The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
https://designers-guide.org/forum/YaBB.pl?num=1367770686

Message started by iVenky on May 5^th, 2013, 9:18am

Title: Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
Post by iVenky on May 5^th, 2013, 9:18am

I have attached the inductive degenerated double balanced mixer. It is said that they are more sensitive to supply and common mode noise. Why is that so? Also it says LC tank circuit is for improving the CMRR. I don't understand the reason behind that either.

Thanks a lot

Title: Re: Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
Post by raja.cedt on May 5^th, 2013, 11:37am

hey,
i don't think so about supply noise sensitivity, since it is differential circuit i would expect high common mode noise rejection. Coming to CMRR at resonance frequency, impedance at tail node very high hence high CMRR.

Thanks,
Raj.

Title: Re: Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
Post by iVenky on May 13^th, 2013, 7:34pm

Thanks for your reply.

How to choose the resonant frequency?
From what I learnt it looks like it is the common mode frequency. It doesn't make any sense to me. Usually I have seen this common mode to be noise. Do we have common mode with with a particular frequency too?

Thanks a lot

Title: Re: Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
Post by @@k@sh on Jun 22^nd, 2013, 5:15am

@raj correct me if wrong
In amplifier with passive load, common mode noise(Id) in transistors causes equal common mode voltage noise in both branches and hence cancel out. Same holds for upper transistors of mixer.
But the common mode noise(Id) in lower transistors will be differential multiplied by the differential LO signal and since we have 2 lower (with RF@gate) transistors the effect is doubled. It's just like increasing the bias current and hence the gain of amplifier with LO signal as i/p

@venky
LC tank is used to filter out the unwanted LO even harmonics, so that RF signal is modulated by fundamental LO freq. Infact you can have multiple tanks to filter out 2LO, 4LO, 6LO .... freq

Title: Re: Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
Post by rfcooltools.com on Jun 22^nd, 2013, 7:33pm

Noise current from one of the input RF devices is present as a noise voltage on the the tank at the tanks resonant frequency. This noise is gained up by the other RF input devices as if it where a common gate amplifier.

http://rfcooltools.com

Title: Re: Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
Post by sheldon on Jun 23^rd, 2013, 1:44pm

iVenky,

This circuit is pseudo-differential. Any change in the common mode
level looks like an applied signal. In a real differential circuit the
tail current source has a high output impedance and the common
mode rejection at dc is the ratio of the transconductance of the
gain devices to the output resistance of the current source, or
very high. For a pseudo differential stage, both the differential
and common-mode signals have about the same gain to the
output so compared to a true differential circuit the sensitivity
to common mode noise is higher. By extension, if ground moves
around on a true differential circuit, then the current at the output
is Vgnd,noise/routput. For a pseudo differential design, the current
at the output is Vgnd,noise/rtransconductance. So the gain from
the power supply to the output is higher or the sensitivity to power
supply noise is much higher.

Sheldon

Title: Re: Why is inductive degenerated Double balanced mixer more sensitive to CM noise?
Post by raja.cedt on Jun 23^rd, 2013, 3:25pm

@@k@sh: Noise cancel's for any kind of load as long as they have good matching, there will be noise cancellation, passive load consideration is wrong. To be frank i didn't understand rest of your post, be clear about common mode noise, for example tail current noise will be common mode for all all transistors only when you apply zero differential signal, infact if you apply some large signal to the extent that differential pair current steered to one said then there is no noise cancellation.

@sheldon: How can you say that it is pseudo differential circuit, At the resonance frequency (or LO frequency), of course at dc you are correct but if we can some how manage good ac coupling and proper layout then it has good rejection.

Thanks,
Raj.