Actually, I can't accept EVER connecting tube filaments in series. I know it
is done in the old 7-tube table top radios, but it is very scary in a 3-500
amp! I would definitley recommend using a transformer that allows you to run
the filaments in parallel rather than in series.
I am not sure i understood your question, Dennis, but the idea of the
filament being +5V higher at one end than the other does not lead to
hot-spotting.
Hot spots are caused by the filament being a little thinner in some location.
Think of the filament transformer as a current source. It is going to put,
let's say, 14.5 A through your 3-500 filament. Now, if you have a spot on your
filament where the ribbon wire is a little thinner, then it will have a little
higher resistance than surrounding areas, and it will run a little hotter. I
would think in a mobile installation, the much greater risk is that the DC
voltage is varies quite a bit!
I've done some research on filaments running on DC. It is interesting. The
big broadcasters (e.g. VOA) have a much more expensive problem when they lose
one of those 100 KW tubes! They have found that they can significantly
increase the life of the tube by two methods: 1) Leave the filaments on all
the time
and 2) Reduce the filament voltage by a very small amount.
They leave the filaments on all the time to combat two problems. When you
switch on a filament, the cold resistance of hte filament is much lower and a
much larger current will flow for a short time until the filament starts
warming
up. This could be as long as a second. It is called "In-Rush Current" and
it is ultimated what burns out a filament. So, if you leave the filaments ON
all the time, you simply avoid this problem altogether. VOA started leaving
their filaments on all the time (instead of turning them on/off a couple times
a
day as schedules required), they suddenly found that tubes that formerly
lasted only months, now lasted years! This is the KILLER of tubes!
The second problem is a little more esoteric. It's called the
"Miller-Larson" effect. What happens is this: as the metallic filament heats
up, the
molecues go through different grain or alignment structures depending on
temperature. At a temperature of around 600-700 degrees C, it goes through the
Miller-Larson region where the metal molecues are in a sort of "plastic" state
where
they are easily stretched, kind of like silly putty. Remember that the
filament is tensioned by a little spring, so, as the filament wire passes
through
this region, the normally strong wire stretches slightly.
If we held constant at this temperature, then the filament would eventually
stretch and may break. But, the filament only passes through the Miller-Larson
region momentarily, as the final filament temperature is MUCH higher.
However, if you keep turning off and on our filaments, you keep stretching
and unstretching them...and this has the eventual effect of breaking the
filament. Let's say you have an imperfect filament....and if you look under a
microscope, they are all imperfect. You might have a very slight thinning in
one
area. Not enough to cause a hot spot, but it is just say .01% thinner than the
rest of the filament. If you keep streching and contracting the filament,
that weak spot will become thinner and thinner at a faster rate than the rest
of
the filament, and sooner or later, you have a 0.1% thinner area, then 1%, then
10% and the next time you turn on the amp, the filament breaks open.
I warned you...it is esoteric! But this is why filaments break when you have
a big inrush current after truning off and on a few hundred times.
Using DC makes no difference unless you dwell in the Miller-Larson region
longer. Most DC supplies are current-limited, so you might heat up slower and
spend more time in the Miller-Larson region stretching your filament apart.
The second effect, reducing the filament voltage, was proven by experiments
at VOA. In that case, they started reducing the voltage in tiny increments,
say less than 0.1%. At some point the cathode started getting
"emission-starved." The easiest was to measure this was, interestingly,
distortion, as the
tops of the waveforms, where you need the most emission, were being clipped.
They would find the point where distortion just started, then bump the voltage
back up by a few notches. They found that they could virtually triple the
lifetime of a tube by running it at lower filament voltages, typically 1-2%
less.
The lower filament voltage lifetime improvement benefit was FAR exceeded by
the benefit of leaving the filaments on all the time, at full voltage. The
cost saved in tubes far outweighted the extra electric cost.
There's a lot of mad science for all the amp guys!
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|