Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
and u can put the filter after the lna this will reduce the NF of the filter
but this is hard if it is integrated reciver coz u will need the filter of chip
if you can't afford input filter and you have SSB receiver, then after LNA use the image rejection mixer (IRM). The use of IRM should decrease the noise figure by 3dB according to theory (image noise is converted to the same IF so the noise power is two times the power comming from the single band).
It will be slightly less in practice depending on image rejection achieved and bandwidth of the rejection.
Strarting from your question and as it is formulated, It seam that you already have a good LNA but it don't mask enough the NF of the following stages.
So the solution is reducing the NF of the chain beginning from the mixer.
Many tecnically experts peoples replied before me.
I'll give you a solution from an economical point of view.
Usually a common double balanced mixer has a NF >= 7 dB.
Usually the IF amplifier has a NF in the 5 dB range.
For my knowledge, the cost of a mixere is higher than the IF amplifiers.
Try to replace the IF amplifier with a Low Noise IF Amplifier. Watkins Johnson, for example, produce low noise, cheaper, leaded MMIC Amplifiers.
Note that 2 dB reduction on IF ampli's NF and 2 dB reduction on Mixer's NF is the same!!
In my idea,to lower the NF of the system,you must select the low insert loss of the select filter and low NF of the first stage LNA.But low NF of the LNA often doesn't have low reflect cofiction,so many people to use balance type. and this being possible,because the 3DB coupler now have lower loss such as from aneren.
In practice, gain of LNA could not be set arbitary high without oscillation ! Also too high a gain will affect your system power handling cap (IP3, P1dB will be lowered).
You have really to strike a balance in your system design. Sometimes, you can switch off LNA gain in order to improve P1dB , using switched LNA design
I have some experience with LNA design. In is often optimized for gain/noise trade off, input VSWR is of secondary importance. However, typically large VSWR (>3) that goes hand in hand with usual gain/noise chosen can missmatch an input filter or antenna and this is not acceptable in some applications, so one have to redisign the LNA and sacrifice some NF to stay tuned.
High input VSWR can trick an unexperienced designer in measuring G/NF by introducing large uncertanity. The essence of the phenomena is that noise power injected from the noise source is reflected at the input, thus neither entering the system nor appearing at the output and the instrument is fooled making the low NF and high G reading. @gilent has the built-in uncertaniny calculator in their new instruments. You can find more about this on their site.
Except everything else that have been mentioned, make sure you don't have a lossy cable connecting your antenna to the first stage (filter or LNA). A look at the Friis Formula will show you what you need to do and why.
Here are some tips. You may not be able to use all of them due to system constraints.
1) Lowest noise figure does not usually happen at the maximum gain. It at slightly reduced gain.
2) Use good dielectric and pay due importance to thickness/bends/width.
3) Use very high Q capacitors.
4) Filter thr power supply.
5) minimize any insertion losses.
BRM
PS: jupitorcuu: why do you feel CDMA requires lesser NF. If you increase the NF then more power is needed resulting in increased interference to other users lowering capacity.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
