Re: MMC Cap Results. WOW!

[Terry Fritz <twftesla-at-uswest-dot-net>]

Hi Garry and All,

Wow! This is cool!  

For the record, Gary's cap is simply 15 x 0.056uF 1600 volt Panasonic caps
in a single string.  The RMS current is from 3 to 3.7 amps depending on how
many caps Garry uses for his system.  As it turns out, RMS current is
fairly consistent for a given coil as the primary capacitance varies.  The
current may get a little high for a single string but the caps tend to be
run under voltage which increase their current handling a bit too.  This
cap has taps between each of the small caps so that Garry can select a
fairly wide range of values.  Garry reports in another mail that the 6.23
tap works best which is slightly below the resonant point for his NST.
MMCs can be made to have very fine adjustment although this has not been
pursued very much.  Probably time to look into this more since find
capacitance tuning is fairly easy and seems to offer advantages...

BTW - Adding more strings will not change the performance.  If you add a
string, you will have to double the number of caps in the strings so the
losses stay the same regardless.  Also, the losses in an MMC are really
insignificant overall and can be ignored other than the effect loss has on
cap heating.

The racing arcs can be stopped by raising the secondary an inch or two
higher above the primary to lower the coupling a little.  Somebody asked
what increments to use for this.  I would "guess" 1/4 inch for a small
coil, 1/2 inch for a medium coil, and 1 inch increments for a large coil.

Many thinks to Garry for enthusiastically helping with my little pet
project!  I could not have asked for better feedback and the opportunity to
study a first timer's experiences with MMCs (actually this is an EMMC) and
how they compare to the usual "cheap" caps.  Garry is only using 9 caps at
6.23nF so the cost at that point is only about $15.  It good to know that
it has not blown up yet too.  Since Garry has blown a number of other caps,
we can see how he finds the reliability of this compared to his others.
The cap is running right at 1600 volts peak per cap at around 3.5 amps now.
 The recent NST study suggested that this configuration, even at resonance,
should be reliable.  

BTW - Be sure the safety gaps and main gaps stay properly set.  The voltage
can skyrocket to like 60kV at resonance if they were to both fail.
Although, there is an odd chance the cap would act as a safety gap, I would
not count on that! ;-))

I don't recommend others go trying single strings.  Garry's cap is highly
experimental and has a fair amount of planning behind it.  Others may blow
the caps easily in different coil configurations and I would think the
chance of getting it right randomly would be low...  Hopefully these
results can be used to find ways to design them reliably, but right now, it
is all "bleeding edge" stuff.

Cheers,

	Terry

>
>
>At 09:00 PM 10/3/2000 -0700, you wrote: 
snip...
>This is the results from my off-sized secondary of 24" of AWG28 wire wound 
>on a 3.85" Schedule-40 PVC form sanded clean.
>
>Toroid 8" X 2"
>
>Spark gap is vacuum static gap using coiled aluminum flashing electrodes to 
>conduct heat away, built in a 4" PVC pipe with a large muffin fan attached 
>to pull air through pipe and expell it outside the enclosure. Gap is 1/2"
>
>NST 12/30
>
>Capacitance Turns Primary for best tuning Arc Length
>
>3.72 22 12-14"
>
>3.99 22 14-16"
>
>4.30 22 15-16" - Three inch increase in streamers here and quickness of 
>buildup to breakout
>
>Streamers are about 12"
>
>4.66 22 18" Archs!!! Streamers gained another 5" to equal about 15-18"!! 
>Instant breakout.
>
>5.09 21 20" Archs!!! 21" huge streamers!!
>
>5.61 21 About same length but more frantic sparking.
>
>6.23 18? 22-24 streamers, flashover until tuned to 18 turns in the primary 
>then results about
>
>21" sparks.
>
>5.61 seems to be the best setting I could get without getting LONG racing 
>sparks on the secondary. I was not sure if the racing was caused by too 
>close coupling or being out of tune.
>
>After 5.61, it bacame difficult to see much difference. I suspect I would 
>get better results from a larger toroid
>
>I tried a 4" by 18" dryer duct toroid but there was no breakout and sparks 
>at the 5.61 setting was about 8" I believe this was because the toroid was 
>so big it was radiating most of the energy instead o storing it.
>
>Thanks again. YOU MADE MY DAY!! I hope this information proves most useful. 
>I am going to insert a big gulp cup into the inside of the secondary to 
>prevent internal breakdown that could be occurring.
>
>I will keep you apprised of further tests.
>