MMIC and Noise figure

I have phase and amplitude detector from -60 to 0dBm and DC to 2Ghz range in test instrument and I want to increase amplitude bandwith.

New Micro Circuits LEE-29 has good linearity in this range and Noise Figure 5.5.
Can I use this MMIC, what about chain of this MMIC's because NF=5.5?
How much must be NF for -100dBm sensitivity, any rule of dipedence sensitivity from noise figure?
Instrument has calibration so, first in chain can be some MMIC with smallest NF(for example ERA 1 has NF=4.3),
what I can expect with this NF?

Noise figure of a chain is calculated regarding to Friis formula e.g...

F(2nd stage)-1 F3(3rd stage )-1
F=F1+--------------------------------------- + -------------------+.......
GA1( Available gain of 1st stage) GA1xGA2

F: Noise Factor ( numerical value NOT dB )
GA: Available Gain ( numerical Value NOT dB )

MDS=kTB+3dB+Ftot+Gtot

k: Boltzman coefficient
T: Ambient temperature
B: Bandwith

MDS= Minimum Detectable Signal

For more information look at the attached file. read carefully and you'll understand.In fact sensitivity depends on many factors such as modulation,signal processing etc.
In Additional to that , wit MAR series amplifiers are not intended for Low Noise applications. Instead of , you should use discrete transistors
(cheaper ) or very low noise amplifiers from CEL or RFMD.

If you don't satisfied I can send you more detailed papers and lecture notes.

Good Luck

You are right for rfmd!
Thanks for article you'r refresh my mind.

Cel has NE38018, it's very good, what do you think about linearisation on 0.1-2Ghz range
with this transistor(gain is not important because other steps).

BTW, several days ago I searched for some VCO on rfmd site and I found only one VCO on all site-that was enough for closing window.
Now I see, I made Big mistake.

Hello
you may look for VCO at Zcomm
they have wide range of VCOs
thanks

Zcomm looks interesting, but I need widest range per VCO because I build Network Analyzer up to 10Ghz.
Agilent series VTO-8000 looks interesting for me.
Agilent's series of amplifiers ABA and INA are the best in this moment for my purposes because high linearity and small noise.
MGA looks interesting too beause very low noise, but linearity is not so good. If somebody have experience or some worked principle how to linearise this MMIC let's post.
I'm find Agilent's AN. with explication how to use MGA series on lower freq. This is very interesting, especially in combination with MGA-72543 and MGA-71543 (switch and LNA in the same pack).
Here is AN:

BigBoss,
I think that this formula is for receiver(detector and mixer) because 3dB.
Only for detector is without:
MDS=KTB+F+G

BigBoss,
I think that this formula is for receiver(detector and mixer) because 3dB.
Only for detector is without:
MDS=KTB+F+G

Another definition for diode detectors ,

Pmin= 10.log(SQRT( 4.k.T.B/M)) [in dBW]

k: Boltzman Coefficient
T: Ambient temperature (generally 300°K)
B: Bandwith
M: Figure of Merit

For more Information see;
I.Bahl , Bhartia.P " Microwave Solid State Circuit Design" , Wiley 1988

( But it's very expensive.. )

GL

Can we expeced that M is less than 1 and in 1/W ?
Can you send more information about this?

*******************************************
In this formula is not used Shot noise effect, only Nyquist noise.
How they are get this SQRT.
Noise effect of diode detector can be approximated with noise of output RC filter:
P=f3dB*K*T/2Pi, f3dB=2Pi/RC
I'm confused because SQRT 8O