[Elecraft] KPA1500 reduces power as it gets warmer?

[Ed Cole]

Slight drop (10%) in output with warming is normal for most solid-state 
amps.  Having fans run full speed might counter that, slightly.  Drop 
from 1500 to 1350 is 0.5 dB in signal (which is likely only noticeable 
if you are running a marginal propagation path (like eme).  In fact that 
is not significant on eme (normally) if you have sufficient antenna gain.

FT8 and many of the wsjt-x modes are weak-signal modes so one wants to 
run as much power as needed for the path.  Eg: yesterday I ran FT8 on 6m 
with a friend that was running a portable station for POTA.  He had 50w 
into a mobile vertical.  That was good for a -16 dB signal at my 6m-eme 
station (two 7-element LFA yagis into a K3).  I ran 1000w which was more 
than needed for the path (received 599 report since he was in 
contest-mode).  He was in the mountains only 100-miles from me.

Most amps are most efficient at their rated power: My converted Harris 
TV Linear runs 1000w at 48.6v at 50A (dc power is 2430w so 1430w is 
being dissipated as heat and indicates eff of 41%).  It uses 16 
transistors to do that.  Five 4-inch fans run full speed (always) 
because I normally am running "key down" digital modes 50-seconds out a 
2-minute Tx/Rx cycle. Good part of this Harris designed this amp to run 
continuous in ch2 AM for TV (100% duty cycle). Keeps shack warm in winter!

You should be running fans full-speed under those conditions if you want 
it to live long!  When I ran a 8877 on 2m running JT65, blowers ran 
continuously (and tube never failed at 1500w).  Talk about noise!

Good part about digital modes is one does not listen to received 
signals, so noise is not a hindrance; wear noise-cancelling headset if 
it bothers you (I just remove my hearing aids).

73, Ed - KL7UW