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

Re: [TCML] Voltage - Gap



Hi Carl,

Capacitors in TC applications are often stressed in a multiple ways simultaneously. Major aging stresses include internal temperature (a function of dielectric losses, metallic plate and interconnect ohmic losses, RMS current and operating voltage) and voltage-related stress (peak voltage and % voltage reversal). Its no wonder that they often fail, sometimes spectacularly.

A somewhat dated but excellent paper by Frederick MacDougall (Capacitors for the Pulsed Power Industry) describes some of these stressing factors and also provides some insight into evolving capacitor technologies:

http://www.aerovox.com/Portals/0/PDFs/t6.pdf

Regarding the effects of DC voltage stress, MacDougall mentions that the estimated lifetime for a number of film capacitor technologies can be expressed as the following power law:

Life at V2 = (Life at V1)*(V1/V2)^n

where n is a function of the dielectric system, and V1 is normally the rated DC operating voltage.

"n"  Dielectric system:
---  -----------------
 5   Kraft paper-Foil
7-9  Kraft paper-PP-Foil
15   PP-metalized electrodes

For high Q ringing applications (such as TC's), you might want to make V1 the capacitor's faceplate rated voltage (Vdc) divided by 2.

Another paper on the same web site provides some useful insight, "Energy Storage in Polymer Laminate Structures - Aging and Diagnostic Approaches for Life Validation":

http://www.aerovox.com/Portals/0/PDFs/t11.pdf

Note that operating a capacitor at 80% of its rated voltage results in lifetime limited by thermal aging. By also keeping the RMS current at or below rated limits, there is no reason why capacitors shouldn't last virtually "forever" in TC use.

Bert
--
Bert Hickman
Stoneridge Engineering
http://www.capturedlightning.com
***********************************************************************
World's source for "Captured Lightning" Lichtenberg Figure sculptures,
magnetically "shrunken" coins, and scarce/out of print technical books
***********************************************************************

Carl Noggle wrote:
True. Such things often have interesting power dependencies. The life if
an incandescent light bulb is inversely proportional to the thirteenth
power of the applied voltage. Bandsaw blade wear, the seventh power of
the excess cutting speed. Wonder what the dependency is for cap voltage,
rep rate, etc.?

---Carl





Another thing to keep in mind is that the shot life is based on the
rated voltage of the cap. If your system's operational voltage is
significantly lower than the cap's rated voltage, the shot life of the
cap will increase just about logarithmically (you can get away with
running a 1 pps rated cap in your coil if you're running 15 kVAC max
but your cap's working voltage rating is like> 60 kV) and it will
probably last a life time for the hobby level coiler.

David Rieben

On Feb 3, 2012, at 6:45 PM, Jim Lux<jimlux@xxxxxxxxxxxxx> wrote:

On 2/3/12 7:39 AM, Joe Mastroianni wrote:
I think David also remembers the incident and has provided some
detail. I was probably exactly on the hairy edge of overvolting, and
definitely over current. If you search back in the TCML archives you
can probably find my postings about it. It was sometime last year.

As I said, when I paralleled the 37667s I never had another problem
with them in my SG coils. So maybe it was overcurrent. Somewhere I
have the JTC files and I'll send them to you off list when I can dig
them out.

That larger, 50kv rated cap had a very low max - rated reversal
rate. Far too low for coil service. I didn't bother reading the
specs when I got it on eBay for really cheap. I thought I was lucky.
Ah, here's the failure
pics<http://www.flickr.com/photos/iceowl/5437138473/in/photostream>


the caplife paper cited earlier gives all the scaling rules. VR isn't
a huge factor, but peak voltage is.

The other thing to look at is the rated shot life. The 1pps flavor
have shot lives like 1E5 shots. That's a very long life in a Marx
bank kind of application where they may do a shot every 10-15 minutes
or something, during an 8 hour work day. At 120bps, you can burn
through the rated life in less than an hour.

_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla
_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla

_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla





_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla