In article <1994Mar8.135014.15825@ccd.harris.com> drs@ccd.harris.com (Doug Snowden) writes:
> In my reading about linear amplifier design and construction I have noticed in a couple of writings that if the filament voltage on say, a 3-500Z tube, is reduced from the typical 5.0 VAC to around 4.8 or 4.9 volts, tube life can beextended several fold.
Yes, this is true, but the increase in tube life is at the expense of
cathode emission. This is fine if you're using something huge and don't
mind reduced output.
> In an article by AG6K on QSK for the TL-922 and SB-220 he mentions that
>a good method of reducing the filament voltage (at the socket) is by
>using smaller guage wire from the transformer to the socket. He mentions that
>it should have good (teflon) insulation. My question is: Is this really the
>best way to reduce the voltage? Seems like varying line voltage would necessatate using some sort of rheostat? Any ingenious methods out there? I am planning onusing a step-start relay to initially start the filament voltage low to reduce
>inrush current on a 4-1000 tube.
That actually sounds like a bad idea to me. Why not just use a lower voltage
transformer? 5V transformers turn up now and then on the surplus market and
are fine for 6.3V tubes if you don't need to get full power. (And with a
4-1000, who needs full power? Anything more than 10W is gravy anyway.)
Inrush current limiting will only extend tube life in that it will reduce
the chances of filament failures. Whether this is important or not depends
on the particular tube type. Easy ways of handling turn-on filament failures
are to use a slightly undersized transformer that is impedance-protected, or
to use thermistors.
> I have already implemented his method of biasing the tubes by putting
>a string of diodes in the cathode return line. This is in place of using a
>high wattage zener. This works fine, as you can fine tune the idle current on
>the tubes by the number of series diodes.
Even slicker, you can use a single zener with a pass transistor. (Can I say
the word "transistor here?" Twiddle the gain on the circuit with a pot, and
you can change your bias, and if you don't twiddle it, it stays rock steady.)
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
In article <1994Mar8.135014.15825@ccd.harris.com> drs@ccd.harris.com (Doug Snowden) writes:
> In my reading about linear amplifier design and construction I have noticed in a couple of writings that if the filament voltage on say, a 3-500Z tube, is reduced from the typical 5.0 VAC to around 4.8 or 4.9 volts, tube life can beextended several fold.
Yes, but at some cost in tube Gm. Note that reducing the filament voltage
is a *must* at VHF+ because back bombardment would otherwise overheat the
cathode and lead to *very* short tube life.
> In an article by AG6K on QSK for the TL-922 and SB-220 he mentions that
>a good method of reducing the filament voltage (at the socket) is by
>using smaller guage wire from the transformer to the socket. He mentions that
>it should have good (teflon) insulation. My question is: Is this really the
>best way to reduce the voltage? Seems like varying line voltage would necessatate using some sort of rheostat? Any ingenious methods out there? I am planning onusing a step-start relay to initially start the filament voltage low to reduce
>inrush current on a 4-1000 tube.
That isn't the best method, it's just a cheap way of adding fixed series
resistance in the filament circuit. A rheostat is better, and a variac on
the primary of the filament transformer is better yet. You want some
adjustment capability because you'll want to raise filament voltage as
the tube ages and emission drops off. And a variac will allow you to
bring up the filaments slowly at power on, which will also increase tube
life.
Gary
--
Gary Coffman KE4ZV | You make it, | gatech!wa4mei!ke4zv!gary
Destructive Testing Systems | we break it. | uunet!rsiatl!ke4zv!gary
534 Shannon Way | Guaranteed! | emory!kd4nc!ke4zv!gary
