Just rethinking that through over lunch - yes - I'm sure
that's correct And I'm unanimous in that! ("Are you being
served?"?)

For the blank to rotate one tooth's distance at right angles
to the hob's teeth, the hob must be rotating at 1/cos times
the RPM used for spur gears.

Orator For Decency wrote:
That appears to match my own analysis except in one
respect. As you say, "This has the effect that as the hob tooth "moves
down"
by one tooth space so does the blank and so the teeth are cut cleanly
and to size."

A movement of the blank at right angles to the line of cutting and
traverse of the hob teeth, is no longer co-axial to the axle of the
blank - the cosine factor comes into play (or, rather 1/cos) so that
the blank rotation is transformed into the movement by one tooths space

"Boo wrote:
Orator For Decency wrote:
If one is hobbing helical gears, and the direction of traverse
is parallel with the cutting teeth of the hob, and the blank
is set over at the required helix angle, and the rotational
speed of the blank is adjusted by the cosine of the helix
angle, then is it necessary to have an automated traverse
feed at a specified rate, or can one do it at manual traverse
as with spur gears?

Hi,

I've never cut a gear so you should not take this as gospel but the
following is what pertains AIUI :

If you have the spindle set over wrt the table travel by the helix angle
of the hob then all you need for cutting spur gears is that the rotation
of the blank is the same as the spindle rotation speed divided by the
number of teeth. This has the effect that as the hob tooth "moves down"
by one tooth space so does the blank and so the teeth are cut cleanly
and to size.

In the case of helicals, AIUI the setup is the same with the only
difference that you additionally must angle the blank by the gear's
helix angle wrt the table travel direction. The same relationship in
respect of the rotation speeds applies : AFAIK there is no sin or cos
term in the rotational gearing it's still just 1 / (number of teeth) as
for spur gears.

John Stevenson's website had good details but the link
http://homepage.ntlworld.com/stevenson.engineers/lsteve/ didn't work
when I tried it a moment ago.

Hth,

--
Boo