As one of those that is interested, and with a wiremaker's perspective, I am
under the impression that mechanical stability of the filament,
the attachment to its posts or anchor points, had much to do about "waving".
Rigid attachment of the filament is important for mechanical
stability of the filament, and is of importantance for maintaining filament
position, and to avoid shorts due to "waving", where tolerances are close.
It seems to me that the inrush damage is not an electrical or magnetic field
issue, but rather a thermal shock, mechanical stability issue. Advances in
ceramic
mounting techniques, and other metallurgical technologies have attempted to
stabilize the filament to hold position.
Low coefficients of thermal expansion, over very large temperature ranges,
would seem to be essential here.
I have not designed filament systems myself, however I have investigated
thermal and electrical characteristics of fuse wires, and it too is fairly
complex
study.
I feel fairly sure that if we had a number of grid to filament shorted
tubes, and the time to study them, we could determine why the filament
changed position, and
if there is a common root cause for this failure.
Just my personal view here.
73 and Happy DXing,
Mike
W2AJI
----- Original Message -----
From: "Tom W8JI" <w8ji@w8ji.com>
To: "Amps Amps" <amps@contesting.com>
Sent: Sunday, August 27, 2006 10:10 PM
Subject: [Amps] Filament forces
> For those who are interested:
>
> 1.) The filament starting current can easily be 30 or 40
> amperes RMS (50 amperes peak) for thousands of turn on
> filament cycles and not cause damage. As a matter of fact
> when Eimac tried to simulate inrush damage to 3-500Z's they
> clamped a 200-amp transformer right to the pins of tubes and
> cycled them over and over again without any sign of damage.
>
> 2.) The filament current is equal all through the filament.
> It does not taper or have peaks or minimums, so the entire
> structure contributes to the MMF.
>
> 3.) With a "oscillation", the peak current with even the
> worse imaginary oscillation we could have is still limited
> to the peak emission current of the filament. That emission
> current is about 11 amperes PEAK. That 11 amperes is
> distributed over the area of the filament. It has the
> highest value at each END of the filament, and each end only
> has the possibility of about 5.5 amperes or so PEAK. The
> emission current at the very center of the filament is
> actually at a null point. It isn't hard to picture this if
> you draw it on paper.
>
> We are supposed to believe a filament that easily takes
> hundreds or thousands of 40 ampere turn-on surges that we
> know cause high current across the entire structure won't
> cause harm will suddenly be harmed by an imaginary saturated
> emission of 11 amperes (that would take a thousand volts of
> grid swing) that has maximum current only at the ends (where
> each end has about 6 amperes peak).
>
> If common sense fails us and we still intuitively want to
> believe that a few amperes of emission current can bend
> things that 50 amps of peak filament supply current can't
> bend, we can fall back on simple formulas. It's easy to
> calculate the MM force between conductors. All we have to do
> is use Biot's law. We don't need to make wild guesses based
> on comparisons between starting currents of multi-horsepower
> motors over many feet of suspended cables that have nothing
> to do with a few inches of imaginary 6-ampere currents in a
> vacuum tube.
>
> Try this link:
> http://library.thinkquest.org/16600/advanced/currentmagneticfields.shtml
>
> In a 3-500Z, it comes out to a few grams of force that peaks
> about 1/5th and 4/5th of the distance from the ends, and is
> almost zero at the middle. The force caused by normal
> filament current is several times larger.
>
> 73 Tom
>
>
> _______________________________________________
> Amps mailing list
> Amps@contesting.com
> http://lists.contesting.com/mailman/listinfo/amps
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|