Eagle module power dissipations as noted in my paper recently published for the AMSAT Space Symposium, are shown below.. The allowable dissipations presume a module temperature of 20°C in a 0°C spacecraft.

Table 1

Mean Emittance, v

Allowable Power Dissipation, Q

0.045

0.4 W

0.15

1.3 W

0.45

3.9 W

0.91

7.9 W

It would be exceptionally presumptuous on my part to assume that all modules will be of 7.9W dissipation as the spacecraft will not have that level of power generation (12 "low power" modules at 7.9W = 94.8W, plus 4 "high power" modules at 50W = 200W, totaling 295W). To achieve that level (7.9W) of power will require the use of internal heat sinking.

On the PCB dissipation, if we are to limit the temperature rise, in a 20°C module, to 40°C or a PCB of 60°C total, then the PCB dissipation would be 5.35W, somewhat less than the 7.9W allowed in a 20°C module. Thus to achieve the 7.9W other heat sinking methods would need to be used.

I fully realize that modern-day electronics are quite compact and have significant power densities, but I personally shudder at some of these above noted power densities in a vacuum environment. This is the basis of my previously noted conservatism in PCB power densities. There are some particularly critical modules that MUST be low power. These are those involved in the commanding of the spacecraft, the command receiver and the IHU. The noted reason for this mandate is that of the cooling effects of the spaceframe during eclipse and the need to maintain the command-loop equipment in suitable operating condition regardless of the radical eclipse cooling of the spaceframe. During a 2-3hour eclipse, the spaceframe can cool to as low as -80°C or below. These conditions have been observed in prior P3 spacecrafts.

This whole ongoing discussion merely further reinforces my concept that close work between the electronic designers and the thermal designer for Eagle is sorely needed. In this environment I feel that the days of wily-nily electronic design, damn the thermal torpedoes, are gone. Module design must be done as a cooperative, multi-technology effort. This is why I again emphasize that all designers should read my recently presented paper.

Dick Jansson --------------------------- mailto:rjansson@cfl.rr.com rjansson@cfl.rr.com ---------------------------