.......
>Whats the best value of resistor in the B+ lead to protect a power tube in
>the event of an arc ??
>
>Obviously, if you make it too large, you dissipate a lot of power, but
>ignoring the power dissipation for a moment, is a high value resistor more
>likely to protect a tube in the event of an arc than a low one ?? My initail
>thoughts were that the higher the resistance, the lower the arc current, and
>so the better. However, I'm not now so convinced this is true.
>
>If you make the assumption (possibly not 100% true) that the arc stops when
>a fuse opens, then if you make the resistor bigger, the current is less, but
>the fuse takes longer to blow.
However, a glitch resistor is not supposed to blow. It is supposed to
limit peak current.
>The parameter i^2 t (current squared times
>time), is approximately constant for a fuse. Hence doubling the resistor,
>halves the current, but the fuse takes roughly 4x as long to open. Since the
>energy in an arc is V*i*t, where V is fixed at about 80~100V across an arc,
>I arrive at the conclusion that increasing the resistor actually results in
>more energy being dumped into the tube in the event of an arc!! This seems
>counter intuitave.
>
>Interestingly, an Eimac application note says use 25 Ohms/ kV of HT, yet on
>the Eimac 3CX5000A7 or YC156 data sheets, which both run at typically 5 kV,
>they suggest using a 10 Ohm resistor. Theres quite a difference between 10
>and 125 Ohms, but I'm now wondering what would be better for arc protecton ??
>...snip...
- I have tested a number of kaput tubes that were removed from
amplifiers that experienced a 'big-bang'. Well over 90% of the kaput
tubes exhibited no gas, so an internal tube-arc seems unlikely. ......
In 3-500Zs the most common mode of failure was a bent filament helix. In
8877s, 3cx800A7s, and the like, the most common mode of failure was thin
layers of gold sputtering off of the grid. Neither type of damage can be
explained by an arc.
(see "Parasitics Revisited" 9/90, 10/90 *QST*)
Svetlana typically recommends 10 to 25 ohms of glitch R in series with
the anode/'plate' supply's positive lead. The glitch resistor must be
able to dissipate the stored energy in the filter capacitor WITHOUT
breaking down. Special, high peak energy-absorbing resistors are best,
although the cost is high. Ordinary glass-coated wirewound resistors
seem to work okay - - if used within reason. For a typical 2.5 to 3kV
supply [SB-220, TL-922, L4B, et cetera], a 10 ohm 10w glass coated unit
appears to work acceptably.
- It is important that the glitch R Not break down between turns during
a glitch. In other words, the glitch R is not a fuse - it is a limiter
of current. If a glitch resistor breaks down during an event, the peak
voltage gradient is too high, and connecting two such resistors in series
is needed.
- IMO, it is also important to connect a string of two or three glitch
diodes from the negative-HV circuit to ground to limit peak cathode
voltage to a few volts during an event. (see "The Nearly Perfect
Amplifier" in 1/94 *QST*)
cheers, David
Rich---
R. L. Measures, 805-386-3734, AG6K
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|