Dear, all.
Let's imagine the situation that we have RF receiver working with passive antenna. LNA of this receiver has acceptable parameters(NF, S12, S21, S22, and so on), except poor S11. Suppose there are no filters between LNA and passive antenna. What would be the drawbacks of such curcuit from the  point of view of overall receiver performance?

One thing thing is that part of input power would reflect back, re-radiate by antenna and may corrupt input signal for other receivers working near the one we observe. But i doubt if this phenomenon is of big importance.
Any comments?
Thanks.

The way you pose the problem is that you have a LNA designed to have low NF with a somewhat poor S11 referenced to 50 ohms. Now you are going to interface it with an antenna, and you want to know the impact.

An implicit assumption is that the antenna impedance is 50 ohms. If it is not (and it doesn't have to be) then you have to figure out what the NF performance is with the antenna impedance (could be better or worse). If the impedance is 50ohms then the performance of the LNA should be about the same as you simulate with a 50ohm port.


It is also possible (although I'm not an antenna expert) that the radiation efficiency of the antenna is impacted by the impedance presented by the LNA. If this happens then some of the EM energy incident to the antenna will be scattered rather than absorbed. The net effect would be the receiver front-end with the antenna would be "less sensitive" that it would be if the radiation efficiency of the antenna were unchanged. The problem here is that you cannot predict this effect without some sort of working model of the antenna. A thermal 50 ohm port doesn't capture this at all.

So, I think you need to dive into the antenna question a bit to find a good answer to your original question.

On second thought ...
It was a poor choice on my part to imply the radiation efficiency of the antenna itself is changed by the matching. But certainly the overall efficiency could be degraded.

Pashtet,

I think what RFICDUDE is implying by overall efficiency, he is implying that consider the LNA and antenna as a system the optimum as a system is not achieved.  
The antenna will have gain vs an isotropic antenna (which uniformly collects energy in all directions).  Antennas have this gain, which is specified as dBi. If you lose a 1dBi because of poor matching but your noise figure improves by .5 dB then your "overall efficiency" is degraded.

http://rfcooltools.com

Pashtet,

Loss before the LNA is dB for dB equivalent to a noise figure hit, by not being matched the system might be loosing the potential to be more optimum.  I say might because an example can be brought up that implies the contrary, but usually that is the case.  But a more subtle issue in the example you provide is what is the noise figure after the LNA can be exaggerated further,  If that noise figure is around 20dB then a low noise figure of the LNA and a gain of 20dB is roughly an additional 3dB noise penalty when referred to the input of the LNA  so a 2dB NF will equate to approximately 5dB input referred.  Where as a 2.5 dB NF and a gain of 26dB will be around 3.5dB NF.

http://rfcooltools.com

Pashtet wrote on May 16^th, 2011, 5:30am:


for your question, I think there are two case:
1. in your case, your receiver is a wireless receiver. then, the reflect power what you say is another description of not max power transmission. but that not max power transition does not affect others, because your receiver antenna radiation power is low, others can't receive it
2. if in cable transmission, LNA input impedance mismatch will affect cable transmission character impedance. thus it will affect transmission efficiency, that means the transmitter should have higher driver capability.

so , in the first case, you may get a NF without matching to 50ohm. however, in the second case, you may loose antenna gain to get LNA better, the total NF way not be good.

the above is just my thinking, just for reference, I am not quite sure about it