Quoted:


Yes, although its been 15 years since I designed amps for a living, and 8 since I taught it at the university.  I'd have to dig out a textbook, haven't done hardcore matching on a Smith chart in 8 years.  There used to be a student version of Ansoft designer for free download (I used that in class, the students seemed to like it).  If you enter the S-parameter file, you can do linear matching networks to check the design.  You need non-linear model data (Raytheon-Statz, Curtis Quadratic, or Curtis Cubic) to check non-linear parameters like 3rd order intercept, 1dB gain compression, etc.

And those aren't FETs, they look like PHEMTs.  If you are going to run them linear you may have problems.  Microwave PHEMTS have a bad tendency to suffer from power slump when you bias them for linear operation.  We couldn't take a family of curves without sag in the curves due to junction heating.  They worked much better in pulsed applications, we even used pulse techniques to get a decent family of curves.  Maybe the low frequency ones are better, as the channel dimensions are probably bigger.

Quoted:




No time like the present to ask questions and see if you have an interest...

Quoted:
Alright here goes.... I went through college as a computer engineer with much work on the electrical engineering side of things but professionally I'm a computer security R&D guy. Circuits aren't a mystery to me but I'm not great at them any more.

My understanding of your amplifier thusly is.........

You have an input signal/s of various frequencies of whatever your antenna is picking up. These are AC/sine waves of very minute voltages? What voltages are we talking about here?


<span style="color: blue;">Yes, these are 'RF signals, just the same as your VHF rig would pick up. The level of the voltages is in the millionths of a volt, we call this microvolts. Or 'uv'. For the level of signals we're testing with, or weak signals coming into your antenna.


So with all these various frequencies inbound do you try to filter them out? And you are using an inductor/capacitor network or traces on the board to do so? When you "filter out" a signal where does it go? Does it just get grounded out or converted into DC that you can block with a capacitor?

<span style="color: blue;">Good questions!

All properly designed receivers have 'input tuned circuits' that filter out [reject] the zillions of extraneous signals coming in off the antenna.  Doesn't matter at what frequency.

Where do they go, if you put your hands over your ears, to block out sounds, the same place.  

The signals never get thru the filter [as much], but unlike sediment in a water filter, the miniscule amount of energy in those 'blocked' signals, is either reflected back into space, to be dissipated as an almost immeasurable amount of heat energy, or due to the Q factor or resistive losses in your filter, the unwanted signals are dissipated as heat, there.

As far as a capacitor, at the frequencies we're talking about, tuned circuits are usually comprised of a network of capacitiors and inductors.  Ergo, why you see coils and caps on the RF circuit boards.

For UHF and higher frequency signals, there are too many issues with lumped constant components, caps, and coils, so usually they are implemented as traces on the printed circuit board, called 'microstrip', having generally well defined size and characteristics..

Okay... At this point the unwanted frequencies are gone and we are down to the ones we want. This is fed into the gate of the FET? Now the power going into the source FET is from a DC source? Whats the voltage going into this FET? And then the drain from the FET is fed into the radio? I think I'm seeing some circuit designs that have resistors or pots on the source and drain of the FET which would allow you to control voltages some. What voltages are we talking about here?


<span style="color: blue;">Right!

On the preamp board I laid out, there is a coil and an input capacitor, variable, to allow for some tuning, but there are only two 'poles' [hell, I don't know how many...]  

...and the input filtering is very broad.

Correct,  the signals next go to the gate.

The amplifier is 'biased' with a DC voltage, provided in this case by the 5 volt regulator. From the spec sheets for the FET, we see that the design specs call for about 3 volts across the drain and source, or in a tube, the plate to cathode.

The drain is equivalent to the plate of a tube and there is a circuit to roughly match the FET drain impedence to the 50 ohm coax. In the pix preamp, it's the black thing to the left of the FET. In my case cause I have bags of them, it's a small binocular ferrite bead with about 6 turns of bifilar wire wrapped thru the eyes to step the high impedence of the FET to ~50 ohms.

It isn't critical and doesn't matter if there is much loss at this point because the noise figure is already set by the input network and the characteristics of the FET.

As far as pots on various circuits, they are likely there allowing for adjustments to FET bias points.

As mentioned above, I put an ultimately 5k pot, in the source lead to adjust the bias current to the 100's of micro-amperes to see what little power could be drawn and still amplify with a good noise figure.

Does the FET invert the frequency 180 degrees? Does that even matter? Is there or what happens if the gain is too great? Does the amplified signal just get clipped?


<span style="color: blue;">
Yes it does and no is doesn't make a rat's behind's worth of a difference, except to folks in academia.  

If the gain is too great, the receiver that's hooked to the preamp suffers the same fate as if EMP hits it, the front panel knobs of the receiver get blown off into the operator's belly.



Oh wait, that's not right...  

It isn't going to happen in this instance, so we don't have to worry abt it.

Low level signals that are too large for the dynamic range of the following stages, do get clipped.

Next questions....