[Amps] cooling big tube seals

[John T. M. Lyles]

There are good reasons against using suction in an amplifier for 
cooling. Any small holes, seams in the chassis will aid the egress 
(or is is ingress?) of dust and dirt. By pressurizing the chassis 
from an external blower, all dust accumulation can be handled at the 
intake scroll of the blower, where a nice filter can do its job. 
Pressurized systems are much cleaner inside than evacuated systems. 
Most commercial broadcast transmitters have followed this practice, 
although there are a few which haven't, and they are filthy inside.

I agree with Bob about the requirement to cool the seals, but 200 is 
pretty darn hot for them. We had a bunch of RCA 7651 tetrodes running 
at that temp (on the filament button contact) and they wouldn't last 
more than 6 months in service. Using Tempilaq paints, I have tried to 
get cooling to where I can be under 200 C, more like 150-175. Another 
technique is to use a Luxtron fiberoptic thermometer to probe the 
seals while it is operating. These don't come cheap however, and they 
are very expensive to operate (>$500 per probe which are fragile 
glass fibers).

At BE many years ago, we learned how to cool the filament stem of the 
4CX20,000A/8990 tetrode without adding another blower. We just put a 
small Teflon pipe pushed up under the tube, so that it surrounds the 
center filament 'button' contact. This is the part which often 
tarnishes black during operation, from the elevated temperatures. 
This pipe then leaked out of the bottom of the amplifier to ambient. 
It was a 'controlled' leak, so that a quantity of cool air (from the 
pressure in the box) would blow out via the filament stem to the 
outside. Completely separate from the anode air, which flowed up 
through fins to the lower pressure upper cavity area where the plate 
circuit was located. For the TH781 and TH628 (250 and 800 kW 
dissipation tetrodes) I am designing with at present, have separate 
blowers (LARGE) for the stem cooling. These are ~1 HP machines 
themselves!

73
John
K5PRO


>
>Mark the critical points that need to be met when cooling a tube is 
>to keep both the plate and cathode-heater seals below their maximum 
>ratings. Typically they are between 200 and 250 Degrees Celsius. 
>Sucking air from the top of the anode in most cases does not provide 
>sufficient airflow over the cathode connections to keep them below 
>their ratings. The water cooled variant of the 4CX10K requires 30 
>CFM to keep the cathode seals below their rating. You can get around 
>this problem directing a second source of forced air at the base. 
>For all of the extra trouble this entails it's more practcal to push 
>the main source of cooling air up through the base in the first 
>place.
>
>Adios
>
>Bob K1TA
>
>Mark Beckwith <mark at concertart.com> wrote:
>Could you wise old sages help me with simple question - I seem to recall
>hearing that forced air cooling was not the same as, shall we say, sucked
>air cooling. Do I remember right? Seems you could destroy your amp if you
>tried to suck the air through the tube(s) tops as opposed to pushing it
>through from the bottoms.