[hpsdr] Norton Amplifiers and Noise Performance

[Chris Bartram]

>     Can anyone comment on what attributes the Makhinson (N6NWP) circuit
> (which is essentially a pair of Norton amps in push-pull) brings to the
> table that would make it preferable over the Norton amp in this
> application.

The main advantage of a push-pull amplifier is that you'll get vastly improved 
second-order inter-mod. performance. The third-orders will improve by about 
3dB as well.

I have to be a little cautious about what I due to professional confidences. 

It's certainly possible to get around +50dBm IPI3 from a push-pull Norton 
amplifier, and the BFG591 which has been mentioned, is capable of approaching 
this in the right circuit. There are better devices... One of the problems 
with the 'traditional' Norton amplifier is that the load line for the active 
devices doesn't coincide with the conditions for a 50ohm match at the sort of 
currents and voltages necessary for the highest IM performance. There's also 
the problem that the transistor bias current flows through the feedback 
transformer, which, of course, causes magnetisation of the core, and limits 
the dynamic range. That usually means that the ferrites specified in a 
'traditional' Norton amp. have to be larger than would be necessary from just 
a signal handling viewpoint. 

There are ways around both of these problems, but I can't discuss them 
further...

Incidentally, measuring +50dBm plus IPI3 is not a trivial matter...

The noise figure required of a receiver is ultimately limited by the noise 
which comes down the antenna cable. At HF and VHF this is a well known 
quantity, and there are graphs in most of the standard text books from which 
it can be calculated. If you work on the basis that the receiver system will 
not degrade the available S/N by (say) 1dB, then receiver noise figures of 
10dB at 10m and 6dB on 6m are more than adequate. I know that there's a 
market for (allegedly) 0.3dB NF pre-amps for 6m, but any system performance 
improvement they bring is more likely to be psychological than real...

Vy 73

Chris
GW4DGU

-- 
Chris Bartram
Christopher Bartram RF Design
Blaenffos
Pen-y-Bont
CARMARTHEN
SA33 6QG

phone:  +44 1994 484650
mobile:	+44 7968 193555
email:  chris at chris-bartram.co.uk
URL:	www.christopherbartramrfdesign.com