Hey Miles,Far be it for me (a rank amateur amidst all these professionals) to offer any kind of opinion at all, but I have gone through exactly what you are experiencing now, with very similar JavaTC results. I worked through my issues with Bart Anderson who wrote JavaTC (and even he stands on the shoulders of giants) and I got things working really peachily. And, I'm sitting here with nothing to do but type to TCML so I'm going to do it, rather than await the flood of helpful suggestions I'm sure you'll get.
Some suggestions for you that I took -You might consider measuring the "k" of your primary/secondary setup directly. It's actually quite easy, though like everything we do, requires some care to avoid getting bit. You need 2 meters to do it, which might be the hardest part.
Disconnect your primary from everything. Find a suitable ballast (an electric hair dryer works really well) and putting that in series with your primary along with an AC ammeter. Make sure the hair dryer is turned "ON" with some of its heating coils enabled. Put 110v across the series connected primary/hair-dryer/ammeter arrangement from a wall socket. Allow the hair dryer to heat up and for the whole thing to stabilize. (Needless to say but I will anyway - remember not to touch your primary...) Use the switches on your hair dryer setting so you get something betwen 5A to 10A through your primary coil.
Take a reading of the AC current. Now measure the voltage from the bottom to the top of your secondary.You will also need the inductance of the primary and secondary as calculated by JavaTC. I used an RCL meter to get a reading of the inductance directly, but it matched JavaTC closely, anyway.
Now calculate:Presuming you're in the U.S. where the frequency of wall current is 60hz (= 377 radians/sec)
Mutual Inductance = Vsec / (377*Ipri) k = M / sqrt(Lpri * Lsec) That will get you an accurate measurement of your actual "k" for your setup.What I found was that even though JavaTC suggested I could run happily at k=0.13 for my 15/120 NST + 6" secondary setup, I was usually better off around k=0.10 - k=0.11. In fact, in my current 8" coil (running 5-6kVA through a pole pig), my coil seems to run very happily at k=0.11 to k=0.12 with absolutely no racing sparks.
The other issue is flashover, of course. I found I was always better off being more conservative than JavaTC's minimum. Remember, that's a "minimum" meaning, perhaps you should stay away from that number. So with a 1.2" minimum suggested by JavaTC, I wouldn't get much closer than 1.5-1.75" in my own paranoid mind. In fact, my current 8" coil is running at 2" closest distance between primary and secondary, and I am quite happy with the performance.
I notice your proximity is actually exceeding the JavaTC minimum, and that would be a red flag to me right there as a flashover candidate.
Anyway, hope some of this is helpful, and I'll bet you get lots of input along similar lines, that will probably be more technically accurate than me.
Cheers, Joe On 12/8/2010 4:01 AM, Miles Mauldin wrote:
DC, I do think your article in the 1984 TCBA is very well written and explains the “why” and “how” of this phenomenon. However, I worked diligently with JavaTC and was very meticulous of my physical measurements and aspects of the secondary prior to construction. And in fact I posted the JavaTC results on this list with nary a comment. I have even gone back and measured the physical dimensions of my primary and secondary and JavaTC still provides the same results. .125k coefficient. 230.3 kHz = Primary Resonant Frequency 14.46 % high = Percent Detuned 0 deg° = Angle of Primary 18.1 ft = Length of Wire 3 mOhms = DC Resistance 0.196 inch = Average spacing between turns (edge to edge) 1.114 inch = Proximity between coils 1.21 inch = Recommended minimum proximity between coils 15.821 µH = Ldc-Low Frequency Inductance 0.02195 µF = Cap size needed with Primary L (reference) 0.098 µH = Lead Length Inductance 59.868 µH = Lm-Mutual Inductance 0.125 k = Coupling Coefficient 0.13 k = Recommended Coupling Coefficient 8 = Number of half cycles for energy transfer at K 17.2 µs = Time for total energy transfer (ideal quench time) There is somethig else at play here. I seem to be missing something and I have not found it. I have searched the archives for your posted parameters but to date have not found them. Can you assist me in a search string that will produce your previously posted JavaTC recommendations? All I find are suggested spark gap settings. Personally, I take offence of your comment “poor design”. If in fact it is a poor design it could only be attributed to the use of poor tools. Respectfully, Miles ________________________________ From: DC Cox<resonance@xxxxxxxxxxxx> To: Tesla Coil Mailing List<tesla@xxxxxxxxxx> Sent: Tue, December 7, 2010 2:53:38 PM Subject: Re: Racing sparks Re: [TCML] Flashover Follow The bottom line is poor initial coil design. JAVATC can do all this work, show us accurately coeff. of coupling for most all designs, aod racing sparks are trul to the parameters I have provided on this list several times before, and racing sparks will go the route of drum brakes. Dr. Resonance On Tue, Dec 7, 2010 at 12:43 PM, Miles Mauldin<teslamiles@xxxxxxx> wrote:Phil, GREAT ARTICLE! Thanks for sharing. Very clear on what's happening and why, better yet how to avoid flashover and achieve great performance. Miles_______________________________________________ Tesla mailing list Tesla@xxxxxxxxxx http://www.pupman.com/mailman/listinfo/tesla
_______________________________________________ Tesla mailing list Tesla@xxxxxxxxxx http://www.pupman.com/mailman/listinfo/tesla