Hi Paulo,
hmm, you are getting closer, at least
OK, first i see that you have wound the coils rather loose. I have them tightly wound around the cores, as you can see in the images in the docu. This will affect the inductance to some extent, so that might be part of the reason.
Next, you have only 3 caps installed for C30/31/32. I have a third one placed on top of them, making a total of 4. I can't see it in you images, but at C20/21 only one cap has to be installed instead of two.
In your post #130, L4_Input shows 56 Volts pp, while RF_Output shows around 80 Volts pp. That can't be right. Are you sure you got those right?
In post #132, L3_Output it shows 232 Volts pp, which seems a bit high considering the output voltages. It might just be that there is too much RF reflected, which then kills the FET somehow. While i can't meassure the voltage there, a image interpolation with a lower voltage waveform tells me that i have around 200 Volts pp there in the image i posted here, resulting in about 85 Volts pp at L5 input, and about 100 Volts pp at the RF output.
Please install that extra capacitor, and make sure that there is only one installed at the input of L3 (positions C20/21), and rewind the toroids so that the windings sit really tight on the core. The XFMR looks fine already. Oh, and bend L3 away a bit from the XFMR1, in the image it literally leans onto it.
Other than that it already starts to look good. I assume you used a "trustworthy" supplier for the IRF510. I have seen cheap China FET's that didn't hold up to the specs in other projects... Keep an eye on the voltage/waveform at the drain of the IRF510, which goes into the center tap of the XFMR. It shouldnt go too much above 60 Volts pp. If it does, too much RF is reflected back somewhere.
Just to make sure, you can try to do the measurements with L6 (right at the RF Out pads) removed. There just might be something fishy going on in that section that throws of the matching. After all, that is the only part that is in direct contact with the RF output signal (besides the output filter obviously, of course).
When i was experimenting with the coils/caps in the output filter section, i blew a few FET's myself, so that filter section is a rather sensitive area.
Oh, and while it may sound stupid to ask, but i can assume you used proper capacitors in the filter section? That is, with the proper voltage rating? As you can see, rather high voltages can appear there, so the caps must have a suitable rating.
At one point i blew the cap at C13, in the supply line to the XFMR, at the moment i switched the unit off. While the circuit still worked somewhat, it also messed with the FET (and overall signal quality). That's why i now have three in that position. Lot's of current draw there. So the caps are a really important factor here.
Greetings,
Chris
Edit: Oh, and you can scale down the scope images, for example to 800x600 like i did. There isn't more information in them than that anyways. Only eats up bandwidth