Craftsman/Atlas Commercial lathe gear cutters

John you are thinking of my sketch as if it were a dividing plate. There are no holes to count. no pins to engage or sector arms. You do need a pointer. Not having a dividing head and never having used one I don't know if they have a pointer. To be honest after reading how to use a dividing plate it sounded like a recipe for disaster.

All you do is rotate the sketch from one line to the next as you cut each tooth on the gear. If you should happen to skip a line or two you can always go back and cut the missed line The amount of possible error depends on the diameter of the circle. larger is better, and the thickness of the lines, thinner is better. I did the math based upon an 8" diameter circle and the thinnest line my printer would print. The maximum possible error was a couple of thousandths on any given tooth.

I did not come up with this on my own. I was researching how to make gears because I need a couple for future projects to be able to cut metric threads. I found an article or it may have been a you tube video where a guy explained how he made gears with the milling attachment on his lathe. If I can find it I will post a link. One of my future projects is to make a simple fixture to cut gears. Here is a rough sketch of my fixture to show the concept.

gear fixture 002.jpg

The sketch is very rough and doesn't show details like how the gear blank would be attached to the shaft or how I would lock the shaft in place.
 
I do not see how the necessary accuracy for gear teeth (spacing) can be accomplished with a piece of paper and a pointer. On my lathe, when I made my metric transposing gears, the required DP made a gear that was of a diameter that exceeded the swing of the dividing head, which is capable of differential indexing, so I made a smaller gear of a lesser DP that fit the change gears of my automatic gear cutter (antique!), then used it to cut the teeth on the gear required to match the DP of my lathe change gears. Other change gears are required to accomplish all the common metric pitches; for one, the 127 tooth gear is used in combination with a 120 tooth gear (they are keyed together on the same shaft).
Some time back, I posted a chart of the change gears that were in a article in American Machinist Magazine many years ago (with permission of AM). The article supposed a lathe with a 4 TPI lead screw, but it also pertains to any other screw pitch with some math being done.
 
John I am sure that you have forgotten more about machining than I will ever know. This is a hobby for me and I am very new to it. I don't use my lathe very much. If something comes along I try to make it if I can. I could just buy the gears that I need. I need 4, maybe 5, gears. I would rather try to make them. And I tend to think outside the box. This sometimes leads to failure. But more often than not I succeed.

I doubt that I could draw a circle and divide it into truly equal lines. I would not even attempt it. However this is easy for a computer and the printer prints whatever the computer tells it to. A pixel is .0104 wide from edge to edge. Make the tip of the pointer .010 and line the edges of the pointer up to the width of the line. If you are careful how far off will you be?

I did a quick search and found the article that I referenced in my previous post. Here is the link. http://users.tpg.com.au/agnet/cq9325rev7.html It has numerous pages where he describes in detail how he makes gears. Here is a picture of his setup.

indexer setup.jpg

It seems to work for him. I will give it a try.
 
When I looked on the link, and how the person was making his own cutters using tapered grinding wheels and burrs, it finally sunk in that he is making SINGLE CURVE gears, not involute; I'd venture to say that the two types would be unlikely to work very well when run, one against the other. Assume that all gears used on end trains of lathes are going to be made to the involute system, whether module or diametral pitch.
 
Well John you got me with a single curve gear. I tried searching for a single curve gear definition verses an involute gear and could not find anything. Lots of info on involute gears and nothing on single curve gears.

What I was using the above referenced article for was his method of indexing to cut the correct number of teeth and not the shape of the teeth. I will buy the correct cutters for gears that I want to make. I appreciate your feedback.
 
I'm thinking that single curve gears are fine for crude gearing run at low speed where efficiency is not a large issue, maybe also for clock gearing, but not for machine tools; I think the issue is having a rolling contact (DP) rather than a rubbing contact (SC)
 
Back
Top