From mgarvi–(at)–anix.com Mon Apr 29 12:10:22 CDT 1996
Article: 13709 of alt.guitar.amps
Path: geraldo.cc.utexas.edu!arlut.utexas.edu!news.io.com!imci4!newsfeed.internetmci.com!panix!not-for-mail
From: mgarvi–(at)–anix.com (Mark Garvin)
Newsgroups: alt.guitar.amps
Subject: Screen grid dissipation
Date: 25 Apr 1996 23:52:32 -0400
Organization: PANIX Public Access Internet and Unix, NYC
Lines: 57
Message-ID: <4lpha0$26--(at)--anix2.panix.com>
NNTP-Posting-Host: panix2.panix.com
I recently posted something stating ‘You’d better have screen resistors’,
and I was asked about this via email. Since Carl has been neglecting
his tech support duties, and hasn’t even found time to tell anyone to
“EAT SHIT!” lately, I figured I’d post on this briefly.
Note that these are my opinions, and further discussion is welcome:
The query was in regard to vintage amps that have no screen resistors.
When screen grids are less positive than the plate, screen dissipation
is usually minimized: electron flow is diverted toward the plate. So
no dissipation problem unless the screens are misaligned (see below).
But in a classic pentode output stage, screen grid voltages are fairly
constant. When control grid voltages go upward and the tube turns on,
the plate voltages swing more negative than the screen grid. This
leaves the screen grid as the most positive element in the tube. Screen
current (and therefore dissipation) goes up, and is usually limited by
the screen resistors. (Note that this is a dynamic (AC) condition,
so screen resistors must be unbypassed to remain effective).
At lower volume levels, even the transient currents thru the screens
are not a huge problem, but this depends on tube quality of course.
Problems arise at higher volumes, as you would expect.
But the REAL problem occurs when that nasty Marshall impedance selector
plug falls out, or if speaker load is much higher than it is supposed
to be. Pentodes have a relatively high internal resistance, so the
plate has to struggle a bit to pull down when connected to a 3K to 5K
primary. BUT the 3K (or whatever) primary spec is counting on the
specified output load! With speakers disconnected, the primary imp.
(reflected back from the secondary) goes very high. Result: no more
struggle for the plate to pull very low. Screens are then absorbing
lots of current…and with no screen resistors they will easily burn.
To my experience, this problem is not so severe when screen resistors
are in place, and slight mismatches of output impedance aren’t going
to instantly toast the screens. But there are usually a lot of
variables in play.
Another consideration is that many new EL34’s do not have properly
aligned screen grids (they are not in the ‘shadow’ of the control
grid), and the grids will catch a lot more current than usual.
I’ll leave all this to be confirmed by our resident expert.
What do you think, Carl?
>Carl:
>>EAT SHIT, YOU STUPUD SCREEN GRIDS!
Thank you, Carl.
I know this is a news group about ugly singers.
I’m sorry to be so far off topic.
Mark Garvin
