6166	2011-04-23 09:28:25	richj_focus	RF isolation amplifier
Message	Date	From	Subject
Hello: I am thinking about what would be the best isolation amplifier between an antenna and the first mixer, or between two filters, or in general just the most most isolation between input and output of an amplifier. I am thinking of a grounded gate FET, or common base BJT, or maybe a cascode. What is the experience of the group please. cheers, rich
6167	2011-04-23 12:09:47	Chris Trask	Re: RF isolation amplifier
> > I am thinking about what would be the best isolation amplifier between > an antenna and the first mixer, or between two filters, or in general > just the most most isolation between input and output of an amplifier. > > I am thinking of a grounded gate FET, > Mmmmmm... Maybe. > > or common base BJT, > Good choice. > > or maybe a cascode. > We have a winner. > > What is the experience of the group please. > There's a super spiffy way of getting 30-40dB of isolation with a common-base amplifier. It requires a 1:N:N transformer, a capacitor, and a resistor. It's described in only one textbook, and I've tried it with PSpice and one prototype. It gives you wideband neutralization and the reverse transmission with the VNA is in the noise floor. Chris Trask N7ZWY / WDX3HLB Senior Member IEEE http://www.home.earthlink.net/~christrask/
6169	2011-04-23 12:33:40	victor	Re: RF isolation amplifier
If you want a low noise and high dynamic range amplifier (high IP3) then common gate, comm
6170	2011-04-23 13:32:27	Chris Trask	Re: RF isolation amplifier
> > If you want a low noise and high dynamic range amplifier (high IP3) > then common gate, common base and cascode amplifiers compromise either > noise or dynamic range. > I designed such a low noise high dynamic range amplifier more than 20 > years ago and published it in RF Design magazine. > I have uploaded the article in the group Files, under my call sign - 4Z4ME. > Yep, seen that many times. I have it filed under "Nifty Amplifiers". I fixed both of those common-base amplifiers with augmentation. With tandem augmentation you get rid of the bulk of the 1/f noise, which is what causes the greater part of NF by way of modulating the bias which then modulates the signal(s). Chris Trask N7ZWY / WDX3HLB Senior Member IEEE http://www.home.earthlink.net/~christrask/
6171	2011-04-23 13:45:58	victor	Re: RF isolation amplifier
Noise figure of common base/gate is limited to a not very low value when the input impedance of the transistor is used directly to define the amplifier input impedance. The use of lossless feedback as in your amplifiers or Norton's solves this problem but not the reverse isolati
6172	2011-04-23 13:49:27	Chris Trask	Re: RF isolation amplifier
> > Noise figure of common base/gate is limited to a not very low value > when the input impedance of the transistor is used directly to define > the amplifier input impedance. The use of lossless feedback as in your > amplifiers or Norton's solves this problem but not the reverse isolation > problem. It is not easy to design an amplifier that fulfils all this > requirements simultaneously: 50 Ohm input/output impedance, Low Noise, > High linearity and high reverse isolation. > That's correct. In any feedback amplifier the reverse isolation is the inverse of the forward gain. Chris Trask N7ZWY / WDX3HLB Senior Member IEEE http://www.home.earthlink.net/~christrask/
6173	2011-04-23 13:52:03	victor	Re: RF isolation amplifier
In my design the reverse isolati
6174	2011-04-23 14:45:46	Chris Trask	Re: RF isolation amplifier
> > In my design the reverse isolation is much higher than the inverse of its gain. > As I recall. I had looed at is wondering if a CB stage could be used, but you would have to add a third transformer. Chris Trask N7ZWY / WDX3HLB Senior Member IEEE http://www.home.earthlink.net/~christrask/
6175	2011-04-23 21:15:27	Ashhar Farhan	Re: RF isolation amplifier
you might want to read this : http://w7zoi.net/bidiamp/bidiamp.html <http://w7zoi.net/bidiamp/bidiamp.html>though it was developed for bidirectional usage, there is no reason to not use unidirectionally, inter-stage. - farhan
6176	2011-04-24 04:25:34	Tim	Re: RF isolation amplifier
I'd like to jump in and point out that these requirements happen to overlap quite well with the classic "electronic T/R switch" wish list even though actual amplification isn't actually a requirement for a T/R switch. (And in fact the T/R switches that forgo any amplification and live with small to moderate loss, seem to be my favorite solution at least this weekend!) Tim N3QE
6178	2011-04-24 08:09:23	Chris	Re: RF isolation amplifier
> > > > > > ... It is not easy to design an amplifier that fulfils > > > all this requirements simultaneously: 50 Ohm input/output > > > impedance, Low Noise, High linearity and high reverse > > > isolation. > > > > > > > That's correct. In any feedback amplifier the reverse > > isolation is the inverse of the forward gain. > > > In my design the reverse isolation is much higher than the > inverse of its gain. > I dug out my decade-old notes about your amplifier, and I had trouble back then resolving an issue with the load impedance at the summation port of T1, which should be 25 ohms if N=1. I also had difficulty with the dissipation of power in the isolation resistor across the two input ports of T1 that results from the two inputs being of opposite phase. I spent the past couple of days revisiting this, and tried a slightly different approach where the amplifier is now a series/shunt feedback type with a 1:1 power splitter added between the transistor collect, the feedback path, and the output. By forcing the output impedance of the transistor collector to be the optimal 25 ohms and adjusting the emitter series feedback resistor so as to double the transistor gain, I came away with a unity gain buffer amplifier having 40dB of reverse isolati
6180	2011-04-24 10:40:20	richj_focus	Re: RF isolation amplifier
Chris, can you show the schematic of your prefered design for this subject? Thanks much for all the input group. rich