Re: MMC cap bank - ion inception testing

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

Original poster: "Terry Fritz" <teslalist-at-qwest-dot-net>

Hi Mark,

At 12:42 AM 7/6/2003 -0500, you wrote:
>Jeepers!  Barely a whisper of MMCs for months, and then the perverbial 
>"stuff" hits the fan during *THE* first, real vacataion I've had in about 
>3 years....  Which also happened to be my honeymoon, too!  8^)

:-))


>Although not quoted, I'd like to mention I completely agree with Gary 
>Lau's comments on his Thursday post.  Dr. R, as a commercial supplier of 
>Tesla Coils for museums, schools, etc, has to design for complete 
>reliability for many years.  Us "hobbyists" probably won't use the same 
>coil setup for more than a few years (or even months!), and may only acrue 
>a few tens of hours of total runtime on one particular TC component.  IMO, 
>designing something that will definately fail isn't by default "bad" if we 
>design the failure at 1000 hours and expect only 10 hours of use.  In 
>fact, it's fairly common engineering practice to do so.
>
>With all due respect, of course, perhaps a less strongly worded reply from 
>Dr R. in the begninning would have lessened this firestorm to merely a 
>back-yard bonfire?  I know *I* read his replies as "it's going to fail; it 
>won't work; you have to do it this way...."
>
><snip>
>
>>Since Geek Group caps are so common, a simple test would be to string say 
>>5 of them together and run 7.07kV across them from an NST (be sure the 
>>cap value is not resonant with a given NST!).  Then just sit and wait... 
>>to see how long it takes for them to fail.  One has to be careful since 
>>it may burn one's house down if they flame out when your not there and 
>>there is a big electrical hazard and all that.  But that test would 
>>pretty well define the lifetime.  It is interesting to note that at first 
>>they will just self-heal.  But eventually, they will obviously go 
>>"bad".  As far as I know, no MMC has ever failed from ionization.  Just 
>>to hard to get 1000+ hours of run time on a coil ;-))  Perhaps your coils 
>>get run much more, so you may have to worry far more than most.  Of 
>>course, you buy caps most of us can't afford too ;-))
>>
>>I am not sure if Mark Broker is looking for a science experiment ;-) but 
>>maybe the Geek group could run this test safely.  It would be interesting 
>>to really know...
>
><snip>
>
>I think it's a good idea, and will discuss logistics and such with Chris 
>Boden.  I think that the hardest part of the test will be to "hurry up and 
>wait"....

I was thinking of something like:

http://hot-streamer-dot-com/temp/CapLifeTest.gif

I think that main fuse should be 7.5 amps not 0.75...

lots of fuses...


>We should define "failure" and agree on testing conditions.  We (TGG) have 
>"many" MOTs on hand, so wiring up 2 caps in series across a 2kV MOT is my 
>preferred setup.  The cap is much larger than resonant, and will 
>self-limit the MOT.  I think we have at least one old laptop around that 
>can be set up with some form of data aquisition to monitor voltage and 
>current, and shut things down automatically if "bad things" happen.

I think lots of fuses will do the trick.  I am concerned if 1-1/4 
cartridges fuses would break 2kV??  Have to short across one to be 
"sure".  Might need a few in series.  It may make some smoke too so not a 
home indoors thing.  But once a cap snaps, those little 100mA fuses should 
"flash".  Also, the thing should be sort of fire safe and safe from pets, 
kids, on-lookers...  I didn't know if you had a good place to do it or not.

The fuses might blow if there is a power surge or any line troubles since 
100mA is very sensitive.

I would say failure would be when the fuse blows due to an internal 
failure.  That would be about when they are really damaged anyway and sort 
of similar to the time they would fail in a tank cap application.  Just a 
daily check of the fuses would be fine.  Computers really are not needed 
unless you are looking for an excuse to play with test stuff ;-))

Hopefully, all 5 could be tested until they all fail.  Then we would see 
something like the NIST graph.

You may be able to "just ask" CD too...  Maybe they already know ;-))  Or, 
maybe "they" would do the test ;-))

Cool if you could do such a test or just ask your friends at CD.  It would 
be a very interesting thing to know.  It would be nice if "I" didn't have 
to do it too :o)))

Let me know if I can help or you are not able to do it there.  Then, I'll 
see if I can get it going here.

Cheers,

         Terry



>Regards,
>
>Mark Broker
>Chief Engineer, The Geek Group
>