Making gear cutters to match an existing gear

Hello all,

This is all very new to me and experts will certainly frown ... but maybe this will help other folks new to machining. TLDR: below is a VERY quick and dirty way I was able to get a better shape for a hand-ground gear tooth cutter.

I want to make gears for the Sherline threading attachment outside of the range that Sherline sells. I would be willing to buy a commercial cutter, but no luck (I got one, 24 pitch, correct "number", but it did not seem to fit and was dull...). By the way, could someone on this forum advise if there is a matching one out there??? Sherline had no info for me on this. OK, so the only option, and Sherline's suggestion, and generally a good thing to have experience in, is to grind one myself.

I ground one as an experiment and it kind of worked, but I could not grind the cutter to the right shape. I believe that was to a large extent because I could not repeatably hold the cutter and the template gear next to each other in the same position to see where to grind off the next tiny bit; every time it would be angled a little differently and I was chasing my tail.

I wanted to make a holder where a gear and the cutter being ground could be held next to each other in a repeatable way. I had / have an overall design in my head, but I realized after agonizing over this for a while that it's outside of the scope of what I can do at this time. But I was able to make a MUCH less ambiguous version with three business cards, six staples, and two spring clamps. OK, and something flat to clamp it down to. It worked VERY well in relation to the effort invested.

Please find the pictures below. I scribed some lines along the center line of the bottom business card to help with alignment; they are probably invisible in the pictures. I cut one card into four pieces, stapled two of them (stacked), put the cutter blank next to them, stapled two more pieces on the other side. This worked out very well indeed; the cutter "clicks" into this "channel", can slide back and forth, but does not move in any other way (to the extent relevant for this low-precision operation). I cut up another card into four pieces, stapled them on top of each other, and put the template gear on top; this is needed because it is thinner than the cutter. The whole thing was kind of flimsy so I used the spring clamps and a piece of plastic as a backing; one of the clamps also holds the template gear; this also worked surprisingly well, the gear would not move, I had to decrease the pressure on the clamp to tweak its position.

It would be nice to have a hole in the area where the cutter and the gear meet so that you can shine the light from the other side. I did not do it because I wanted to see those scribed lines on the card for aligning the gear, but I think I will find a way next time.

I started with a standard bench grinder, then a diamond wheel on a tiny benchtop polisher to get closer. Resulting shape is not perfect but a lot better than last time. Most importantly, this time I felt that the thing I am struggling with is the grinding, not the holding-of-gear-and-cutter-next-to-each-other, so it feels like a step forward.

Extra observation 1: Sherline's gears have big "burrs" at the end of their teeth -- I am not even sure if that's the right name for it, they are like turned-over slices of aluminum. I guess they were made by a broaching-like process and that's a side effect? Any way, that probably does not affect the functionality because this outer edge of the tooth does not come into play, but it makes tooth shape irregular, makes gears harder to measure, looks ugly, and more to the point makes it a lot harder to use them as a template for a gear cutter. So I removed this burr on the lathe and lightly sanded the teeth. TBH I have not yet tested how that works, but it sure helped with the grinding operation. See one of the pictures (sorry about the poor quality but you will know exactly what I mean if you look at these gears under magnification) showing "as is" and "cleaned up" gears next to each other.

Extra observation 2: Sherline's pictures suggest that one would use the entire standard-length cutter blank in their gear cutting attachment. I had better results (as in, not stalling the lathe all the time) after cutting the blank in half (with a Dremel) so that the cutting part just slightly sticks out of the holder. (I suppose it might also help to use a heavier holder; maybe adapt their "chuck blank"? What do you think?)

Very best,
-Pavel
 
The cutter does have a sharp edge and I think reasonable "side relief", but what bothers me is that the "top" part is flat.
I make my gear cutters on the manual mill and CNC lathe and all have a flat top. Because the cutter has a relieve, grinding the top to an angle will change the gear profile.
If I am not sure about the pressure angle, I 3D print a gear to see if it matches.
If you have a manual mill, milling a gear cutter isn't that difficult.
Your gear seem pretty good!
This is how I make my gears.
 
Acetal worked SUPER and the gear came out very nice and worked great. Probably because it cuts so much easier and so there is a lot less deflection. And my third cutter was perhaps a little better. I would say that's absolutely the way to go for this application.

My big constraint is that I don't have a mill yet, only a Sherline lathe and a mill column for it, which restricts my options. Here's an almost comical setup, one of a few I tried (it did work though, I am using that gear too): https://www.hobby-machinist.com/threads/hello-from-boston-ma-usa.112713/post-1137943
 
P.S. Since the thread is soooo far away from the original topic (how to hold a blank consistently so that a more precise gear cutting is possible), it only seems suitable to post the first part threaded with these new gears. The scale in the background is 1mm. Thread pitch is 0.1mm or 254tpi. Thread shape is more like a "U" than a "V" to match the original part (also shown). The little brass cylinder (wall thickness 0.07mm) was used to hold the half-finished part (0.56 mm in the largest diameter) in a 0.7mm collet so that it does not get scratched up.

Sherline collets are *not* precise enough for this; if anyone is interested I can post a simple workaround I found for that.

This part gets screwed into the pallet fork of an old watch movement. I will need to hold it in a pin vice to do that, and I plan to use the little brass cylinder again, for the same purpose of protecting the new part from scratches.
 

Attachments

  • pivot2.png
    pivot2.png
    4.2 MB · Views: 16
This was actually surprisingly easy once the gears were made. I am not sure if you are familiar with how Sherline does it: after installing the gears you crank the spindle by hand. In this case the gears included a mix of two included in the kit, one bought separately from Sherline, and two custom made (18 and 60 teeth, 24DP, apparently a non-standard tooth shape). I think I said something negative about this arrangement above, and I think that is fair for normal threads, but it really is _perfect_ at this scale because you have exceptional control. I plan to make another post with some minor bits of advice, for beginners like myself, about this process.

The cutter was custom ground from 1mm carbide rod (in a pin vice in the boring bar holder in the quick change toolpost ... far from ideal but worked fine) to approximately match the existing thread (looks like it was made by some kind of rolling process, and it really does not seem like thread form precision is important here). Depth of cut was on the order of 3-5um per pass (1/2 to 1/3 of a graduation on the crossslide wheel), with a few spring passes at the end. The material was O1 steel (I think it should ideally be heat treated, but there is no way I am experimenting with heat treatment on _this_ part).
 

Attachments

  • sherline_100um_pitch_gears.jpg
    sherline_100um_pitch_gears.jpg
    198.1 KB · Views: 11
Back
Top