Problems with microstrip Broadband and selective amplifiers

Hello,
I'm doing a university project and it is my first time to work on amplifiers (or any hardware at all) and I have run at some problem I cannot solve.
My assignment is to build a broadband amplifier, using a gain block circuit and a selective amplifier using a transistor.

I designed the circuits in and simulated them in Qucs to get the wanted results.

However when I measure, between the results and simulation, there is quite a difference:
- Broadband amplifiers gain is both lower than simulated and declining in the part it should be level
- Sel Amp has similar shape to the simulated, but the gain at design frequency is 11db vs 17db(simulation)

Somethings that could cause small differences:
-the substrate could be with er 4.4+/- 0.1
-the amplifier leads are not connected to the center of of the microstrips
-the matching stubs in the sel amp may be not at the exact place as calculated


Can someone suggest a possible explanation for big differences, because I have run out of ideas.

Thanks in advance

Hi N3x,

The stubs not being in the right place could definitely be an issue. What is the difference between the ideal simulation and reality?

With your amplifier fired up, finger the ground plane. A stick with copper tape on it can also be used. If the signal jumps around, the ground vias are a problem. Also try your DC bias lines too out where the power supply connect. I see solder around a couple vias. With a 50 ohm term on the input, do you measure any output power, i.e. is it oscillating?

I'm hoping you are not trying to get 10MHz to 8GHz out of your amp. Non-ideal effects of ground vias, ground around your SMA connectors non-ideal capacitors, etc will not allow 8GHz.

Hi ge,

I think the stubs are off by half of the width of the strip line (~1.4mm) because I start measuring their length form the top of the line, not from the middle.
However today I'm going to test a new board with the stubs positioned the "right" way.

I'm going to try the test later, when I get to the lab.

For the broadband amp I wanted to get constant gain for 100MHz to 4(5)GHz. I hoped to achieve results that would be close to the simulation, because I tried to add most of the components in a "as real" way as possible.

Thank you for the tips so far :)

Sometimes differences between simulations and measurements take place because maybe there are EM coupling in your structure.

If you send me the screenshot of the layout maybe i can see if there are some parts critical to be simulated for example by momentum or sonnet. See also
www.uwaverf.altervista.org/forum