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

RE: Auto Quenching - Re: OLTC update



Original poster: "by way of Terry Fritz <twftesla-at-qwest-dot-net>" <Marco.Denicolai-at-tellabs-dot-com>

Hi Terry,

> Original poster: "Terry Fritz" <twftesla-at-qwest-dot-net>
> 
> It may seem that high coupling is good, but it tends to restrict the
> secondary voltage from ringing up to the best value too.  It 
> seems that
> hitting the secondary with a well planed single burst at 
> lower coupling
> allows the highest secondary voltage to build up in the "high 
> loss" case.

As you know, in a TC (or, better said, in two air-coupled resonant 
circuits) *every* parameter change influences the operating conditions. 
That is, even changing a single parameter among:

 - primary or secondary L
 - primary or secondary C
 - coupling coefficient

you'll be changing the two coupled resonance frequencies that is, in the 
end, the primary/secondary voltage/current waveforms.

What my paper says is this. If you have built a TC (with primary and 
secondary "tuned" as usually), you can really get 18% higher secondary 
voltage by reducing the toroid capacitance and increasing the coupling 
exactly to k=0.546.
But, as we all know, a coupling tight (not tide :) ) like that is 
difficult to achieve. And quenching should be fast too. For lower "tight 
couplings" the voltage increase is miserable.
So, it turns out that it's more useful to keep the TC "tuned" (T=1) and 
to play simply with k.

As you wrote, there are no closed solutions (i.e. simple equations) for 
the lossy case. So it's a manual process of running MicroSim (or 
whatever simulator) and picking up the secondary voltage maximum. You 
must also decide, will you always consider only the first notch, or will 
also the second one be a candidate, should you get a higher peak there 
(although difficult with losses > 0).

In the end, as Antonio wrote, a higher coupling k will always provide a 
faster energy transfer to the secondary (and back). You will spend less 
time loosing power, so you should always loose *less* energy.

In your simulations:
- are you sure you are always feeding the OLTC with the same amount of 
energy (for the single burst, I mean)?
- did you check, is there any voltage left on the primary cap after the 
burst?

Another approach is to test your findings at low voltage, rising and 
lowering the secondary, and scoping the secondary waveform. Grab 3-4 
readings including the sweet spot, then measure k at the same heights.

This is interesting stuff!

Best Regards