The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> IQ dwon mixer design
https://designers-guide.org/forum/YaBB.pl?num=1207643361

Message started by silentgun on Apr 8^th, 2008, 1:29am

Title: IQ dwon mixer design
Post by silentgun on Apr 8^th, 2008, 1:29am

hi, all
for most receiver, there need IQ down mixers. for active mixer design, there's two way to
achiev the IQ down conversion
one way is designing a seperated mixer in I-path, and just putting a copy of mixer in Q-path ;
another way is designing an IQ combined mixer which shared one transconductance stage.

it seems IQ combined mixer has lot of advantages over seperated mixer, like less power, better NF and
better linearity.
but it seems to me that seperated IQ mixer is more popular in curcuit design.
does IQ combined mixer have any disadvantage which make it unpopular?

thanks

Title: Re: IQ dwon mixer design
Post by pancho_hideboo on Apr 11^th, 2008, 9:12am

silentgun wrote on Apr 8^th, 2008, 1:29am:

hi, all
for most receiver, there need IQ down mixers. for active mixer design, there's two way to
achiev the IQ down conversion
one way is designing a seperated mixer in I-path, and just putting a copy of mixer in Q-path ;
another way is designing an IQ combined mixer which shared one transconductance stage.
it seems IQ combined mixer has lot of advantages over seperated mixer, like less power, better NF and
better linearity.
but it seems to me that seperated IQ mixer is more popular in curcuit design.
does IQ combined mixer have any disadvantage which make it unpopular?
thanks

[Merit]
I-Path Gain and Q-Path Gain are well-matched.
If it is used as image rejection mixer, NF is better than separate transconductance amplifier.
Local leak to RF will be 4 times frequency of Local.

[Reference]
A merged CMOS LNA and mixer for a WCDMA receiver
Sjoland, H.; Karimi-Sanjaani, A.; Abidi, A.A.;
Solid-State Circuits, IEEE Journal of
Volume 38, Issue 6, June 2003 Page(s):1045 - 1050

[Demerit]
If Local signal doen't have enough amplitude and then gilbert switchs are operated in finite transition time,
I-Local and Q-Local have some interference via common transconductance amplifier.
It causes degradation in image rejection performance.

Title: Re: IQ dwon mixer design
Post by pancho_hideboo on Apr 13^th, 2008, 8:28am

What opinion do you have for the following topic ?
http://www.edaboard.com/ftopic305231.html

Title: Re: IQ dwon mixer design
Post by silentgun on Apr 14^th, 2008, 8:08pm

hi, thank for your repley

i used saw those resistors in mixer design, as for as i know
they act as current steering things between I/Q
when current flows through I-path, making sure Q-path current minimal
that will improve 1/f and gain

Title: Re: IQ dwon mixer design
Post by pancho_hideboo on Apr 15^th, 2008, 12:34am

silentgun wrote on Apr 14^th, 2008, 8:08pm:

hi, thank for your repley
i used saw those resistors in mixer design, as for as i know
they act as current steering things between I/Q
when current flows through I-path, making sure Q-path current minimal
that will improve 1/f and gain

You mean 1/f as fricker noise contribution to mixer output ?
If so, why does it improve fricker noise ?

Title: Re: IQ dwon mixer design
Post by silentgun on Apr 15^th, 2008, 7:49pm

as we known, switches contribute flicker noise to mixer output when LO cross over
making current flowing through swiches as low as possible at the time LO crossing over will compress 1/f noise contributiong of switches.
that's why the resistors there.
when I swiches are totally opened, Q switches are just at the crossing point.
the resistors steer all current into I-switches, make the 1/f noise contribution of Q-switches minimal