Re: 15kva 14.4 kV Transformer/MMC failure (fwd)

["Tesla list" <tesla@xxxxxxxxxx>]

Moderated and approved by: Gerry Reynolds <greynolds@xxxxxxxxxx>



---------- Forwarded message ----------
Date: Sat, 23 Dec 2006 15:36:54 -0700
From: Terrell Fritz <terrellfone@xxxxxxxxxxxxxxxxxxxxxxx>
To: Tesla list <tesla@xxxxxxxxxx>
Subject: Re: 15kva 14.4 kV Transformer/MMC failure (fwd)

Hi,

The polypropylene capacitors we use are "also" used in thousands of 
industrial applications.  They really don't make them just for "us" 
:o))  They typically last "forever" for all practical purposes running 24 
hours a day.

There are just three numbers to worry about:

Peak Voltage - If you stay under the rated voltage they will "never" be 
damaged from over voltage.  Some of us have "pushed" the voltage really 
high without problem.  The caps "self heal" too.  It is pretty hard to blow 
an MMC from over voltage unless it is very extreme and prolonged.

Peak Current - The caps have a "rated" peak current.  "Never" exceed this 
one since the damage can be instant and permanent.  But those rating are 
pretty high and hard to exceed (unless you have an SISG ;-)).

RMS Current - Another fully rated number.  You can wildly exceed it for a 
"short" time since it causes heating and melting.  But if you are fast, it 
will not get too hot.  If you run continuously, then you really need to 
stay within the RMS current rating since dielectric melting is "bad".

But if you stay within these three.  There will be no problem.

After about 75 hours the caps will fail due to AC corona bombarding the 
dielectric edges.:

http://hot-streamer.com/temp/CD942C20P15K-MMC-Capacitor-Lifetest.pdf

However, DRSSTC and SISG type coils suffer from corona far less due to 
their DC nature.  Probably like a thousand hours then.  You could way over 
design for voltage though and fully eliminate this factor and then the caps 
will really last "forever"...  I think DC Cox has determined a good factor 
for this.

Voltage, Peak Current and RMS Current "can be calculated" and the caps have 
all the specs to go by.  There is no magic or mystery...  You "can" make an 
MMC last forever.  You just have to figure out the numbers.  In general, 
RMS current is the biggest worry for long cap runs and probably accounts 
for most of those failures.  Solid state coils tend to have higher RMS 
currents too.

Most of the solid state coilers are very aware of the electrical numbers so 
it has never been a big issue.  However, coils like the PIRANHA series 
"easily" push 35 amps RMS >:-)))  So MMC caps burning up is a real problem 
there and thus those caps are "really big"...

Since PIRANHA runs at 15kV I was able to get the RMS primary current down 
to "only" 30 amps :o)))  That's about 5X a big NST coil!!!

So if you blow an MMC, there is a good reason and it certainly can be found 
and corrected.

Cheers,

         Terry



At 11:33 AM 12/23/2006, you wrote:

>From: Dr. John W. Gudenas <comsciprof@xxxxxxxxxxxxx>
>Subject: Re: 15kva 14.4 kV Transformer/MMC failure
>
>Steve and DC
>I have been using a completely rebuilt DRSSTC for the past year. It
>essentially has Steve Wards electronics with modifications I made.
>Steve and I have had regular correspondence about this coil. The
>first failure was the MMC. I redesigned the MMC bank roughly for a
>20% voltage increase, but a 50% current increase.
>This bank has not failed, however, on July 4th this past summer I was
>having a prolonged run outside. The coil was running at least an hour
>@ 140 v from the variac until a gust of wind directed the streamers
>all over the place. The secondary arced and opened about 3/4 up and
>the IGBT's blew. After I killed power I felt the MMC and electrolytic
>supply and both were warm, but not hot.
>
>Since then I repaired the secondary (I think Steve will get a look )
>replaced the IGBT's and the coil performs great. I did a demo for the
>science club at the university Physics lab and the coil
>ran incredible. It is designed to produce streamers from a discharge
>point in the toroid center like a VTTC. This lets a small geometry
>coil streamers go up toward a ground area above (The hanging and
>grounded light fixtures in the Physics lab.) A newspaper reporter was
>there as well as faculty and about 45 students. The reporter had a
>great picture for the local newspaper  and I was told (I did not see
>as I was watching an LED in the coil base) that
>when I cranked to 140 v on the variac the streamers made it to the
>light fixture that I measured afterwards to be about 48 inches away.
>This coil draws around 8 amps from the 120v main.
>
>This was a 20 minute run at multiple power levels and all components
>were cool.  I don't have direct measurement data to support your
>conclusion, but I made the same inference that you did.
>I only have the "build it bigger and stronger" experience that
>supports yours and Steve's notions. I experienced equivalent issues
>with my VTTC exploding high voltage mica caps (10x primary voltage)
>until
>it became obvious that it was a current issue.  One can likely get
>away with a lot if you run your coil on short duty cycles. But the
>stress factor really goes up for long run time.
>
>John W. Gudenas, Ph.D.
>Professor of Computer Science
>
>
>On Dec 22, 2006, at 10:59 PM, Tesla list wrote:
>
> > Moderated and approved by: Gerry Reynolds <greynolds@xxxxxxxxxx>
> >
> >
> >
> > ---------- Forwarded message ----------
> > Date: Fri, 22 Dec 2006 21:39:00 -0700
> > From: resonance <resonance@xxxxxxxxxxxx>
> > To: Tesla list <tesla@xxxxxxxxxx>
> > Subject: Re: 15kva 14.4 kV Transformer.. (fwd)
> >
> >
> >
> >
> > I'm beginning to think that due to the higher efficiency of IGBT
> > switching
> > (output sparks are much hotter) that the pri currents are higher
> > and much
> > harsher on the MMCs than a typical spark gap switched cap bank.
> > Less losses
> > equals more current in the pri and higher stress on the MMCs.
> >
> > I don't have hard data to support my theory --- just a hunch.
> >
> > Dr. Resonance
> >
> >
> >>
> >> Excellent DC!
> >>
> >> I find that my MMC eventually fails (before the IGBTs provided no
> >> "special events" occur).  I have one DRSSTC that has gone through 2
> >> MMC banks now, but still has the same IGBT set.  It consistently
> >> pumps
> >> out 6 foot sparks at 1300W input.  If given the chance, id also try
> >> running one of my coils for extended periods of time to find out
> >> where
> >> the bottleneck in the design is.
> >>
> >> Steve Ward
> >>
>
>SNIP -------- SNIP ------------ SNIP