On 6/3/2011 2:53 AM, Steve Thompson wrote:
> The anode might not get hotter when the blower is turned off, but
> the other parts of the tube will, as heat from the anode is
> conducted into them and they no longer have the cooling effect of
> the air going past them. If there's little convection because of
> the airflow design, the temperature rise could be substantial and
> prolonged. Maybe not to the point of threatening tube seals, but
> possibly enough to degrade socket connections and/or close spaced
> components.
There is a bit of thermal inertia for the plate, and probably lesser
amounts for the grid and screen if it has one, but the critical points
are the seals. The cathode is the hottest part of the tube likely with
a bit more mass than the grid of screen.
Due to the thermal resistance in the leads there might be a time lag as
the cathode, grid, and screen cool.
> Steve
>
>> The anode and tube seals would not
>> increase their temperature, because no elements in the tube would be a heat
>> source (except the filament, which presumably is at a hotter temperature
>> than the anode, but which has such a low thermal mass that its effect can be
>> neglected)
Both the grid and screen have small mass and low dissipation ratings so
they could get quite hot, but there would be few BTUs to transfer to the
seals. OTOH what is the thermal transfer coefficient for the insulation
in the insulators. If low then the actual lead to seal temp could get
quite hot and the air flow would be important.
The cathode certainly has less mass than the anode, but it can (in some
instances) have Both a high enough temperature and mass to contain quite
a few BTUs. Far more than the grid or screen. Enough to make a
difference in seal temp?
The point is that unless one of us has run actual tests, or we have some
one who has done power tube design work (or is at least thoroughly
knowledgeable about it, we don't have enough knowledge to say whether
the cooling air could be turned off with the filament. OTOH for tubes
that use thoriated tungsten filaments/cathodes that run in the 300
watts and up range "I'd think" there would be enough mass and high temp
to present a need for a cool down. Perhaps even the indirectly heated
cathode in the YC156, but I expect the tube contacts are massive enough
compared to the cathode and grid to take care of the problem. Where
does the line fall between those that need cooling and those that don't?
I have no idea.
One thought that comes to mine though and that is the 3-500s melting the
solder out of the pin connections that fit the socket. That means the
tube seals for those leads (under those conditions) are likely exceeding
the seal temp limitations by a substantial amount, so I'd think it would
be a good practice to make sure the pins fit the sockets with very
little resistance and there is plenty of air on those pins when power is
applied. The thermal inertia and capacity of the graphite plates is
substantial. That transfers to the plate cap. On those tubes I'm fairly
sure that those seals run close to their design max so I'd want to give
the tube time to cool before turning off the power after a long
transmission. Based on the color I'd guess those plates reach some
where between 800 and 1200 degrees F. The graphite plates are massive
and the lead is small so the plates likely lose most of their energy
through direct IR radiation. Still there is a lot for that little leak
to carry through the envelope seal.
I've run 4-400s and other tubes with metal anodes well past bright red
so they are reaching at least 1400 to 1600F, but that plate has
relatively little mass, so the BTUs contained in that white hot tungsten
anode are likely considerably less than whats in the graphite anodes of
the 3-500s. I've also blown holes in a couple so with those the cool
down was no longer a concern. <:-))
I know most commercial amps don't do it with 3-500s or 4-400s, but with
those tubes I'd want to let the fan run a good minute or two after
turning off power to the filament. Course if the amp had been just
sitting there with the filament on, I doubt it'd be a problem to just
turn it off. It would be interesting if we had enough amps set up both
ways to run a comparison on tube life, but it would suffice to monitor
the tube seal temps.
Any one have the capability to measure the seal temps during operation
and after power down to see how they change and if the typical
commercial amps actually keep the seals within the limits on the data
sheets.
I think the more important point would be to let it set there on stand
by for a few minutes before turning it off if you've been pushing the
tubes for all they are worth
73
Roger (K8RI)
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|