"Henry Kolesnik" wrote in message ...
Other questions now come to mind. Why are paper caps usually tubular but
mylar or poly are rolled but on a very flat oval cross section? I've been
told to stay away from the green and brown dipped caps that come from asian
areas as they can be a source of problems.

?? As Roy said, you probably won't find many modern paper caps.

Polyester (aka Mylar) caps come in a wide array of shapes. You can
get them as round, axial leaded parts in either molded or
shrink-sleeve packages; as rectangular-cased "radial" leaded parts; as
dipped radial leaded parts, and even (in large values) with screw
terminals. You can get them in surface-mount packages. It's really
pretty difficult to tell what's inside by the look of things outside.

FWIW, I've been conducting a long-term experiment on a couple
polyester and a couple polypropylene caps to find the self-discharge
time constant. The caps I'm testing are WIMA MKP10 polyprops and M/C
708D1 polyesters, both 0.1uF. The polyesters are exhibiting a time
constant around 4 years, and the polyprops are up in the neighborhood
of 50 years. In other words, modern caps are quite a bit lower
leakage than you'd expect from the info in sources like "Reference
Data for Engineers," and unless you are doing some very low current
work, better than you're likely to ever need, in terms of dielectric
resistance. That was NOT the case in the old days! But there are
other things to worry about. Dielectric absorption can cause trouble.
There are small nonlinearities that are a problem in very low
distortion equipment. Dissipation factor is important especially when
dealing with fairly high RMS currents. Stability can be a factor in
some designs. And of course, voltage rating and tolerance are obvious
ones. I'd guess with a little web-searching, you can find some
guidance about selecting caps with respect to these and other factors.

Cheers,
Tom