why the LNA in o.18um,in the different processor TT,SS,FF,the output impedance changed very much .in the TT Z=35-j*41
in the SS Z=60-j*62;and in the FF,Z=7-j*7
who can tell me why ,my LNA is cascode instructure .
First of all you should have an expression, although aproximate, of you output impedance. Then you can understand which component with process spread is giving the main contribution and if it is only one or the combination of more than one effect.
Than you can try to understand more by simulation: let vary only one type of component (inductor, resistor cap and active device) and see what happens.
Mazz
First of all you should have an expression, although aproximate, of you output impedance. Then you can understand which component with process spread is giving the main contribution and if it is only one or the combination of more than one effect.
Than you can try to understand more by simulation: let vary only one type of component (inductor, resistor cap and active device) and see what happens.
Mazz
Do you have resonant load? And, if so, is narrowband?
You may have a look at output capacitors (total capacitance to the output node) load, as you know caps have big spread in IC.
Mazz
Do you have resonant load? And, if so, is narrowband?
You may have a look at output capacitors (total capacitance to the output node) load, as you know caps have big spread in IC.
Mazz
hello Mazz:
I try again ,is the cause of output matching caps,my output
matching network is series a cap ,and parallel a cap.How to decrease the caps spread? thank you very much.
Do you have resonant load? And, if so, is narrowband?
You may have a look at output capacitors (total capacitance to the output node) load, as you know caps have big spread in IC.
Mazz
Do you have resonant load? And, if so, is narrowband?
You may have a look at output capacitors (total capacitance to the output node) load, as you know caps have big spread in IC.
Mazz
Try to decrease the Q factor of the output load. This will not influence a lot the NF in the cascode (I don't know if you are designing this 802.11a amplifier to make a product or an article: maybe you can survive with small higher NF), but maybe the output matching network in this way (of course has to be modified) has to do an easier job and the spread of the components will have less impact.
Mazz
Try to decrease the Q factor of the output load. This will not influence a lot the NF in the cascode (I don't know if you are designing this 802.11a amplifier to make a product or an article: maybe you can survive with small higher NF), but maybe the output matching network in this way (of course has to be modified) has to do an easier job and the spread of the components will have less impact.
Mazz
why decrease the Q of theoutput load ,can descrease the effect of the spread of the caps,I think lower Q load ,the NF maybe is the same to higher Q,but the gain maybe descrease.
The component model has different parameters at different corner. So your LNA will show different output impedance. The higher operating frequency the more different output impedance.