Here are a few changes which will make this amplifier work better but keep in mind that it is Class-E and demands a good 50 +j0 ohm load. These FETs don’t handle reactive loads well at all. There are additional mods for those interesting in a CW-specific amplifier (see the CW-specific amplifier schematic on the previous page for details on the waveform shaping circuit).
1) Add a squarer to the input. This will allow 1mW low level drive from many transceivers like the K3S. NOTE: CW OPS should use the waveform keying circuit and alternate drive approach as direct drive might result in key clicks. This modification is intended for constant phase and amplitude computer-generated digital modes.
2) If you are using a modified transceiver with expanded transmit capability, consider using a low pass filter between the rig and amplifier input. Those outputs are often quite dirty.
3) Instead of using PVC coil forms, use T106-2 Ferrite. Here are some details:
L1 is 12 uH, which is 30 turns of #18 mag wire (50 inches)
L2 is 6.5 uH, which is 22 turns of #18 mag wire (35 inches)
L3 is modified as this is the low pass filter and is not ample to be legal in the US. Change the value to 13 uH, which is 31 turns of #18 mag wire. At the ferrite, parallel a 2200 pF pulse rated poly capacitor, preferably 1% (or good quality, 500v silver dipped mica TX cap). Minimum voltage should probably be 400 V in spite of really only needing about 100 V. This capacitor takes a beating.
4) Be careful with the pin-out on the BC 550 and BC 560. They are opposite of normal TO-92 transistors. Check the data sheet…twice.
5) Capacitors should be pulse rated (WIMA) but you can use silver dipped mica’s at 500V in the output network.
6) Don’t forget the additional capacitor specified in the center of the schematic labeled “472-479 mod”. They will make a big difference.
7) Use a nice sized heat sink on the FET. A fan is also good. This is one of the reasons that I can boast 15000 hours of RF emissions on the same FETs.
If you follow these instructions, you will get an amp that is better than 93% efficient that will make over 100W at 24V. Don’t push it too far, however.