Nano SSTC!

[The Lightning Stalker]

That will help a lot to identify the value of that unmarked coil.

 

[Fiddy]

ok L1 is 46.8uH and i have C11 in 2.2nF 3.3nF and 4.7nF

Dialing this into a resonant calculator, i get:

496kHz,404kHz,339kHz.

where the secondary is resonating at 417kHz, the closest i could get was 3nF capacitor which will get me around 423kHz

or im trying to find a series or parallel arrangement with 2 capacitors to get 3.1nF to get me closer.

 

[Sigurthr]

Unless it is a DRSSTC you have some fundamental issues in your idea to change the capacitor value. First off, SSTC's do NOT use a resonant primary, the primary reactance is simply used to limit the surge current through the switching devices. By changing the toroid you've changed the secondary's resonant frequency, which the oscillator in the driver no longer matches.

This driver appears to be using discrete analog secondary base feedback and a very odd mazzili-esque center tapped primary style topology where power is fed through center and then opposing phase sides of the primary is pulled to ground. Problem with this is that if there is any phase error between the output and the oscillator you wind up in cross conduction where both sides of the primary are passing current. Not only detrimental to the coil's performance but it presents a null impedance load to the switches which causes massive currents to flow through them, likely blowing them up. Quite frankly I am amazed this driver worked at all (a compliment to the designer who actually got it to work as supplied!).

 

[Fiddy]

Interesting. What would you suggest I do to match the oscilator to the new toroid?

 

[The Lightning Stalker]

Well, if Sigthur is right (and he probably is), find one of the higher harmonic peaks with the signal
generator again and tune to that. What was the original resonance frequency and original value of C11?

 

[Sigurthr]

I would return the coil to the as-bought state (old C11 and old toroid) and then measure the f0 with a frequency counter. Use that and C11's value to determine L1's value. Then put the new toroid on and pump the secondary with a signal generator and scope the output to determine the new f0. Then use the new f0 and L1's value to determine what value you need for the new C11. Then replace C11.

That's pretty much the ONLY way to go about it. Honestly if I was in your shoes I would just build a brand new self resonant driver. All it would take would be 1 Schmitt trigger inverter, 1 UCC31321, 1 UCC37322, and 2 MOSFETs of your choice. No GDT needed, you can use the same phase/antiphase center-fed single ended primary topology. Will work with any feedback method and will auto-tune so the coil won't fail to oscillate when you draw an arc or vary the capacitance via surroundings. If you're wondering why this isn't used more if it only uses three ICs and auto-tunes it is because the center-fed primary causes poor coupling factor and can induce high voltage stresses on the anti-phase leg when secondary energy sloshes back to the primary.

 

[The Lightning Stalker]

Isn't that where the interrupter comes in?

 

[Sigurthr]

No, the interrupter simply causes the streamers to grow instead of form the ubiquitous CW bush.

A large overhead voltage margin and recovery diodes help to keep things functioning.

 

[Fiddy]

i already know the value of L1 its 47uH* or measured at 46.8uH*

 

[DashApple]

Id suggest using a signal source to find the new resonant frequency of the secondary with the new topload , then work out the required LC values

Im assume you mean the inductor is 47uH , Not uF ?

 

[Sigurthr]

I assume you mean Henries, right? 46.8uH?

The thing about measuring inductance is that an LCR meter blasts it with a certain frequency and then measures the reactance at that frequency, typically 100kHz. Not all LCR meters down convert that to Henries (inductance), some just display the reactance (ohms). So, given the mixup in your post, double check you've got it right if you're going to skip the manual determination of L.

The rest would be the same; you need to find the new f0 with a scope and signal gen and then determine the new value for C.

 

[Fiddy]

yeah i meant uH, i can change the test frequency on my inductance meter, i test at a frequency as close as i can to what it will resonate at, in this case its highest test frequency was 100kHz