Odd how these subjects come up. I was just discussing this with someone earlier today that should know better than to answer the phone when I call . Refer to AN762.pdf for all but the MRF454. The mathematics are clearly discussed in AN779.pdf for plastic TO220 transistors but the principals still hold for anything from a 2SC2166 to a 2SC2879. I've supped 2SC2166/2SC1969 into the AN779 design with little effort.
I don't have the details for the MRF454 anymore but I did the math and applied it to a bare bones 2 transistor amp with predictable results.
Bias adjustment is critical to linearity and the lifespan of the transistor. Deviate from the spec at your own risk or a quality test fixture.
Nomad is correct, no matter what the white paper seems to say. Biased "on" to AB the MRF454 IS VERY SENSITIVE TO DRIVE. It's something to be whispered at , not hit with a hammer. I learned the hard way and ended up using a 5db input pad as well as making sure AMC and ALC were in place and verified to be working. ALC overshoot and undershoot is a topic for a very long discussion .
I just posted this elsewhere today so ...
It just "might make it at 3-4 AM and 10w SSB
This absolutely will. That takes care of the input .
That takes care of the output.
The rest of the porn, I mean parts can control boards can be found here .
https://www.w6pql.com/parts_i_can_provide.htm#23cm
So where does it stop? Well it starts with a $120 two transistor box and ends up being a 1.8 to 6m piece of technological coin. Getting close to $600...
or
Ask me about it sometime...