amp gurus, can you explain me why ...

[hunter]

My H&K Triamp MkII has a half power switch, and when I biased a set of Mullard EL34 valves, I had set them to save 35mA at 480V.

But when I activate half power/50W, it seems all 4 valves stay active, but on two of them the plate voltage rises to 500V and the bias goes from 35mA to 40mA.

Is that any kind of weird circuit in the amp? I'd like to use both modes (100W with the metal band, 50W with the cover band) but I fear that the New Sensor Mullard RIs might die if they run at 40mA/500V. Sound wise the 35mA setting is perfect.

What so you think is happening in the circuit and what would you advise?

I was considering getting a quad =C=s as I guess they could cope with the higher current, but 4 of them is a bit expensive (considering I should get a retube on the preamp, which is 9 valve$$).

Thanks for any help, highly appreciated!

[HTH AMPS]

When you switch to half power, the circuit is turning off the current through two of the EL34s.  This means that the plate voltage rises slightly as the output stage isn't fully loaded.  I wouldn't worry about the change of 20v, it's not gonna harm anything.

If you do the maths, biasing the two EL34s with a plate voltage of 500v gives 35mA (per valve), whereas all four EL34s with the plate voltage at 480v is 36mA (per valve).  This is really not worth worrying about. 

[hunter]

When you switch to half power, the circuit is turning off the current through two of the EL34s.  This means that the plate voltage rises slightly as the output stage isn't fully loaded.  I wouldn't worry about the change of 20v, it's not gonna harm anything.

If you do the maths, biasing the two EL34s with a plate voltage of 500v gives 35mA (per valve), whereas all four EL34s with the plate voltage at 480v is 36mA (per valve).  This is really not worth worrying about. 

Thanks for your reply!

Ok, that sounds good. So the fact that it's rising to 40mA per valve is no issue in the 50W mode?

[HTH AMPS]

nah, it's not a problem Hunter.  if you're really concerned, knock the bias down to around 36mA on the pair that run hotter.