On 4/29/2012 11:02 PM, Jim Brown wrote:
> On 4/29/2012 5:24 PM, Roger (K8RI) wrote:
>>> Most of the tube designs have been significantly "over engineered".
>>>> That is, the tubes have excess plate dissipation for the power output
>>>> and grids that will take a significant beating.
>> Which is a plus for tubes.
> Naw, it's a plus for conservative design.
> One thing that seems to be missed in this discussion about home brewing
> amps was clearly pointed in a comment in this thread -- the key to solid
> state amp design is THERMAL design and protection. Those are very
> different skills than most amp designers of our generation learned, so
> at least in part, it's an "old dog new tricks" problem, not a limitation
> of the technology.
It helps to have a computer systems design background<:-)) Although
many systems are running right at edge of the device limits.
Even with the best heat transfer using the latest heat transfer
compounds such as Arctic silver which has a thermal resistance of less
than 0.010 watt per square inch you are then limited by the internal
thermal resistance which may be more than the compound between the
device and heat spreader.
Arctic Silver 5 has a thermal conductance of 350,000 W m^2, and a
thermal resistance of 0.0045W per square inch. Achieving these figures
takes mastering the technique of spreading the compound and then
thermally cycling the required times. Often the mating surfaces have to
be polished, or lapped flat enough that when clean surfaces are placed
together you can not pull them apart with a straight pull. The
installation requires working the compound down to a thickness of 0.001"
to 003". The copper heat spreaders between the devices and heat sinks
are usually 1/4 to 1/2" thick.
Many devices rated at a KW are only good for a couple hundred watts on
SSB and there's none of this long tune up, key down.
The protective devices or rather circuitry is usually far more complex
than the amplifier circuitry. Also as the devices are usually very
broad band you use low pass filters band switched in the output as even
the smallest harmonic output from the exciter gets amplified just as
much as the fundamental. There is also such a thing as a device with too
much gain. which can lead to instability when operating on the lower
end of the frequency range. There is also protection from input on the
wrong band as these are no where near as forgiving as a tube amp driven
for a short time on the wrong band. It only takes one over voltage
spike to take out a semiconductor although they are improving.
The protective circuits when taken one at a time are quite simple, but
taken as a whole can appear quite intimidating. OTOH setting them up
with the proper values and time constants can be a real chore.
Fortunately that can normally be done without the expensive and fragile
devices in the circuit.
> 73, Jim K9YC
> Amps mailing list
Amps mailing list