Radford STA100 improvements?

[bigwill]

I have been tasked with rebuilding a Radford STA100. I have heard these things eat KT88s as they're run them at full tilt with 600V on the screens and anodes (being ultra-linear).

The circuit can be found here: The Free Information Society - Radford STA100 Electronic Circuit Schematic

I was thinking it might be worthwhile to run the amp with fixed screens at 400V or so (using a simple regulator of sorts), seeing as I've heard that modern production KT88s don't appreciate such high screen voltages (especially in UL as they sit at a higher voltage than the anodes )

I was also considering wiring the phase splitter valves as triodes with a CCS in the tail, depending on the anode loads I choose will probably lower the gain a bit (I haven't seen what the EL184 looks like in triode mode), but this may be compensated by the increase gain by running the output valves as pure pentodes.

Am I on the right track here? I don't want to totally warp the design into something new but I feel there are worthwhile improvements to be made with this circuit with regards to linearity and reliability...

[artosalo]

Hi supply voltage like 600 V for KT88 does not "kill" them if the anode current and total power dissipation is kept within the specification.
In the attached schematic the total power is some 36 W, which should be very safe for the tubes.

I was also considering wiring the phase splitter valves as triodes with a CCS in the tail, depending on the anode loads I choose will probably lower the gain a bit (I haven't seen what the EL184 looks like in triode mode), but this may be compensated by the increase gain by running the output valves as pure pentodes.

To modify from UL to pentode mode is surely not an improvement.
I think that the phase splitter with two pentodes is just fine. Why do you want to modify it ?
If this stage is converted to triode mode, the gain drops significantly.
This also necessitates that the GNFB-loop should be re-built.

[bigwill]

Originally Posted by artosalo

Hi supply voltage like 600 V for KT88 does not "kill" them if the anode current and total power dissipation is kept within the specification.
In the attached schematic the total power is some 36 W, which should be very safe for the tubes.

I've heard many people say that 600V on the screens of current production KT88s is not a very good idea - and the STA100 was known to wear them out quickly anyway. I'm not going about this as a serial 'fiddler' as some people like to do, I just want it to be reliable.

Regarding the loss of gain with triode wiring the phase splitter - I was fully aware of this happening and I was just throwing ideas out there. I just glanced at some curves of an EF184 in triode mode and it appears to have a mu of 55 and an Rp of 3.7k, which is a decent amount of gain, especially as the first stage is a cascoded 6DJ8. Whether or not it could drive the EF184s as triodes - I'm not sure, I'd have to look into it but I'm currently standing up writing this as I'm between some jobs outdoors