head into a usable unit.. Cheers from Louisiana.. Ramsay 1

 

[EmGo]

The smallest K&T with a dividing head and lead attachment will work but even with over 40,000 different leads, one cannot get the exact lead

You frequently can't get the exact lead on a hobber, either

One can, however, produce a gear that will turn the gearless boat anchor of a vertical head into a usable unit..

Yup. I thought he might want to make useful stuff, but then he started talking 4" chucks ... In that case, a little mikron or something wouldn't take any space and he could make instrument gears properly, so no point to all this home-made stuff.

 

[JST]

Sparky,

Here is a Burke #4 set up for gear cutting using gear hobbs. Maybe a little complicated for one off's.

I did NOT know those were "universal" mills. I wish the small horizontal I have was universal. I will think better thoughts about Burke mills in future.