0
\$\begingroup\$

I'm designing the frontend of a receiver and I want the output impedance to be 50Ω so that I can get as much power into the mixer. I tried placing a bandpass filter (image filter for 88-108MHz) in between the amplifier and the transformer primary side, but it didn't workout (signal became very noisy). The transformer has a 1:1 ratio and is used to isolate the amplifier from the mixer. I don't know where to place the filter and impedance matching network for 50Ω output impedance, passing 88-108MHz. Any help would be greatly appreciated.

Schematic : enter image description here

Edit : enter image description here

\$\endgroup\$
8
  • 1
    \$\begingroup\$ Skipping the transformer/load, this stage obviously is not a power stage. Yet, on the right is a 50 Ohm termination, which tells me that this should be a power stage. That juxtaposition bothers me right away. And when you say it "became noisy" this says there was a before-this circuit. What is that circuit? Is it just this one without the 1:1 transformer and 50 Ohm load? Finally, noise matching would be critical in a non-power stage. What analysis did you do to match your input load to the conjugate noise impedance of the source (unshown?) I'm not good at RF and I'm out of questions. \$\endgroup\$
    – periblepsis
    Commented Apr 28 at 4:57
  • \$\begingroup\$ Well, it might be a power stage accidentally, if the transistor's buit-in \$C_{cb}\$ and the parasitic \$L_e\$ in its package are providing the right voltage and current feedback. (Some transistors are popular for that property...) But most likely it is not (so you are right). \$\endgroup\$
    – Jos Bergervoet
    Commented Apr 28 at 7:22
  • \$\begingroup\$ Well, 2SC3838K headlines a "11V, 50mA, 3.2GHz" rating, so the 12V supply is a bit marginal, but there's plenty of room to reduce the resistor values. Have you considered a matching network, other transformer ratio, loosely-coupled (double tuned) transformer, etc.? Please show the proceeding stage, and perhaps input network up to the antenna connection, as well. \$\endgroup\$
    – Tim Williams
    Commented Apr 28 at 8:18
  • \$\begingroup\$ A common base amp with a 3GHz transistor, without any measures against instability, is bold ;-). The input impedance is about 25 Ohm. The output impedance is high. There is no match unless you force the output to 50 Ohm with an appropriate collector shunt resistor and corresponding transformer ratio. Additionally, the current gain of the transistor is 1 so you hardly get any useful gain out of the amp. Maybe consider switching to a common emitter stage with shunt and series feedback for 50 Ohm In/Out. \$\endgroup\$
    – Raonoke
    Commented Apr 28 at 9:57
  • \$\begingroup\$ The circuit before this was that but the transformer removed and RL is 5.2k. I used the common-base amp configuration since it does not suffer from Miller Effect, which is crucial at high frequencies. Although there is no current gain, there is lots of voltage gain (Rc / re). I'm trying to make a matching network that will pass 88-108MHz and matches 5.2k to 50. The full circuit is shown in the edit. I have not currently chosen which mixer I'm going to use, but looking at the options, it seems they all have 50 ohms input impedance. \$\endgroup\$
    – TheButterMineCutter
    Commented Apr 28 at 17:24

1 Answer 1

Reset to default
4
+50
\$\begingroup\$

Take a look at the following two examples of common emitter, 50 Ohm I/O preamplifiers. The first one, with current and voltage feedback, is broadband, has medium gain and doesn't need much power.

enter image description here

enter image description here

The second example has high gain, uses inductive emitter degeneration for low noise and is I/O matched to 50 Ohm over the entire 88-108 MHz band.

enter image description here

enter image description here

\$\endgroup\$
3
  • \$\begingroup\$ Thank you. Where did you learn to design amplifiers like that? I've read "Art of Electronics" and they don't seem to teach it with inductors. \$\endgroup\$
    – TheButterMineCutter
    Commented May 3 at 3:54
  • 1
    \$\begingroup\$ @TheButterMineCutter if you use DC feedback from the collector to the base you can more readily use an inductor for AC emitter degeneration. If you used a fixed bias on the base it becomes infeasible to use inductor emitter degeneration. \$\endgroup\$
    – Andy aka
    Commented May 4 at 13:10
  • 1
    \$\begingroup\$ @TheButterMineCutter * AoE can not cover all types of amplifiers out there. More info about the principle and design of LNAs with emitter degeneration can be found for example in "Thomas H. Lee, Planar Microwave Engineering - A Practical Guide to Theory, Measurements and Circuits, 2004, Cambridge University Press" p.451. \$\endgroup\$
    – Raonoke
    Commented May 11 at 6:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.