Ampdog
R.I.P. 23 June 2022
Technical Question:
At all members old/wise enough to still understand 'ye olde pentode' valve operation (we are a dying breed ... :'( . . . F-D ??)
Gentlemen:
It concerns the indecision of some designers on what to do with an input pentode's screen grid bypassing in a typical pentode-input power amplifier. Over the years I have put this question to several forums without receiving a definitive answer except agreement with the puzzle.
Classically one bypasses cathode, screen grid and other elements of a particular stage to common. Where global NFB feeds to the input stage of a power amplifier, this normally goes to the cathode of said pentode, or said common of such a stage - meaning that the screen bypass cap keeps its connection to the cathode to maintain pentode operation* (Leak TL12, Mullard-510 and -520 et al). But some designers keep having the screen bypass cap going to circuit earth, thus effectively putting an extra 'signal generator' in series with the screen. This may be disregarded in the case of an unbypassed cathode (though hardly), but certainly not where global NFB goes to said cathode, inserting 10X - 20X the input signal there in anti-phase! In the classic Leak TL12 circuit it is done in the usual way, but suddenly in his TL25 and TL50 topologies screen bypass goes to earth as explained. This is also observed elsewhere e.g. the Radfford and a few others.
When putting the question elsewhere the general opinion was that it was an error simply out of habit. When the same designer e.g. Leak does this, it is a little more perplexing. I was hoping that fellow ably-versed colleagues would with the aid of an equivalent circuit be of help. (One could ask why worry if you do it correctly; alas, it is the scourge of researchers to 'want to know' . . . . )
To those willing to investigate this:
Thanks!
____________________________________________
* To the uninitiated: In a pentode, pentode operation is achieved only when the screen grid is an a.c. short to the cathode. As e.g., an a.c. short or then simply a direct short of the screen to anode gives triode operation.
At all members old/wise enough to still understand 'ye olde pentode' valve operation (we are a dying breed ... :'( . . . F-D ??)
Gentlemen:
It concerns the indecision of some designers on what to do with an input pentode's screen grid bypassing in a typical pentode-input power amplifier. Over the years I have put this question to several forums without receiving a definitive answer except agreement with the puzzle.
Classically one bypasses cathode, screen grid and other elements of a particular stage to common. Where global NFB feeds to the input stage of a power amplifier, this normally goes to the cathode of said pentode, or said common of such a stage - meaning that the screen bypass cap keeps its connection to the cathode to maintain pentode operation* (Leak TL12, Mullard-510 and -520 et al). But some designers keep having the screen bypass cap going to circuit earth, thus effectively putting an extra 'signal generator' in series with the screen. This may be disregarded in the case of an unbypassed cathode (though hardly), but certainly not where global NFB goes to said cathode, inserting 10X - 20X the input signal there in anti-phase! In the classic Leak TL12 circuit it is done in the usual way, but suddenly in his TL25 and TL50 topologies screen bypass goes to earth as explained. This is also observed elsewhere e.g. the Radfford and a few others.
When putting the question elsewhere the general opinion was that it was an error simply out of habit. When the same designer e.g. Leak does this, it is a little more perplexing. I was hoping that fellow ably-versed colleagues would with the aid of an equivalent circuit be of help. (One could ask why worry if you do it correctly; alas, it is the scourge of researchers to 'want to know' . . . . )
To those willing to investigate this:
Thanks!
____________________________________________
* To the uninitiated: In a pentode, pentode operation is achieved only when the screen grid is an a.c. short to the cathode. As e.g., an a.c. short or then simply a direct short of the screen to anode gives triode operation.
