Re: Help! My Coil Blew Up

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

Original poster: Finn Hammer <f-h-at-c.dk> 

Sean,

Thanks for reminding me of Ken`s promising method. I seem to forget badly 
these days, it is only 3 months ago that I tried to make it work, but couldn`t.
I got vastly different positive to negative pulses, but have since learned 
that it is no good to use 1n4007 diodes for that task.
I`l stock up on RF diodes and have another shot at it.

Cheers, Finn Hammer

Tesla list wrote:

>Original poster: Sean Taylor <sstaylor-at-uiuc.edu>
>On Thu, 05 Feb 2004 18:04:03 -0700, Tesla list <tesla-at-pupman-dot-com> wrote:
>
>>Original poster: Finn Hammer <f-h-at-c.dk> Tesla list wrote:
>>Steve,
>>I`m fond of the antenna feedback circuit, but I`m not going to flame you 
>>for pointing out that it is flawed.
>>It is still the best we`ve got right now.
>
>What about the capacitor in the ground path idea that Ken Herrick uses?
>It gives a reliable signal, and a 90 degree phase shift is built in so 
>when the voltage across the cap is at a max, the current is at zero - or 
>more easily detected is the zero voltage crossing - where current is a 
>maximum.  Ken can go into more detail since he's had a lot of experience 
>doing this but I believe a capacitor in the ground gives a much more 
>reliable signal for sensing when to switch and keep the system in resonance.
>
>>However, when I swich from 494 to comparator, it is obvious that the 
>>antenna does not drive the coil at it`s res. freq.
>>
>>It appears to me, that it would be good to ad an adjustable phase shift 
>>option between the antenna and the comparator but I don`t know how to do that.
>
>There are several methods for adjusting phase shift, but I suppose you 
>want something with a constant phase shift over a wide range of 
>frequencies.  All you need to construct is an all-pass filter with a dual 
>variable resistor (I think that'll work - just need to keep the poles and 
>zeros reciprocals).  But more likely would be something with a constant 
>time delay to compensate for switching time.  Of course, a time advance 
>can't be made, so a delay of a full cycle is the only option, but the 
>obvious problem is that the length of a full cycle changes with Fres, 
>which is exactly what we're trying to track changes of.  So, a PLL is the 
>best thing I can think of to actually work well, but of course that is 
>getting way to complicated.  Maybe some DSP on the primary current 
>waveform to try to keep it as clean as possible by changing the switching 
>time? :-)
>Okay, enough "good" ideas for one night . . .
>Sean
>