An inquiry about biasing techniques for the LNA

[mostafa_maarouf]

Hi all;

I'm desigining a front end for WCDMA receiver and i have designed the LNA and working in the mixer..
But i need to bias the LNA (and of course the mixer later ) so what type of bias circuit i can use..

I read a little about the constant gm bias but this needs an external resstance which i think it is not that pretty solution..
so any help in any topic will be highly appreciated..
Best Regards,
Mostafa El Khouly

 

[khouly]

AA
i have worked with alot of chips from very big companies , in CDMA fornt ends , and i can say everyone have internal constant gm bias , and we add a very high quality resistor off chip , so the constant gm circuit work , so it is good solution

other wise u need to a good bandgap reference and bias cell for ur LNA and mixer 4 sure

khouly

 

[mostafa_maarouf]

thanks alot khouly, so it is normal speaking that i use a constant gm bias circuit and put a high quality off chip resistance ..
if i wanted to make this resistance on chip , will there be any practical case or it will be catastrphe

 

[khouly]

to put it on chip there will be about 20 to 30 % tolerance , the resistor i used with the IC's were about 1% tolerance so take care

i am not expeinced in IC design , but i think it is hard to make the constant gm circuit fully integrated

khouly

 

[toonafishy]

We always use 1% resistor off chip. As khouly points out, the 20% on chip is too much to handle.

 

[Grig]

Hello
Try cmos based resistor.
Note: gm stability (process) will be less, simulate this one before.

 

[nxing]

Actually, in LNA design, the absolute current is not that critical and I think the constant Gm curcuit will work.