Re: Good & Bad Maxwell Caps, MMC questions?

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

Original poster: "Bert Hickman by way of Terry Fritz <twftesla-at-qwest-dot-net>" <bert.hickman-at-aquila-dot-net>

Bill,

Sorry to hear about the premature death of your cap! I'll address your
first set of questions, and will let others address the questions about
MMC's. The particular Maxwell cap you have is indeed a "good" one - it's
rated for 1000 pulses per second (PPS) with an expected lifetime of 200
million pulses (90% survival), and is designed for 20% voltage reversal. It
is an excellent choice for Tesla Coil use. However, you most likely
overvolted it by driving it from a 15 kV NST bank. Why? The answer is a bit
complex...

Your NST's output voltage 15 kV RMS. This means that the peak voltage will
be 1.414 times greater, or 21,200 volts assuming you don't aren't getting
additional rise from resonance charging effects, and you're not purposely
"boosting" the input voltage to 140 volts via your variac. Let's assume
that the tank cap is charged to the peak voltage. When the main gap fires,
the capacitor is discharged into the primary inductor, forming a "ringing"
LC circuit with an effective Q of around 10-15 (assuming typical loading by
the secondary). The circuit begins to oscillate, and one half cycle later
the capacitor reverses polarity, hitting 85%-90% of its initial voltage,
but with reversed polarity. The sudden voltage reversal severely stresses
the capacitor's dielectric system - it's equivalent to applying the SUM of
the initial voltage AND the reversed voltage - about 185% to 190% of the
initial voltage! 

For your system, it's equivalent to applying 39 - 40 kV to your 30 kV cap.
Even worse, these voltage reversals are quite rapid, occurring in tens of
microseconds or less. The typical failure mode is corona breakdown, surface
tracking, and eventual breakdown of the dielectric along the edges of the
capacitor's plates (where the local E-fields are most intense). A useful
rule of thumb is to take the RMS rating of the NST multiplied by 3 or 4
times to arrive at the MINIMUM DC voltage rating for the commercial pulse
capacitor. A 15 kV NST should have pulse caps rated for at least 45 kV and
(better yet) 60 kV. 

Next time, try to get get two of these caps and connect them in series. Or
construct an MMC... :^)

Good luck!

-- Bert --
-- 
Bert Hickman
Stoneridge Engineering
Email:    bert.hickman-at-aquila-dot-net
Web Site: http://www.teslamania-dot-com


Tesla list wrote:
> 
> Original poster: "Bill Vanyo by way of Terry Fritz <twftesla-at-qwest-dot-net>"
<vanyo-at-echoes-dot-net>
> 
> I just blew my Maxwell cap (0.06uF, 30kV, catalog #37321 - type with
> terminals on top).
> 
> I was never sure whether this was a good cap for Tesla use.  I've seen
> guidelines in the past that said those with the terminals on top
> (instead of on opposite sides) generally didn't have a long life.  It
> worked great for a while, but now it's dead (cap makes a sort of
> fizzling sound when I first power up).
> 
> Can anyone confirm that this is indeed not a good cap to use, or is this
> what I could expect with any Maxwell cap?  And if mine was bad, which
> are good?  The reason I'd like to get another Maxwell (if I can find a
> more reliable one) is that it's "plug-and-play" - no assembly required,
> and compact.
> 
> Are MMC's really any better?  To what extent are they "self healing"?
> To build a really sturdy MMC with the geek group caps (942c), for use
> with 15/180 NST (three 15/60's), what is the minimum string length and
> number of strings to handle the voltage and current they'll see?  I can
> adjust upwards to get my desired capacitance, which is between 0.06uF
> and 0.045uF --- 0.06 because that's what I was using and it worked,
> 0.045 because that's the "right" LTR value for 15/180 with a static gap,
> but I don't know (I haven't calculated yet) if my primary has enough
> extra turns that I'd still be able to tap it in tune with 0.045 (I
> currently tap around 7.5 with my biggest topload, and have 10 turns).  I
> understand that the strings need to be of adequate length to handle the
> voltage (though I'm not sure what that length is to get a long MMC life
> with 15kV AC in tesla coil use).  Am I right that there is also some
> minimum number of strings needed for adequate current handling (180ma,
> in my case)?
> 
> BTW (and I'm not sure, but I think I've heard this reported before), my
> coil seemed to be performing better than it ever had shortly before it
> died.
> 
>         Thanks,
>         -  Bill Vanyo