Rich Measures wrote:
>>
>>If you float the grids, it's a good idea to take additional steps to
>>protect the exciter from a HV dump through the cathode circuit
>>because the odds of having it happen increase.
>>
>>
>>73, Tom W8JI
>
>With glitch diodes in the (cathode) metering circuitry, and a glitch R
>in the positive lead, the designer avoids such.
>>
I just tried a SPICE transient model for the effects of an internal tube
arc from anode to grid, from a typical B+ supply with 25uF charged to
3kV, with a 47 ohm glitch resistor.
With typical "SB-220/clone" grid bypassing, say 3*200pF in parallel with
a 1mH RFC (10 ohms series resistance), the grid voltage spikes sharply
positive, by hundreds of volts for several milliseconds.
During this time, the grid and cathode (or filament) will behave like a
diode, and will pass a high current - as much electron current as the
cathode/filament can emit. That current also flows down through the
cathode bias circuit, so there's every possibility of damage.
It's the inductance and resistance between the grid and chassis ground
that causes the problem. Using 27 ohm "grid fuse" resistors instead of
the RFC will avoid the initial sharp spike, but will still result in
high positive grid voltages (how else can the resistor "fuses" blow?).
Grounding the grid as directly as possible to the chassis (say 100nH +
0.1 ohm for a pin-based tube) reduces the maximum grid voltage spike to
a harmless 10V.
Admittedly, internal arcs are only one of the many possible kinds of
"glitch"... but the non-grounded grid is very vulnerable to that
particular kind.
If anyone wants to try some SPICE transient modeling, pick up the free
demo of SiMetrix from www.newburytech.co.uk and give it a spin.
73 from Ian G3SEK Editor, 'The VHF/UHF DX Book'
'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.com/g3sek
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|