LNA design

[Leiser]

Hi all,

I have a question about lna design. I´m designing a 1.4GHz LNA. In order to get the best noise performance I was thinking placing the band-pass filter at the ouput of the amplifier.

Does anyone know if this makes sense?

Thanks in advance!

 

[bogeyman]

Makes sense if you want the lowest possible noise, it won't work in real life though.
If/when there is a large out of band signal in your operating site (an airport L band radar @ 20 miles or local cellphone mast) the amp will saturate and not do much by way of amplifying the frequency of interest, if your filter is a good one there will be less than 1 dB of extra noise if the filter is before the amp.
If low noise is important you will need to spend a wee bit of extra cash on the filter and amplifier to get a good system.

 

[logbook]

... hence this is a system problem. How much out of band signal do you expect? If you can pick up strong out of band signals which will then saturate your amplifier, then the output is not the right place for the filter. The input is almost never the right place for a bandpass filter in an LNA system. You might for example split the amplifier and put the bandpass filter after the first 10dB of gain.

1dB insertion loss in a filter in front of the LNA increases the noise figure of the LNA by 1dB. Not good.

 

[Leiser]

Hi,

Thanks for your comments. There is something I don´t fully understand, if I pick up out of band signals, how can they saturate my amplifier? is it because harmonics and intermodulation products of those signals can fall in the desire bandwidth?

Yes, we are thinking splitting the amplifier, however the design is quite finished and now where are trying to assess the possible problems of the design, putting the filter at the output.

Thanks.

 

[logbook]

Well if the amplifier were acceptable as the bandpass filter you presumably wouldn't put a bandpass filter on its output. I therefore assume that the amplifier has at least slightly more bandwidth than the bandpass filter.

You are right that the harmonics of a lower out of band signal may cause problems, but then being in band they would pass through the whole system anyway.

Depending on the strength of the out of band signal, it may cause in-band intermodulation products in the amplifier, but it would have to be pretty strong. I wouldn't like to quantify that effect but youo could test for it if you knew what interference you were expecting. A mobile phone is always a good bet for interference. Try several different phone bands.