Duncan,
><<Could we not mount the regulator next to the radiators somehow instead
>of on the firewall. >>
>
>Do we need to? Mine is mounted on the firewall close to the air (exit) vent.
>A temperature label stuck to hte body of the regulator has read a maximum
>temp. of 70C in 200 hours of service; and this was probably during heat soak
>after shutdown or during a long taxiing hold, when the regulator would have
>little to do.
>
>
Yes, there may be easier ways to get some moving air, but I think some
moving air is needed.
>It's agreed that mounting the regulator close to the exhaust headers
>(numerous examples of this are around) gives little chance of survival..
>
One can easily be lucky for quite a while (not accusing you of this -
you measure) without being safe.
The worst dissipation of 80W probably only happens, in most cases, when
the battery is repleted fast after cranking. The heat capacity of the
heatsink smoothes out the effect of such short term excesses.
To attempt an estimate of this for the still air case:
- specific heat for aluminium is 0.9 J/gK,
- heatsink mass is 100g (just a guess),
- then heat capacity is 90J/K.
- this in parallel with a thermal resistance of 2K/W gives a relaxation
time of 1131 seconds (2 x 90 x 2 x pi).
So it takes about 19 minutes for the junction temperature to attain 2/3
of its final temperature rise, from the onset of a certain constant
dissipation (I am neglecting the 0.3K/W series resistance). A 1/3 rise
takes about 8 minutes, a 40% rise 10 minutes.
For 80W - with a final temperature increase of 184 degrees C - 10
minutes gets you up to 75 degrees C above ambient.
If then dissipation drops to 32W or below the junction temperature will
rise no further and everything will still be fine - even in still air.
If, on the other hand,...
Jan de Jong
|