Q on down mixer gain and noise....

[chacha]

I am designing an active down mixer for BT applications.
The receiver is a direct conversion type. I am facing a big difficulty in getting gain and noise figure requirements.

The low frequency noise is due to the bias current of the mixer. I cant reduce the bias current as i lose the gain.

Can any one suggest me how can i improve noise figure without losing gain??

 

[noiseless]

when reduce current, you can increase the resistor load to maintain the gain. If the voltage drop is too large on the resistor, you can design a folded one

 

[chacha]

for a folded mixer the current consumption will be more.

iam already using more current for high gain. max allowed is 4mA. I am using 3.8mA.
how can i reduce the noise due to switching quad in the gilbert mixer??

 

[noiseless]

more current, more noise, in which, current is referred to rf transconductance stage current. in the folded ones, more current is coming from load transistor, not rf stage.

 

[chacha]

what i found from the simulations is that if i decrease the bias current of the gm stage and increasing the same in the switching transistors then noise is increased.
now NF is worsened( in the case of folded mixer).

 

[bigworm]

maybe you can modify the size of your mos while keeping the ratio since major noise come from 1/f

 

[safwatonline]

hello,
are u using any noise cancellation technique like dynamic current injection or filtering or dynamic matching (in case of active loads)
try reading simple model for noise by Darabi and noise cancellation technique by Darabi too

 

[chacha]

hi safwatonline,

i have implemented that also.
in that how can we check the switching of the back to back connected pmos transistors??

in the noise report, it is given that the maximum noise is contributed by the current bias transistors(in mixer core and post mixer amplifier)...
is there any way to reduce that???

 

[stober]

Are you trying to down convert 2.4GHz rf signal directly to baseband? I think this is not a good construction.

 

[chacha]

yes, iam doing the direct down conversion, 2.4G to 1M.

 

[khouly]

dear safwatonline

can u upload some data and papers about noise cancellation techniquces

this will be helpfull

khouly

 

[safwatonline]

dynamic current injection and filtering and dynmaic matching

 

[safwatonline]

dynamic matching