Re: Maxwell AC vs. DC ratings & failure modes

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Crow Leader" <tesla-at-lists.symmetric-dot-net> 

It has the symtoms of a failed multisection foil cap- increased capacitance.
Metallized caps die the other way, decreased capacitance as they fail.

It's made of a series of small capacitor sections. When one fails, and foil
caps always fail short, the total capacitance goes up.

0.077uF / 0.061uF = 1.26 or about 1.25 or 5/4

Let's say the cap is really 5 series sections of 0.30uF. That gives us
0.06uF the rated value.

Say one section shorts. We now have 4 0.30uF sections for a total of 0.075uF
about what you got. I'm not sure how many secions are in this cap, but it's
probably something close, maybe 1 of 6 sections are failed.

This is how they determine if power company power factor caps are bad. There
is a chart and if the measured capacitance is too high, they assume a
section is failed, and the cap is bad.

KEN


----- Original Message -----
From: "Tesla list" <tesla-at-pupman-dot-com>
To: <tesla-at-pupman-dot-com>
Sent: Wednesday, April 07, 2004 9:02 AM
Subject: Re: Maxwell AC vs. DC ratings & failure modes


 > Original poster: "Scott Hanson" <huil888-at-surfside-dot-net>
 >
 > Terry -
 >
 > Although several list members have claimed that they have had multiple
 > commercial pulse caps failed in Tesla coil service, only one list member
has
 > actually come forward with a commercial pulse cap for failure analysis.
Some
 > time ago, Bill Vayno (are you still out there?) sent me a Maxwell
capacitor
 > that is (suspected) to have failed while in use in his coil. Bill's cap is
a
 > Maxwell model 37321 in the single-ended plastic case. This capacitor is
 > rated .06 uF at only 30 KV. While the label gives the original "measured"
 > capacitance as .061 uF, it now measures .077 uF on my Fluke 87 III, which
is
 > quite a difference from the initial value. The white plastic case is
 > slightly swollen, as are many of the "used" Maxwells I have seen. All I
know
 > about Bill's setup was that it used a 15 KV, 180 ma NST bank for power,
and
 > he supplied up to 140 VAC input to the NST bank. This could have charged
the
 > cap to 24,750 volts, assuming no resonant rise. This did not leave much of
a
 > safety factor, given the capacitor's 30 KV rating.
 >
 > Now for the "problem" with this capacitor: I have not yet been able to
 > confirm that it is in fact bad, which is why I have not opened it up for
 > analysis. Although I'd really like to see the internal construction, I am
 > holding off on any exploratory surgery until I can confirm that the cap
has
 > truly failed. Resistance measures off-scale megohms with my multimeter,
and
 > I have not yet had a chance to take it in to the lab and check it with a
1KV
 > megohmmeter. After that check, I will measure leakage current with a 25KV
 > hi-pot tester. If all still looks good, I'll functionally test it by using
 > it to replace the (2) .06 uF, 40 KV series-connected Maxwells (P/N 31642)
I
 > have been using to drive my 6" coil (serious retuning obviously required).
 >
 > If testing confirms that the cap has failed, I'll open it up and submit
the
 > photos to the list Moderator for archiving at Hot-Streamer-dot-com.
 >
 > Regards,
 > Scott Hanson
 >
 > ----- Original Message -----
 > From: "Tesla list" <tesla-at-pupman-dot-com>
 > To: <tesla-at-pupman-dot-com>
 > Sent: Monday, April 05, 2004 10:31 PM
 > Subject: Re: Maxwell AC vs. DC ratings & failure modes
 >
 >
 >  > Original poster: Terry Fritz <teslalist-at-twfpowerelectronics-dot-com>
 >  >
 >  > Hi Scott,
 >  >
 >  > I will also be interested in what truly breaks these capacitors.  I
think
 >  > we can rule out current ;-))  But that leaves us with pure over voltage
or
 >  > dielectric failure do to ionization.  I find it hard to believe that
 >  > ionization would do it and 35kV is pretty high for a breakdown voltage.
I
 >  > just don't see why these super robust commercial caps pop like pop-corn
 >  > when hooked to a simple NST...  I guess we would also have to consider
 >  > mechanical stress, but these are multi kamp caps...
 >  >
 >  > Cheers,
 >  >
 >  >          Terry
 >  >
 >  > At 03:49 PM 4/4/2004, you wrote:
 >  > >Mike -
 >  > >
 >  > >Thanks for the additional info on the failed Maxwell capacitors. As I
 >  > >suspected, there were no measurements made on the actual operating
 > voltage
 >  > >when the caps failed.
 >  > >
 >  > >Just because the system was connected to a 15 KV NST does NOT mean
that
 > the
 >  > >voltage across the capacitor was 15 KV; it may have actually been 2X
or
 > 3X
 >  > >or higher, depending on resonance effects, safety gaps, etc, etc.
Please
 >  > >post the results of your failure analysis when you open up the
 > capacitors.
 >  > >If you like, I'll pay shipping costs for you to send one of the failed
 > caps
 >  > >to me and I'll perform an intensive failure analysis in my lab at
work,
 > and
 >  > >post the photomicrographs & findings at hot-streamer-dot-com for everyone
to
 >  > >review. We've seen several failure analyses of MMC caps that died in
 > Tesla
 >  > >coil service, but nothing on a commercial pulse cap that died under
 > similar
 >  > >conditions.
 >  > >
 >  > >As for the Tesla coil driven by the 90 KV X-ray transformer, a posting
to
 >  > >this list from Mike.Marcum-at-zoomtown-dot-com dated Sat, 04 Oct 2003
 >  > >05:09:29 -0600 went into great detail describing the system, with the
 > X-ray
 >  > >transformer cooled by a car radiator & pump system, driving a giant
MMC
 >  > >array of five-hundred (!!!) Cornell-Dubilier 943 capacitors, with the
 > whole
 >  > >system drawing 24KVA (!!!!) and being capable of operation at this
power
 >  > >level for a period of 10 minutes before "the neighborhood browns out".
 > This
 >  > >posting is archived at www.pupman-dot-com.
 >  > >
 >  > >Likewise, the claim of a 50 KVA Jacob's ladder running in your
backyard
 > (two
 >  > >34.5 KV, 50 KVA pole pigs):
 >  > >
 >  > >"Don't know what the actual power draw was, but it was popping 800A
worth
 > of
 >  > >breakers (!!!!) in 5 seconds or so (made the drop wire from the pole
 > quite
 >  > >warm in the process). The arc spread over 17 feet before the breakers
 > popped
 >  > >(which was over 4" thick and bright as a welder". This also is
archived
 > at
 >  > >Pupman-dot-com, posted by Mike Marcum -at- zoomtown-dot-com, on Sun, 23 Nov 2003
 >  > >22:04:59 -0700.
 >  > >
 >  > >Are there somehow TWO different Mike Marcums -at- zoomtown-dot-com?  Could
there
 >  > >somehow be an "imposter" posting incredible claims under your name?
 >  > >
 >  > >Regards,
 >  > >Scott Hanson
 >  > >
 >  > >
 >  > >----- Original Message -----
 >  > >From: "Tesla list" <tesla-at-pupman-dot-com>
 >  > >To: <tesla-at-pupman-dot-com>
 >  > >Sent: Sunday, April 04, 2004 1:48 PM
 >  > >Subject: RE: Maxwell AC vs. DC ratings & failure modes
 >  > >
 >  > >
 >  > > > Original poster: "Lau, Gary" <gary.lau-at-hp-dot-com>
 >  > > >
 >  > > > Some information about the operating voltage might be inferred if
you
 >  > > > could supply just a few more details.  What size NST was used
(voltage
 >  > > > AND current), what size cap was used (I assume just the single
 >  > > > .03uf/35kV), gap type (sync/async), and safety gap/MOV used?
 >  > > >..........
 >  >
 >
 >
 >