From: Jason A. @ 
Date: 3/9/2002 6:08 PM 
Subject: For kg--Miller Capacitance? 

ken,

I wanted to ask about the effect of Miller Capacitance in output stages. I read in one of G. Weber's books a while back about how the effect of Miller Capacitance in EL-84 amps [such as the AC-30] has a play in tone. Apparently, the author of the article stated that a lot of the magic of EL-84's is lost when trying to use more than 4 EL-84's in a standard push pull output stage. Personally, I have never tried more than two parallel tubes/winding in an output stage [quad], but I notice that a lot of HiFi guys like to design using more than 4/each winding. I would think that some of the better manufacturers [like VTL in my opinion] get great results and have some damn nice sounding amps that go beyond what most of us have tried here. Tubes like the EL-84 and EL-34 have a bit higher amplification factor than other power tubes, and I was wondering what you thought about this. The way I see it, is if some excellent designers can get away with producing a superior sound using a lot of parallel tubes in thier HiFi output designs, why would this be so horrible in a guitar amp?

Jason

From: kg 
Date: 3/9/2002 7:46 PM 
Subject: Re: For kg--Miller Capacitance? 

miller capacitance can be easily dealt with... you just design a driver with lower impedance. miller effect per se isn't really much of a problem in multigrid tubes operated as such, since the screening effect of g2 works so well. there is still the heavier capacitive load imposed by the parallel tubes to contend with, which is where the lower driver Z is required.

the problems with multiple parallel output tubes as i see it are:

a) the grids can interact with each other very easily. to isolate them, you can install grid stoppers for each tube, but this can also lead to other problems, described in c) below...

b) there is some merit to the long held hifi belief that a single tube sounds clearer than multiple parallel tubes. this has to do with the impossibility of having an exactly matched pair of tubes. as a result, the tubes will "fight" each other to some extent as to what exactly the output SHOULD be at any given instant. folks have said the parallel tube tends to defocus the sound when compared to the single tube.

c) including those grid stoppers described in a) can actually exacerbate the problem, since now the electrode voltages will be dependent on the electrode CURRENTS. 

this is almost like the problem of synchronizing the carburetors in a multi-cylinder motorcycle.. if they are not properly synced then one cylinder will always be working harder than the others, and you will never get max power output.

this may also be why the ultra-hifi purists tend towards single ended designs, with a single output tube, since a PP amp is really operating the tubes in parallel.

so there are some mechanisms which would explain the differences in tone when parallel multiple tubes, but i don't find increased miller capacitance to be that convincing, since it can and should be dealt with in the driver stage. however, imperfect valve matching is highly likely, especially in terms of dynamic characteristics, and to me sounds more feasible an explanation for "loss of mojo."

ken