in your screenshot, the duty cycle seems to be already very close to 50% but the drain voltage is alrady rising again at turn on. This might be caused by a load impedance which is lower than the design value. I think the point Eike was trying to make is that it you could intentionally reduce the duty cycle a bit to compensate for some degree of mismatch.

 

Best wishes,

Markus (DF6NM)

Hi Eike, 

 

you are right, the drain voltage waveform is not perfect! I have since 

corrected the duty cycle on the gate to be exactly 50%. I have also 

added two 200 V zener diodes from the drain to the source. This will 

clamp the voltage to 400 V which is quite safe for a 500 V FET. The 

zener diodes together with the fuses, I think, make this design pretty 

bullet proof. 

 

73, Dimitris 

 

PS Nothing wrong with your english :-) 

 

 

2013/2/5 Eike Katzera <[email protected]>: 

> Hi Dimitros, 

> the pulse shape in this picture does not look good: 

> http://people.physics.anu.edu.au/~dxt103/472/nltt/vk1sv-drain.jpg 

> The reason seams to be the duty cycle at the gate. Please try to change R5 

> to a 100k potentiometer and trim it to a clean single Pilse an the drain. I 

> solved the same problem tha way in my PA, wich killed my IRF540. 

> Kind regards 

> Eike DL3IKE 

> PS: sorry for my english ;) 

> Gesendet: Dienstag, 05. Februar 2013 um 08:45 Uhr 

> Von: "Victor van Kooten" <[email protected]> 

> An: [email protected] 

> Betreff: Re: LF: My transmit converter description 

> Looks very nice Dimitris,hope to see/hear you soon, 

> 

> 73 

> 

> Victor 

> 

>>>> Dimitrios Tsifakis <[email protected]> 05/02/2013 01:50 >>> 

> Hello group, 

> 

> My transmit converter is now finished and tested - it survived two 

> sessions of 50% duty cycle WSPR transmissions on 475 kHz and the 

> results are great. I thought others may find this useful, so I have 

> put a write-up on my web page: 

> 

> http://people.physics.anu.edu.au/~dxt103/472/nltt/ 

> 

> It is a 50 W class-e PA designed for 50 ohm impedance, so there is no 

> need for an impedance transformer. Doesn't get simpler than that :-) 

> The driving circuit is based on G3XBM Roger's design with the 

> difference of a 10 MHz OCXO LO and the addition of a beefy MOSFET 

> driver. 

> 

> 73, Dimitris VK1SV 

> 

> 

> 

>