[Amps] MOSFET amp filtering - was: auto-tune

[Manfred Mornhinweg]

Steve,

> If the topic is filtering and maintaining a good match - then why can
> we not transform up to the input impedance of an inexpensive
> multiband matching system?

The problem is imperfect transformers. Introducing delays, and having
poor coupling. We don't even need "high" impedance. 50 ohm would be fine.

> What would be the input impedance of a multiband matching and
> filtering system that was half the price and complexity of the
> currently used per-band filter?

I would say, it's convenient to design it for anything between about 30
and 100 ohm input impedance. This would even allow to relay-switch
complete per-band matching/filtering networks. Each of them might use
just three parts. Or more, if more bandwidth is wanted.

But this ends up very close to the usual switched low pass filter approach!

> Would the loss of that transformer be a magnitude worse than a 50ohm
> output transformer?

It would be a 50 ohm transformer! :-)   Designers came up with this in
the early 1970s...

Bill,

> Harmonic filters do need to be switched in and out depending on the 
> band, but they don't need "tuning".

I can conceive one truly broadband approach, that would need no badn
switching of any kind: A class D amplifier with pulse width modulation,
having a carrier frequency around 200MHz or so. It would need a single
but good 30MHz low pass filter at the output, used on all frequencies
from zero to 30MHz.

The little problem with this approach is the lack of suitable power
devices...  But maybe in the near future we will get them!

Cathy,

> Tuning is a fact of life, 

You are right, and that brief sentence has almost philosophical weight!

I try to set up my antennas so that I can have an SWR below 1.5:1 over 
all band segments I use, but still occasionally I need to move out of 
those segments, or have ice on the antennas, and then I need a tuner.

Which makes an SS amp with tunable output network an attractive option. 
The tuning can be automatic or manual.

Instead of the current approach of using a broadband transformer to step 
up the impedance, then a switchable low pass filter to reduce harmonics, 
and then a tuner to adapt to practical antennas, it should certainly be 
less expensive and more efficient to use a single tunable tank network 
that does all three things in one. But to make it band-switchable, we 
need high voltage transistors that can run at a few hundred volts. 
That's the single problem. Unless somebody can think of a way to 
band-switch at a 3 ohm impedance level. I can't.

Manfred


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