[Amps] High voltage MOSFETs

[Manfred Mornhinweg]

Jim,

> ###  has anybody actually used a tuned circuit in any 
> of these proposed configurations ?? 

I have often tuned up solid state amps. While I have never so far seen 
one that has tune and load controls on the front panel, or anything like 
that, many solid state amps are tuned designs. That includes almost all 
VHF and higher amps, and most HF amps designed for single-band use, such 
as CB radios, single-band ham transceivers, and commercial/industrial 
things like RF welders.

 > I just cant fathom
> manually tuning up a SS amp..... unless  you were 
> really careful, and pulse tuned it with a low 3-10 % duty 
> cycle, and also starting with low drive power. 

It's not as touchy as you imagine. Most typically it's a matter of 
taking a screwdriver and tuning two trimmer caps for maximum output. 
It's good to watch the current at the same time, because there is often 
a range over which the power output almost doesn't vary, but the current 
does. In those cases you tune for the lowest current on the max power 
plateau, or in some cases such as solar-powered VHF repeaters you tune 
for best efficiency.

And if a trimmer runs out of range, you simple bend the turns of the 
coil a bit together or apart, to change its inductance, and then 
readjust the trimmers.

I have done it this way with dozens of ham repeaters from brands like 
VHF Engineering, RPT Electronics, Hamtronics, lots of VHF radios by 
Motorola, General Electric, and the various Japanese brands, countless 
CB radios, and of course with my homebuilt HF and VHF radios and 
amplifiers, such as the little 5-watt 40m QRP transceiver, or the 80W FM 
stereo broadcast transmitter, both published on my web site.

It's no rocket science.

With tube amps and their high Q networks, starting at full power and 
with the network seriously out of tune can melt the tube plate, or the 
envelope. I have seen both. In transistor amps instead, with their much 
lower Q networks, they are never totally out of tune. It's really only a 
touching-up what one has to do! So it's much harder to dramatically 
exceed the normal dissipation level.

 > This
> would be one application, for a  6:1   or  10: 1
> vernier drive, with well calibrated  large diameter  skirts.

Absolutely not! The loaded Q in these networks is far lower than what's 
used in tube amps, and for that reason the tuning is very broad and 
quite uncritical.

> And will
> it handle higher swrs ?   Or do we have to go through the
> rigmarole of a manual or auto tuner ?  

All such tuning networks do handle somewhat elevated SWR. How much, 
depends mostly on their Q. Since the minimum required Q is typically 
much lower than in tube circuits, most designers would indeed use a low 
Q to reduce size, cost, touchiness, increase bandwidth, etc, and this 
will allow only a limited range of load mismatch to be tuned out. But 
there is absolutely nothing preventing the designer from deliberately 
using a higher Q in the matching network, achieving a behavior closer to 
that of a tube amp: Ability to match crummier loads, along with more 
sensitive tuning (here your vernier might come in!), and more loss in 
the tank, requiring a big coil to keep loss at bay.

But then the typical autotuner anyway allows only a 3:1 SWR range or so. 
That level of SWR doesn't require a very large Q in the matching network.

Think of it like this: Instead of having low voltage transistors 
followed by a broadband transformer, a set of low pass filters, and then 
an autotuner,   you would have high voltage transistors and then just 
the autotuner, doing away with all the other stuff in between! But then 
you need an autotuner having good enough harmonic suppression. The most 
commonly used T or L networks don't, but a slightly more complex tuning 
network does. This slightly more complex autotuner (or manual tuner) 
goes directly between the high voltage transistors and the antenna, 
allowing to work into an SWR of 3:1 or so.

Tuning can be done at reduced supply voltage and limited supply current, 
so that the transistors cannot be thermally overstressed. Once tuning is 
complete, the full supply voltage is applied. With an EER system, this 
is trivially simple to do, since the variable power supply is already 
present. Instead a conventional class AB linear system would need the 
supply voltage switching to be added, but that's quite simple too.

When the transistors are selected according to conventional wisdom, that 
is, that their dissipation rating should be well over twice the maximum 
output power, I would even feel comfortable running such an amp without 
further protections (but paying attention while operating). Instead in a 
high efficiency amp that has transistors rated for a dissipation roughly 
similar to, or even way lower than the output, additional protection 
circuitry would be required to keep the transistors alive in case of 
mistuning, a bird sitting on the dipole tip, a coax connector coming 
loose, or whatever.

Manfred

========================
Visit my hobby homepage!
http://ludens.cl
========================