[Amps] Ceramic valve exhaust air temperature

[Alan Ibbetson]

I'm looking for a sanity check on some calculated ceramic tube exhaust 
air temperatures predicted by the data sheets.

Given the blower inlet temperature, the Specific Heat of air (Cp) and 
the rate of air delivery, simple arithmetic predicts the exhaust 
temperature at a given tube dissipation: temp rise = watts/air mass per 
sec/Cp. Or have I got this wrong?

The reason I ask is that the Eimac data sheet for the 8877 says it needs 
7.8 cfm air at 50C ambient to keep the seals at a safe temperature when 
the tube is dissipating 500W (plus 75W for grid/heater). If I have the 
"simple" calculation right this results in an exhaust air temperature of 
134C above ambient, so around 160C (320F) in an averagely warm shack. Is 
it really going to be this hot? It seems to me that the cabinet 
paintwork could be damaged as well as people/pets.

The data sheet air volumes at higher dissipation rise in much more than 
simple linear progression, presumably due to the kinetics of the airflow 
over the tube and socket. Hence the predicted exhaust temperature for an 
8877 at 1500W anode dissipation with the recommended 38 cfm airflow is 
74C above ambient (so maybe 100C). This will still burn you but at least 
the paint on the amplifier cabinet probably won't melt.

Has anyone actually used an 8877 at 500W Pd (a fairly typical UK-legal 
figure) on, say, RTTY with a puny 7.8 cfm blower? Did the cabinet feel 
like it was about to catch fire? Or are my numbers all messed up?

The reason I ask is that there is no warning about the possible dangers 
of these exhaust temperatures in the Eimac data sheet, which might lead 
novice builders astray. I would feel happier with a target of 30C 
temperature rise, which requires 35 cfm for 500W anode dissipation, 65 
cfm for 1KW and 100 cfm for 1.5KW, all at 100% duty. I suspect that the 
backpressure will defeat most practical-sized blowers even when asked to 
deliver 65 cfm, let alone 100 cfm, especially in a conventional 'through 
the SK2210 socket' configuration. Maybe you could get closer with the 
K2RIW approach of pressurising the anode compartment?

73, Alan G3XAQ