Different Input/Output matching
I need help with this "problem".
I have device A, which has 50 ohm output port (900 MHz). I also have antenna with 50 ohm input. So There is no problem.
Now, I wanna make a new device "blackbox". Input of the blackbox should be matched to the 50 ohms (so there is no problem with reflected power into the device A), but output port of the black box should have different impedance (20+10j). So basically, S11 of the blackbox should be -30 db (matched) and the S22 of the blackbox should be - 7 dB.
I've basic LC matching network. But (as you know) I can do S22 = -7 dB, but S11 also equals to - 7 dB.
Do you have any ideas how I can do this black box? So there is no reflected power into the device A and also there is mismatched output for the Antenna?
Thank you
I don't "know" this, neither the setup seems to make sense. Why would you measure S22 on the 20+j10 port with 50 ohm reference impedance? You may measure the complex impedance and calculate S22 for the target impedance.
Correct me If I'm wrong, but if I have input 50 ohms, then LC network (using for changing the impedance), and then I connect antenna with 50 ohms, it will change both S11 and S22 the same way. What I do is actually mismatching the device A port, so connected antenna is will not be perfectly matched. But I don't want any reflected power into the device A port.
I hope I've written my thoughts correctly...
I need to clarify, how can I have different S11 and S22 with using same impedance at input and output.
Measuring of the S22 with reference 50 ohm reference impedance is representing, that the antenna "see" mismatched input (cause antenna is also 50 ohm9
To be honest, I don't understand your problem and what is your purpose. Maybe you could make a simple diagram to show what you want to do and what is the problem.
Simple simultaneous matching.
http://literature.cdn.keysight.com/l..._gamma1().html
http://literature.cdn.keysight.com/l..._gamma2().html
Yes, I also think that some "Pi" network will be enough. But I can find calculator, or procedure for design that network.
In theory the black box does exist. But what is the point to design such a mis-matching network?
If you would like the network to be lossless and composed by LC, it is not possible. From your requirement of 50 OHM input impedance, we have S11 = 0. Since it is lossless network, S21 = 1. Since the LC network is reciprocal, S12 = S21 = 1.
Again use the energy conservation formula, we have S22 = 0. The output impedance must be 50 OHM.
Such a network must be lossy, or not reciprocal.
mis-matching network should simulate, that the device A does not have matched output.
There will not be problem with some lossy or that the circuit won't be reciprocal.
Do you have any ideas? I was thinking about band pass filter - input impedance matched, output will be mis-matched.
After seeing the post #6 diagram, I really don't understand what you want to achieve.
1. You don't explain why you design intentional mismatching.
2. As stated by others, if you have the mismatch, you'll see it in S11 as long as the matching "black box" is lossless and reciprocal.
3. The technical solution to isolate a mismatched antenna from the source is a circulator.
As FvM said, use circulator to absorb the reflection from antenna is common way to separate TX/RX channel. It has been used in radar system for many many years.
if you just worry about the output impedance of A is not 50 OHM, a matching network will solve the problem. If your band is narrow, two components are sufficient.
Simply insert attenuator at left side.
https://www.edaboard.com/showthread.php?384685 is never useful for your purpose, since antenna impedance is 50ohm.