>
> Greetings,
>
> Ok... I think I need to try something different but
> wonder what those with experience thought. I was
> sooooooo encouraged that this would work.
>
> I took my existing circuit and had to switch to a
> lower voltage transformer so I could use the higher
> voltage one to try the test. I adjusted it to 350V
> and then from a supply built with the bigger
> transformer, applied 556 volts to it's output through
> 7.5K of resistance. Voltage rose only 4 volts. I
> then switched to a 5K resistor.... voltage rose to
> 360V at which point 38mA was measured. Dropping to
> 2.5K passed 54mA but the voltage was 403.
>
> Unless someone thinks otherwise, I think I need to try
> something a little different but...????
>
> On the topic of negative screen currents.... how does
> this happen anyway. I mean wouldn't the grid have to
> 'emit' like a cathode or filament?
>
>
> Many thanks and Happy Holidays!
>
> Randy (AD5GB)
Randy,
In a power Tetrode the conditions on the Screen change
through the RF conduction cycle - causing current to flow
into or out of the Screen (assuming that the voltage is kept
constant).
If you use a series pass regulator you will have good regulation
over the part of the cycle where the Screen 'consumes' current
but when the Screen 'sources' current the voltage at the Screen
electrode will rise (because you are not sinking the current)
this rise in voltage leads to a rise in anode current and if you're
unlucky a runaway condition and flashover.
The 'proper' resulator for a Tetrode screen supply is a 'shunt'
stabilisor, this can be acheived in several fashions, the simplest
was use of stabilisor tubes like the OA1/OB1/OA2 etc. or
nowadays a string of zener diodes. Both provide a (semi)
regulated voltage source with the ability to sink current without
going out of regulation.
Rich suggested a resistor across the output of your existing
supply - this works too by providing a constant load on your
series regulated source. When the Screen grid sources current
it passes back to ground via the resistor and your series regulator
works less hard (Norton's rule regarding summing of currents into
a junction applies).
The problem with the simple parallel resistor approach is that
there's a limit on how much current you can sink back into the
resistor, ie. before your series regulator sources zero current
and drops out of regulation. Under certain fault conditions (poor
loading, flashover, etc.) this doesn't offer as good protection as
a real shunt regulator.
Explaination and examples are shown in Eimac's "Care and
Feeding of Power Grid Tubes" if you can find a copy to look at.
Back on your regulator - the ideal configuration has your main
regulator tube effectively operating as a "programmable zener
diode" in this arrangement you can adjust the voltage to suit
circuit conditions, the anode of your regulator tube is the output
to the Tetrode's Screen grid and negative screen current is
simply 'sunk' by your regulator. Under adverse conditions (flash
over) the regulator may end up sinking the bulk of the anode
HT supply. In a well built amp this won't be a disaster since
(a) you'll have a glitch resistor in the anode supply and (b) you'll
have a VDR across the screen to deck chosen to be 50-100V
higher than the normal screen operating supply.
See if you can find the UK's RadCom magazine for October 1977
there's a good example there that uses an EL84 as a shunt screen
regulator - its a simple, neat and very reliable design.
Regards
Mike G8TIC
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|