[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GE caps - the long answer ;-))



Original poster: "Jon Rosenstiel" <jonandcarol-at-adelphia-dot-net> 

I've used the 42L's, (0.22 -at- 2000V) on my 5kVA pig powered coil.  Had to run
30/string so they'd last a reasonable amount of time. At 20/string their
life expectancy was a minute or two. (Ends blown out)  I bought them
surplus, (less than $1.00 each), so having to run 30/string wasn't such a
big deal.

I should also mention that when using Terry's original 0.056uF Panasonic
caps I had to run 15/string, ("normal" was 8 to 10/string), or they'd
overheat and rupture the case. I suspect the reason I had to run so many
caps/string is that  my ballast, (Sears welder), was shoving a fairly large
"inductive kick" into the caps, well beyond what one would think. As the
saying goes..... "your mileage may vary"!

Jon Rosenstiel




----- Original Message -----
From: "Tesla list" <tesla-at-pupman-dot-com>
To: <tesla-at-pupman-dot-com>
Sent: Sunday, November 09, 2003 2:07 PM
Subject: RE: GE caps - the long answer ;-))


 > Original poster: Terry Fritz <teslalist-at-twfpowerelectronics-dot-com>
 >
 > Hi Dan,
 >
 > I should have explained more...
 >
 > The 42L3332 caps are fairly high current snubber types caps.  However,
they
 > are full metalized film rather than foil / film.  The metalization does
not
 > attach well to the end cap material in such caps and high peak currents
 > "may" blow them out.  The super high current caps use actual metal foil in
 > the plates to get very good attach and the end caps.  The 42L3332 are
 > pretty strong for a full metalized cap and they often do survive and many
 > people have not had trouble with them.  But, some people have had them
blow
 > up too.  It all depends on what the peak current in them is.  So as a
 > general rule, I would not recommend them.  The full metal foil caps are
far
 > stronger.
 >
 > Of course if one wants to use them or they are not going to super stress
 > them, I can't stop them.  But I just wanted to mention the warning:
 >
 > http://hot-streamer-dot-com/temp/GE42L3332.jpg
 >
 > The 42L3332 was actually considered a "good" cap for about a year until
 > they started failing some people.  I got a bunch of them sent to me and
 > finally found the data sheet...  Then we discovered the error.  So many
 > people said they worked fine...  If one is going to use full metalized
 > caps, these are probably your best bet.  But just don't come yelling at
 > Terry if they blow up.  Enough people have already yelled at me for having
 > them on the "good" list :-((
 >
 > The data sheet for the 42L3332 is at:
 >
 > http://hot-streamer-dot-com/TeslaCoils/MMCInfo/GE%2042L%20capacitor%20data.pdf
 >
 > Compared to the CD caps:
 >
 > http://hot-streamer-dot-com/TeslaCoils/MMCInfo/942C.pdf
 >
 > The dV/dT of the GE caps is 1000V/uS (remember you run them at 3000 volts
too!)
 >
 > The dV/dT of the CD caps is 2890V/uS.
 >
 > As an example, suppose we have a tank circuit running at 250kHz and we run
 > the caps at full rated voltage.  What is the peak current?
 >
 > GE42L3332:  Gives 3000V/uSpeak
 >
 > CD9420C20P15K:  Gives 2000V/uSpeak
 >
 > V = I x t / C  or I = V x C / t  so:
 >
 > GE = 990 amps peak
 >
 > CD = 300 amps peak
 >
 > Since the GE caps are higher value and higher voltage, they get hit with
 > 3.3 times their peak current (330 amps)!!!  In general, you can usually
get
 > away with running them at 2X the rating for a short time in intermittent
 > use.  But officially, you should not run the GE caps over 1000V at
 > 250kHz...  I suspect Dan's string has fairly low voltage on the caps
 > allowing them to survive?
 >
 > The CD caps are rated for 432 amps peak so they are "just fine!!" ;-))
One
 > does have to derate a little for higher frequencies and all, but the CD
 > caps just happen to fall in a nice "sweet spot" for value, voltage, RMS
 > current, peak current, cost.....  Thus is why we all recommend them so
much
 > (always).  There really is nothing that can go wrong with them unless you
 > run them far over 2000 Vpeak, the frequency is super high, or the break
 > rate is so high they overheat due to RMS current heating (a whole other
 > animal!).  Since LTR coils that run off NSTs are well defined, the given
 > charts are very safe and the reports of failures are just about zero if
one
 > also follows the "tips" too:
 >
 > http://thegeekgroup-dot-org/mmc/constrguide.html
 >
 > Forgive if we sometimes just "tell" people to get the CD caps.  A ton of
 > time, effort, experimentation, trial and error, failure analysis...  has
 > gone into all this stuff (6710 posts!! in the last two years, and I think
I
 > wrote 1/3 of them ;-)))...  So, we sometimes just hope people with "trust
 > us" ;-))
 >
 > I am not sure the original poster understood there is a more to MMC caps
 > than "polypropylene, voltage, and value".  MMC caps are super high stress
 > parts and we have to pick and choose them very very carefully!!  They also
 > have to take high frequency peak currents and substantial RMS currents
 > too.  It can get to be very messy trying to figure it all out.  Thus, we
 > just try to get people to use the well known and proven types of caps with
 > nice design charts so all the work is done.
 >
 > http://hot-streamer-dot-com/temp/MMCcapSales.gif
 >
 >
 > BTW -
 >
 > RMS current data is at:
 >
 > http://hot-streamer-dot-com/TeslaCoils/MMCInfo/MMCPower4.html
 >
 > CD9420C20P15K caps can be gotten from the Geek group:
 >
 > http://thegeekgroup-dot-org/mmc/
 >
 > Or from Richardson Electronics
 >
 > http://www.rell-dot-com/
 >
 > Lead times are usually 4 weeks from either.  Rell sticks you good for
 > shipping, but if you are going to get large quantities, you can save some
 > money there.  The Geek group has far better support, info, fast
 > answers....  In normal quantities, they are the best source by far and
they
 > are our pals too ;-)))
 >
 > Cheers,
 >
 >          Terry
 >
 >
 >
 > At 08:23 AM 11/9/2003, you wrote:
 >
 > >I have to disagree with that assessment.  I have been using these GE
 > >type capacitors from the beginning
 > >with zero problems running systems up to about 4kW on a single string of
 > >these.  The only time i ever smoked one
 > >was while running when the coil is severely out of tune.  I would
 > >suspect other people have had the same thing.  But, i've run these for
 > >hours (intermittently of course - say one minute on time, five minutes
 > >off) on end (for a few demonstrations i've done in the past year) with
 > >no problems.
 > >
 > >Dan
 > >
 > > > Hi,
 > > >
 > > > I must mention that people have had trouble with these caps
 > > > for MMC use!
 > > >
 > > > http://hot-streamer-dot-com/TeslaCoils/MMCInfo/good-bad.txt
 > > >
 > > > Cheers,
 > > >
 > > >          Terry
 > > >
 > > > At 06:45 PM 11/8/2003, you wrote:
 > > > >They are at Allied Electronics
 > > > >
 > > > >http://www.alliedelec-dot-com/
 > > > >
 > > > >Mfr.'s Part #: 42L3332
 > > > >Allied Stock #: 591-6130
 > > > >Manufacturer: GENERAL ELECTRIC CAPACITOR
 > > > >Description: CAPACITORS, SNUBBER, METALLIZED POLYPROPYLENE, AXIAL, 5%
 > > > >TOLERANCE, .33uF, 2000VDC
 > > > >Current Page #: 999*
 > > > >Previous Page #: 155*
 > > > >
 > > > >Qty. In Stk.: 305
 > > > >
 > > > >U / M:
 > > > >1 Ea.
 > > > >Min. Order Qty.:  1
 > > > >
 > > > >1-24 $4.800 Ea.
 > > > >25-49 $4.360 Ea.
 > > > >50-99 $4.000 Ea.
 > > > >100-249 $3.700 Ea.
 > > > >250+ $3.400 Ea.
 > > > >
 > > > >Terry Blake
 > > > >http://www.tb3-dot-com/
 > > > >
 > > > >----- Original Message -----
 > > > >From: "Tesla list" <tesla-at-pupman-dot-com>
 > > > >To: <tesla-at-pupman-dot-com>
 > > > >Sent: Saturday, November 08, 2003 6:38 PM
 > > > >Subject: GE caps
 > > > >
 > > > >
 > > > > > Original poster: Koen van den Berg <cerberus_rex-at-planet.nl>
 > > > > >
 > > > > > Hi there everybody,
 > > > > > just another (short) cap related question: Does anybody
 > > > know where I can
 > > > > > buy GE 42L3332 capacitors (rated at 2kV DC .33uF)? One
 > > > lead took me to
 > > > > > Newark Electronics, but I can't find them there, or
 > > > anywhere else...Any
 > > > > > help would be much appreciated.
 > > > > >
 > > > > > TIA,
 > > > > > Koen
 > > > > >
 > > > > >
 > > >
 > > >
 > > >
 >
 >
 >