Yes. That's right. Also the comment about number of teeth meshing while cutting (about 3 or 4 in that example). You shouldn't be too far off with making say 10 passes with the blank rotated a bit and the hob going up/down.
Nick
Another amusing gotcha. If you do end up with a good involute profile, it won't run cleanly with other gears if they've been cut with B&S type cutters. The B&S cutters use a compromise profile that is significantly different from a true involute for small numbers of teeth.
Mark Rand RTFM
Is that the voice of (bitter?) experience, Mark?
Tim
Not that bitter. I was expecting some issues because I had read articles that mentioned problems. I have made one 14 tooth pinion with the shaper and it doesn't roll against the other gears when hand held as smoothly as I would like or as smoothly as pairs of 'compatible' gears work. The pinion seems to work reasonably with less engagement than it should have.
I think I need to make a depthing tool as used by clockmakers. Then I can experiment to find the best engagement and cut the pinions to suit. I can also experiment with changing the form of the cutter to get a better action. If necessary I can use a conventional modified addendum but I'd have to re-cut the mating gears slightly.
There are three of these 14t pinions in the HLV apron gearbox and they all need making on the shaper. either because the adjacent journals are larger than the root diameter of the pinions or because, on one of them, the teeth are 22/29 stub form.
Mark Rand RTFM
Had me puzzled too....
Regards, Tony
Even with a hob, because the hob has distinct teeth and the cut is therefore interrupted, the cut is actually a series of distinct steps. A large number of them though.
Regards, Tony
Yep - you are right. I spent a while slaving over a hot CAD system to show exactly what happens with these kinds if cutters. With a small tooth count like he was using, only 3 of the rings of teeth in his "hob" actually do any cutting, and you end up with a tooth form that is a very crude approximation to the involute form. Even with relatively high counts the extra rings of teeth don't actually do much to improve the tooth form.
If you rotate the blank by a fraction of a tooth and move the Z axis "down" by the corresponding fraction of the module size, and repeat the process, then the tooth form approximation improves. You can make the form as crude or as good as you like by deciding how many of these shifts you want to make per tooth. By 4 passes per tooth, the yooth form is starting to get pretty reasonable.
In reality, for non-critical applications, you can get away with the one cut per tooth approach; for example I recently cut a small gear to replace one in the winding mechanism of a friend's clock The tooth form isn't great, but it fits and it works for the desired function (and saved him lotsa ££ in repair costs). So it is a valit method; the compromise is time versus desired degree of accuracy for the application. For a one-off where you can't justify the cost of a commercial cutter it makes a lot of sense.
Regards, Tony
That's the way.
For a clock, I wouldn't hesitate to do it this way. Their teeth aren't/needn't to be (really?) very precise due to the very little load. I also think they do have something quite different to involute or hypocycloidal (SP in English?).
Nick
I disagree - the gears that are nearest to the mainspring or the barrel that carries the driving weight actually carry quite heavy loads for the size of gear & the material (usually brass). The big issue in clocks is running friction in the "going" train, so I wouldn't choose to use the "one cut per tooth" approach for a clock, as the bad tooth form that resulted would also result in higher friction. However, with a higher number of cuts per tooth, it wouldn't be a problem.
Traditional clocks use a cucloidal tooth form; however, this particular clock is relatively modern and the gears are involute. Actually, this gear carries quite a high load, but as it is in the winding rather than the "going" train, the fact that it will probably be not ideal from the point of view of the tooth friction won't be an issue.
I believe the choice of cycloidal for clocks was based on the idea that, in the typical clock application where you are massively "gearing up" (large count gear driving small count pinion), the cycloidal offers lower friction than involute. However, as the cycloidal form used in clocks is much more sensitive to accurate "depthing" (centre-to-centre distance) than involute, I suspect that most clocks lose all the (small) theoretical advantage of cycloidal as the mechanism wears. Certainly, modern mass-produced movements seem to use involute & they runn well enough.
Regards, Tony
[cut]
Another benefit of the epicycloidal form, is that with the generating circle 1/2 the diameter of the pinion pitch circle the pinion dedendum is a straight line which makes it much earier to cut (in steel). As the (smaller) pinion is always driven, providing it has a reasonable number of leaves (say >10) the addendum profile is not particularly critical. Hence early pinions are cut with what is in effect a slitting saw and the addendum is rounded by some less precise hand method. However since involute pinions can (now) be made easily by hobbing this argument only remains relevant for very small gears (e.g. watch pinions < 0.5 modiule) where hobbing is not possible (or at least so I'm told).
The "minimize contact before the line of centres" friction reduction argument in favour of cycloidal gearing is certainly valid when compared to badly cut involute gears! However, it does not appear so valid if the gears are well cut with smooth profiles.
Another benefit of involutes is that the depthing (centre to centre) distance is not nearly so critical with the involute profile, whereas cycloidal gears (at least traditionally) need to be "depthed" in a tool to ensure that the action starts at the correct place in relation to the line of centres.
Alan Bain
There's an interesting and possibly handy site at:
Richard
On Wed, 03 Oct 2007 20:10:39 +0100, Richard top-posted:-
I've just downloaded a program from Marv Klotz's site written by Johan Smit. It calculates the button sizes, spacings and infeeds to make B&S type cutters. So I could make a single point cutter out of a bit of O1.
Mark Rand RTFM
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.