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

Re: [TCML] MMC Caps

I've been using Mitch Tilbury's book ("The Ultimate Tesla Coil Design").  He's got models and spreadsheets on line.

From Tilbury's book - measuring a chain of 18 CDE 942C20P15-Ks in my coil - input parameters from JavaTC - I used max V instead of applied V for this calculation.  There was no allowance for calculation with a different peak than average voltage.

These numbers are from his chapter 5, section on polypro/paper-oil caps.  

Applied frequency of primary oscillations in kHz (f)			191.9 kHz	191870 Hz		
Capacitance value of each capacitor in µF (C1)				0.1500 uF	150000.0 pF	
Total network capacitance value in µF (Ct)					0.0083 uF
Number of capacitors used in network (N)					18		
Capacitor reactance in Ohms (Xc)						99.5392
Parallel (1) or Series (2) arrangement (SP)					2		
Loss Factor (tand)										0.5894
Capacitor Dissipation Factor value (DF)					58.94%
ESR @ applied frequency in Ohms (ESR3)					0.0059
Applied rms current (primary) in Amps (I)					11.2500 A		
Power dissipation in capacitor in Watts (Pd)					0.7494W
Rated rms current in Amps (Ir)							2.8 A		
Current multiplier at applied temperature (Imt)				1.88
Current multiplier at applied frequency (Imf)				0.05
Estimated rms rated current in Amps (Ire)					0.02 A
Derated rms current in Amps (Idr)							0.25 A

Lowest rated voltage in capacitor network in kV (Vr)			2.0 kV		
Rated voltage of capacitor network in kV (Vrn)				36.0 kV
Applied voltage (step-up transformer output) in kVrms (V)	21.0 kV		
Applied peak voltage in kV (Vp)							21.0 kV
Cell (E19) must be less than cell (E18)	
Life Expectancy			
Allowable surge voltage duration in minutes per day (SV)		8.9E+14
Life expectancy in hours (LE)							3.5E+12
Ambient temperature in °C (Ta)							25.0 °C	77.0 °F	

Now, despite all that, I have no way of knowing exactly how I have been stressing these caps with all the mistuning they've gone through.  I have no idea if the safety gaps, which have been set 2mm or so beyond the point they would fire at max V under normal circumstances - I have no idea if they fired because the voltage was 16kv or 90kv, and I do not have the instrumentation to measure that.

All of this, perhaps, is my way of suggesting that in building and designing ones own coils - guardbanding is warranted to ensure the builders spark fun enjoyment. 
Therefore, despite the cold hard facts that the 942 caps may have a longer life expectancy than the 940s by some 10s of percent (using Mitch's numbers) - they probably still live longer than most people run their coils.  

And I don't know why that guy put that info on his website  that says "KNOWN BAD CAPS FOR TESLA COILS".   Seriously, don't know if he had a bad experience or just heard a rumor.  So you can't believe what you read on websites and on newsgroups. Imagine that.

And if there is some way a guy building a coil in his garage with the instruments he's got at hand is supposed to be able to measure all that precisely, let me know.  I've got as much test equipment as the next guy, but not everyone does.

So if the price differential is there, buy the lower cost devices and have at it.  Fill coke bottles with salt water, whatever.  This is supposed to be fun, I think.

Unless you're a professional, you probably don't have thousands of dollars of test gear at hand, so it's seat of the pants engineering and word-of-mouth.

I believed everything people told me here, and most of it worked for what I was doing.  Some of it didn't.  I only slapped a scope on the coil once, and that's when someone here doubted I'd put together my Freau phase adjuster wrong, which made me doubt it.  Nope.  Worked just fine.

I'm just trying to be happy and keep my coil working.


On Aug 9, 2010, at 5:04 PM, Scott Bogard wrote:

>> The 942 series clearly has better pulse handling rating- dv/dt is an order of magnitude greater for 2kv .15uf
>> models. Current at 100khz is about 50% greater in the 942 than 940.
>> I personally have no idea how to translate that into a design suggestion for the man,
>> because the on line resources are all over the map.
> Perhaps someone more knowledgeable than I could write up an MMC calculator that shows peak current (based off of input power supply) so that we could factor this into their calculations, most of us newbs to MMC just assume voltage and capacitance are the only important factors...  It should input power supply RMS voltage and current (usually known) MMC capacitor capacitance, MMC capacitor peak current, MMC capacitor peak voltage, number of caps per row, number of rows, cost per capacitor, possibly temperature dissipation and any other important factors on the data sheet, and it should output recommended resonate, LTR and STR values, MMC current capacity/current capacity per string, MMC current draw, MMC voltage total, MMC capacitance, and any other goodies one actually needs to know when seriously constructing one of these to be bullet proof, basically everything you need...  It would be awesome if it would give basically a yes it is good in this configuration, or a no it is not, (or even a maybe but use at own risk.)  I'd reckon this should be possible on EXCEL but I don't off hand know all the formula required else I would do it, I'm still not working yet so I've nothing better to do...  Perhaps primary frequency is also important, I don't even know if that contributes to capacitor failure or not or if it is just a temperature thing, and higher frequency means more heat, though I suppose one could factor temperature dissipation and expected frequency also.  It seems to me I had a document that did this but since I can't find it, either it didn't do all that or I was thinking of something else...
> Scott Bogard.
> _______________________________________________
> Tesla mailing list
> Tesla@xxxxxxxxxx
> http://www.pupman.com/mailman/listinfo/tesla

Tesla mailing list