How to improve linearity without losing the gain?

i am facing a big trouble with gain and p1dB of my active mixer.
I need a gain of 20dB and a p1dB of -12dBm for a low voltage application.
I am not sure whether it is possible or not.

Is there any technique to improve linearity without losing the gain?

Thanks..

if no specification on current, you can parallel the transistor to increase the current.

helo , can u explain more

try to think in other mixer architectures , and see which one can best fit with ur spec's

don't stick in one topology , scan the topologies and then decide

khouly

there is a constraint on current also.
iam at the max current limit now.

Ok, lets say i am doing a mixer and i have reached some spec on Noise figure and Gain ,and i have a spare current budget can i spent this current to increase the IP3 without decreasing the gain or the NF
\thnx

this is similar as PA design, you can reduce the load impendene to increase the current, then the total output power is increased.

so it is important to match the circuits.

hi,

your question is difficult to answer without seeing your circuit topology. what's your supply voltage?

Gain is different , compression point is different thing.
If you P1dB CP is -12dBm ( I assume output )and you gain is 20dB, that means when you apply -32dBm power, you mixer will be saturated...

Don't confuse the terms..

gain and compression point are usually traded

I have an amplifier that has P1dB CP is 44dBm but gain is 6 dB ..

or another amplifier P1dB CP is -22dBm but gain 35 dB..

So where is the trade ?

Dude u stated the trade?!
the lower gain with higher CP and the higer gain with lower CP?!!

Dear BigBoss,
can you get 35dB gain with 44dBm p1dB?
i feel it is very difficult...

No you can't..
There are always limitations and constraints in amplifier design but that DOES NOT mean that there should be cross correlation between CP and gain...

Theoritically and practically they are independent..

If I intend to implement an amplifier having a gain 5 dB and P1dB CP -25dBm, I can do that.Nothing can stop me to make it..

CP is ONLY and ONLY depended on NONLINEARITY behaviour of the active components nothing else..

So, it should not be confused with gain and CP terms..

dear BigBoss,
linearity is always traded with Gain , if u are able to do bad design then this doesn't prove anything, the idea is simple u cannot get high gain with high linearity and hence, u trade between them and thats it , if u get both low gain and low linearity so what it is just a bad design that doesn't need a trade-off

20 dB gain and -12 dB seems to be feasible, because I even implemented an amplifier with 25 dB gain and -13 dBm P1dB. but this totally depends on your supply voltage and circuit topology. the linearity depends mainly on your voltage headroom of the mixer load.

how about a 1.1V supply and with 20dB gain , what should be a good IP3 or P1dB?

dear estradasphere,
my design is for a low voltage application and it is a folded mixer.
the supply is 1.2v.
freq range is 2.4G.

i would like to know what is the best gain i can get with a good p1dB if you have any idea about this.

well, to be honest with a supply voltage voltage of 1.2 V, it may be quite difficult to achieve 20 dB gain and -12 dBm P1dB. i think, the folded structure is not proper for low voltage applications, as it consumes too much voltage headroom. you can easily calculate or simulate the maximum achievable gain, because i'm not familiar with the technology you're using. passive mixer is more suitable for low voltage applications, but you should amplify the signal in front of the passive mixer to get the required gain.