Having an "odd" number of gear teeth also is used to make sure you dont get the exact same gear teeth mating each time.

If there were an imperfection on one tooth on one of the planets, and it kept hitting the same tooth on the ring gear, it could quickly cause a problem on that one ring gear tooth.  By having an odd number of teeth, any problem is spread out more evenly over all the ring gear's teeth, increasing the life of the gear set.

So in other words it may be completely on purpose.

Kevin,
Mike and fxlarry are both right.  The sun
probably has a 50 percent plus profile shift
to strengthen the teeth.  You really need
to know or determine the exact center distances
of the parts.  First the cd of the sun and planet
and then the cd of the planet and internal gear or
annulis.  By using 3 different near size pins
and checking the measurement over pins for the
sun and planet, you can determine the involute curve
with some certainty.  You could also check the
dimension under pins for the gear.
Would be glad to solve this with this information
submitted or given.  If you cannot measure the
dimension over pins, you can probably just use
depths from the od and check the chordal tooth
thicknesses at 3 different depths to determine
the involute curves.  I would imagine that the
a standard dp or module was used, but the pressure
angle is probably also one of the standards like
14.5 20 22.5 or 25.  Would be glad to help you
solve this in acad if you can furnish the center
distances and the od of the idler, sun, and id
of the internal ring gear.  You can call me at
noon eastern standard time at 330-562-4000
ext 238 if you wish.  How soon do you need a
specific answer?  This sound very much like a
job I was working on for someone up in Canada
about 12 years ago.  Home phone 330-678-0226
in the evenings. No charge for services.
Your original statement is correct in that in
most cases you can add twice the idler to the
sun to get the number of teeth in the annulis.

Kevin and all,
This 18 11 and 49 planetary system is
interesting in that the primes as
2 3 11 and seven.  Since the sun and annulis
have different primes, they cannot be
spaced at exactly 120 degrees apart.  Does
anyone know a formula of how to calculate
the exact angles that they would operate.
Normally the sun and annulis have common
primes and determine the number of planets.
I have not seen this addressed before.
Could the above use more than 3 planets to
have different starting points like fxlarry
pointed out.  I know it is convenient to use
primes to determine the planets, but is this
necessary?  This could probably be a new
question or post.

diamondjim,
post your e-mail. I have the formula.
gearguru

This planetary (11T sun, 18T pinion, 49T annulus) can have 3 or 4 or 5 or 6 equally spaced pinions.
gearguru

In my previous statement please correct:
3 or 4 pinons equally spaced. They also have enough space to be in an assembly. Theoretically 5 and 6 equally spaced could be inserted, but they had to be ofsetted axially, because there is not space between the pinons to accomodate them.
Also: If for some reason the pinions have to be located at a different angle, then they have to move 6.0 (or 12 0r 18...)degrees from the exact - (120 or 90 degree) location.
gearguru

gearguru,
My e-mail for is
J.Geisey@juno.com
I have been able by trial and error
rotate the pinions in acad and get
them in the right position but a formula
would be nice. When you say axially
I think you mean circumferentially about
the sun?  Axially means to me up and down
on the shaft.
Thanks in advance.

diamondjim,
I really meant axially. In this planetary there is no space to locate 5 or 6 pinions next to each other. If we want to use that many, they must be "behind each other", using wide sun and annulus.
gearguru

PS

Merry Christmas and a Happy New Year to all of you who read this messages!