Hi Philip have a look at this website http://www.valvewizard.co.uk/pre65 wrote:
I just get a headache thinking about the theory,have tried reading MJ but it goes over my head,just like it did when i was an apprentice at Marconi in the 60s.That's why i left and joined the motor trade !!
It's biased towards guitar amps but the guy has a very friendly and easy to understand style of writing. He goes into the all the circuit types needed to build a valve amp eg triode gain stages, pentode stages, output stages both PP and SE.
If I had known about this site when I started, it wouldn't have taken me about half a dozen amps before I learned how to design power stages properly. Every one of them I did was woefully under-run until I read his article on loadlines for SE power stages. He explains how to do them far more clearly than MJ.
Even then, after taking my latest SE amp to Nicks place, after he had measured it I was still under running (not enough HT) The penny finally dropped about a week ago, and the amp sounds absolutely brilliant after I set it up properly by changing the mains TX for a higher voltage one.
I suppose what I am saying is that it takes a lot of experimenting, making mistakes and help from such as Nick, Steve S and Paul plus copious reading up before one can finally be confident that a design will perform as one thought it would.
That I finally realised the other day why most of my amps other than the Rocky EL34 were often underpowered and shouty, is more of a tribute to my own dogged perseverance in trying get a handle on power stages than any inherent ability on my part. However now I have a good idea what I am doing I can approach the design of the PX25 breadboard free from doubts. That's with about 8 amps under my belt.
What I would say Philip is to have a go at designing something like a KT66 triode connected SE amp, posting the schematic and asking for comments, advice etc. Also peruse the website I have linked to.
The feeling you get when something you have designed yourself starts making music is hard to describe.
Steve