[Amps] step start

[jim.thom jim.thom@telus.net]

Date: Wed, 4 Oct 2023 11:28:25 -0500
From: MU 4CX250B <4cx250b at miamioh.edu>
To: Paul Decker <kg7hf at comcast.net>
Cc: amps at contesting.com, amps-request at contesting.com
Subject: Re: [Amps] step start


Hi Paul,
I agree generally with your take-always, though I still don?t see the
merit of using so much filter capacitance in the HV supply.  I?m
guessing you?ll be using a big bank of electrolytic capacitors, which
is likely to be the most trouble-prone part of the entire amplifier.
You?ll also be generating a lot of unnecessary heat in your bleeder
resistors and, frankly, it would make me nervous having that much
stored energy in a 2500V power supply: you could have a real death
trap on your hands if the bleeder resistors ever opened up! Your HV
supply will be more reliable if you reduce the filter capacitance to
ten percent of your intended amount. (In this regard, it?s not a bad
idea to follow the lead of commercial amplifiers, which use a much
smaller capacitance than you?re planning.)
      I?m not sure what screen voltage circuitry you have in mind, but
if you want to use a ?stiff? power supply which achieves its voltage
regulation by being able to deliver much more current than is actually
needed, then there are better choices. In my homebrew 160/80m amp, for
example, which uses three parallel, grid-driven GU74B tetrodes, the
maximum screen current of all three tubes is only 45 mA, and yours
will probably be roughly similar. (Note that some amplifiers, the
Collins 30S1, for example, use a circuit that grounds the screen grid
and routes the full cathode current through the screen grid power
supply. That is an unfortunate design, in my opinion, and should be
avoided.)

My homebrew amp uses a screen voltage regulator which supples 330V DC
to the screen grids, with 1% regulation, and which costs about $10 to
duplicate. It is important to protect the tube if there is flashover
in the HV supply (or, worse, a screen-to-plate internal arc in the
tube) If this, unfortunately common occurrence happens, the plate
voltage of the tube will drop to zero and the screen grid will act as
the plate, and bear the full cathode current. This is a situation that
can destroy the tube almost instantly. A properly designed screen
voltage regulator will protect the tube in the event of a flashover,
and is a simple addition to a screen regulator circuit.

To me, overbuilding high voltage power supplies (screen or plate) in
order to gain a bit of voltage regulation isn?t generally advisable,
since simple, inexpensive, and reliable alternatives are available.
73,
Jim   W8ZR

## If he is using a series string of lytics in the HV B+  supply, typ a
100K @ 3 watt  MOF is wired across each lytic.  Diss in each 'eq' resistor
 is typ .9 watt  ( Just under 1 watt).  The chances of any of the 100k eq
resistors opening up is virtually zero.  If the voltage across each lytic
is 80% or less of it's stamped 450 vdc rating, the lytics  will last a long
time, and be trbl free.   I bought 300 x 100K  @ 3 watt mof's  from mouser,
in 1% Tolerance..and all 300 of em measured exactly  99,980 ohms.
Tolerance between any of em is zero, and the V drop across each lytic is
identical. If run at 80% or less of it's 450 vdc rating, the leakage
current  through the caps is through the floor.  The typ 3-4 ma of eq
current more than swamps out the leakage current by a huge ratio, like 30
X  or more.

##  A  50 ohm glitch resistor in series  with the B+ would limit fault
current to a safe value.

##  A 500 volt  GDT (gas discharge tube) wired between the screen  and
chassis, would / should save the screen supply in the event of an anode to
screen arc.

##  Glen,  AE0Q  at Alpha tells me the GU74B, when it fails, it's typ an
anode to screen arc, which destroys some of the screen supply
components..and typ is a $30 fix in screen component parts.  Alpha used a
800 Volt  GDT, between screen and chassis, and the arc is so fast, the GDT
does not fire. I previously suggested they use a GDT, but he said they
already did, but it's probably sized way  too big.

##  Dunno why the GU74B, when it nears the end of it's life, has the
dreaded  anode to screen arc.  It would have to either gas up, or lose it's
vacuum, or perhaps screen grid migration, for that to happen.   The only
'fix'  would be to periodically  hi - pot test each tube...between anode to
screen....and replace when it's too low, and an arc is imminent.

##  In the OP's case, he really needs a much lower value bleeder on his
screen supply.  Like 10-25 k.

Jim   VE7RF