Homebrew ~ 400w solid state amplifier - needed no, but fun, YES!



The need for more power can be debated.
i have always wanted some more ooomph when i sat at the wrong end of
an exotic pileup. And i did also promise myself that i, one day, would build
a small'ish solid state PA. i already had the band decoder for the bandpass
filters, as you can see in one of my other project pages.

I based my project around the BLF278 FET and the PHV305 board
from greek based company RF-Source.

They rate the board very cautiously at 300 Watt's @ 1 amp bias current.




The 4 components at the bottom of the image is 4 thickfilm resistors
forming a 6dB input attenuator. This way im able to run my radio at
20 watts output, instead of 5w, to get full output from the PA.


I needed some lowpass filters for the output stage too. the PA runs in class
AB, so it's not THAT noisy, but nonetheless, i wanted a clean output so i
turned my attention to the filter kits sold by Comunication Concepts in USA.

Im running 5 filters for the 8 bands of interest,
80m - 40m - 30/20m - 17/15m - 12/10m.




Here's some images from the construction.



All the basics are mounted in the box



The band decoder i build earlier was put in a metal enclosure to secure
the logic circuits against the RF emmission inside the PA box.
This works fine. i have decoupled all the in & outputs with capacitors.








here the PA is mounted with the bandpass filter relays




The filter selection is full auto with the binary code coming from the radio.
and all the switching is done with the relays, incl. the PTT and bias circuits.




Then it was just a matter of sticking it all together and adjusting the bias
for the finals. Everything worked at first go (lucky me :-) but i had to do
some adjusting of the bias and RX/TX relays.

The RX/TX relays has to shift in a certain order like this.
when you push the PTT switch, the output relay has to turn ON,
then the input relay ON, and finally the bias relay ON.

When you let go of the PTT, the bias relay has to drop first, then
the input relay, and finally the output relay.

This is done to protect the output transistor. it has to have some
place to deliver all the energy, and if this sequence is not in place
you WILL destroy the output stage at some point.

So some form of relay sequencer is needed. I have not used a micro
processor for this project, so i had to do some tinkering with diodes
and capacitors to make it play the way i wanted, and in the end it did.





The back of the PA enclosure




The front of the PA enclosure




And here it is, alive at last :-) working like a charm, with a good clean output too!.




Giving around 350-400 watt's on all bands, im a happy camper.
now i will have that little edge in the pileup.


DX Cluster