The problem is capacitor heating and using capacitors with a large
temperature coefficient. Typical of general purpose capacitors. You need
NPO or COG type dielectric, which are larger and more costly. I'd bet on
the tuning padder, then the loading padding capacitor. At 160 meters the
fraction of the tuning C that is the tubes coupled through the coupling
capacitors makes variations in the coupling capacitors insignificant.
They'd be more of a problem at 10m where a large fraction of the tuning
C comes from the tubes.
Some ceramic capacitors can loose 75% or more of their capacitance going
from room temperature to 100°C. These are the least expensive and most
compact, and the worst possible for tuned circuits. Been there, tried
that, and failed.
Sometimes engineers specify the right capacitors and purchasing skips
the engineer's specification to buy the cheaper capacitors and then
things like tuning drift show up.
You can check the capacitors for drift easily if you have an antenna
analyzer or noise bridge. With the amplifier powered down, connect a
resistor from plate to ground equal to the normal plate load impedance.
Connect the antenna analyzer or noise bridge to the antenna connection
of the amplifier (might need to jumper around the relay to get connected
to the output tuned circuit). Tweak the tuning for a perfect match
indicated by the test equipment. Apply heat from a hot air gun or hair
dryer to one capacitor at a time, when you find the capacitor doing the
drifting, it will show rapidly. You can confirm the drift by cooling the
suspect capacitor with freeze mist (used to be at Radio Shack, I've not
checked lately). Photo shop canned air will work for cooling also, but
not as effectively as canned refrigerants. This technique will directly
identify all the capacitors that are drifting, with no hazard to the
operator making the tests.
73, Jerry, K0CQ
73, Jerry, K0CQ
--
Entire content copyright Dr. Gerald N. Johnson, electrical engineer.
Reproduction by permission only.
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|