>
>Hmmm, if there are no good rules of thumb and no peak current/energy
>ratings for anode
>supply components, how does one go about sizing the glitch resistor? I
>guess you start
>out small and keep making it bigger until parts stop failing, or perhaps
>start out big and
>make it smaller until things start failing :):)
>
// The purpose of the glitch-R is to protect the electron tube from
failure during an intermittent oscillation condition. For thoriated
tungsten type tubes, the danger is to the wire filament/cathode. During
an intermittent VHF or UHF oscillation., much grid current flows in an
HF amplifier because the tank is a low-pass design. . Since grid current
originates in the filament, the resultant, perpendicular magnetic-force
pressures the filament sideways during the oscillatory event. In an 811A
or 572B, this force can cause the relatively thin, 4A filament wire to
break and fall into the bottom of the envelope. In a 3-500Z, the
thicker, 14.7A filament wire doesn't break. it bends in the direction of
the grid. How much it bends seems to be a function of how much
peak-current can be delivered by the anode supply during a glitch.
Limiting peak-current is the job of the glitch resistor.
- For oxide-cathode type tubes, the cathode can not move about because
it is a rigid nickel cylinder - so, the danger is elsewhere
Oxide-cathode type tubes employ gold-plating on the grid to improve
conductivity and to reduce primary electron-emission from the grid during
signal conditions. According to Eimac, during an 'oscillation
condition', gold can evaporate. This may seem unlikely since gold has a
high boiling/evaporation point. However, due to RF skin-effect, current
is concentrated at the surface of the grid, so only the surface of the
gold is heated to the evaporation point. When gold vapour cools to
1063ºC, it forms smallish solid gold meltballs that are visible with a
30x microscope. Since solidified gold is conductive, an anode to
grounded-grid arc may occur. Another problem is that since gold inhibits
electron-emission, gold meltball contamination poisons the cathode. Gold
evaporation is driven by stored energy in the HV filter-C, so limiting
peak current in the HV supply is a way to prevent gold evaporation. A
suitable glitch resistor in series with the positive HV bus will happily
do this job - but only for those who have more brains than money.
>BTW, speaking of grid components, how come Drake grounds the grids on the
>L-7 thru
>a parallel combination of RF chokes and capacitors? Seems like one could
>achieve the
>same goal by grounding the grids directly with a low inductance strap
>right to the chassis.
>
>73 de Mike, W4EF..............
>
>----- Original Message -----
>From: "Tom Rauch" <w8ji@akorn.net>
>To: <amps@contesting.com>; "Ian White, G3SEK" <g3sek@ifwtech.com>; "Mike"
><W4EF@dellroy.com>
>Sent: Tuesday, September 11, 2001 5:44 PM
>Subject: Re: [AMPS] SB-1000 mods?
>
>
>>
>> > Wouldn't the 4CX5000 tend to have a larger filter capacitor, and thus
>> > a bit more energy to dump, or is peak current the real concern? With
>>
>> The 4 by 5 will have higher voltage, but the tube will easily handle
>> more impact than a small tube. Every system needs to be looked
>> at in view of the total system.
>>
>> > Rich's example of the 10 ohm glitch R in the SB-220, and assuming some
>> > stray R in caps and chokes, you are still talking about a peak current
>> > pulse on the order of 180 amps!! Seems like an awful lot of current.
>> > What can the average plate choke withstand??
>>
>> There is no average plate choke. That is the problem with rules of
>> thumb.
// The HV-RFC in a 1500w-pep HF amplifier has c. 1-ohm DC.
>>
>> The most important thing is to not put a glitch resistor in the grid. It
>> belongs in the HV lead.
>>
>> 73, Tom W8JI...
- R. L. Measures, 805.386.3734, AG6K, www.vcnet.com/measures.
end
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|