The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> RF Design >> cmos down conversion mixer
https://designers-guide.org/forum/YaBB.pl?num=1202841373

Message started by udaykumarchintala on Feb 12^th, 2008, 10:36am

Title: cmos down conversion mixer
Post by udaykumarchintala on Feb 12^th, 2008, 10:36am

Hii,
I want know the design methodology for the down conversion mixers............help me.........

Title: Re: cmos down conversion mixer
Post by Stefan on Feb 12^th, 2008, 11:06am

There are plenty of books available.
Check for Lee, Razavi, Gray, Johns, Baker ...

Maybe also check for the phrases please, kind or other complimentary flowers of speech :)

Kind Regards,

Stefan

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Feb 17^th, 2008, 5:34am

hii,
What is the effect of matching circuits on the performance of the mixer?????Is it compulsory to design them??????????

thanks in advance,
uday

Title: Re: cmos down conversion mixer
Post by Stefan on Feb 18^th, 2008, 12:18am

That strongly depends on the rest of your design. A mixer not capable of driving the load is useless. However you can simulate the mixer with any matching around it that you desire, whether that's a "real world simulation" is another question ...

btw. check your "?"-key, I guess it's stuck someway.

Regards,

Stefan

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Feb 25^th, 2008, 5:08am

Hi,
In principle for a Mixer the LO signal is a square wave,but in most of the papers I have gone through they have carried out the simulations by taking the sine wave as a LO signal.Why?What is the difference?Do we get the desired output in both the cases?

Thanks in advance
Uday

Title: Re: cmos down conversion mixer
Post by Stefan on Feb 25^th, 2008, 5:10am

Sine-Signals do work better with simulations like HB or alike ...
It's not THAT realistic then, but the higher order harmonics are often of no interest, because they are filtered sufficiently.

Title: Re: cmos down conversion mixer
Post by didac on Feb 25^th, 2008, 12:23pm

Hi,
You should always make a design as close to the reality as possible, in this case the question is:which kind of signal will deliver the oscillator to the mixer? At very high frequencies it's not as easy as it seems to generate a good square wave, but if you consider the phase-noise requirements(to avoid reciprocal mixing) tunned oscillators generating a sine wave usually are more suitable to achieve the target specifications.

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Feb 27^th, 2008, 11:23am

hii,
How do we design a mixer for a desired gain,generally in a cascode the gain is gmRl.but in case of mixer how do we design for the same at a high frequency and the output frequency is also different.How to get a such a large gain at GHz ranges?

Thanks in advance

uday

Title: Re: cmos down conversion mixer
Post by loose-electron on Feb 28^th, 2008, 4:46pm

Mixers freqeuently dont have that much gain (12db at an absolute max? YMMV) :-/but the gain is in around them, pre and post.

Jerry

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Mar 1^st, 2008, 1:27pm

Hi,
In some of the previous designed mixers they used folded cascode structure,
the only difference between folded cascode and these mixers is that a switching
transistor .....is it right?Is the design procedure is same?

Thanks in advance

Title: Re: cmos down conversion mixer
Post by didac on Mar 2^nd, 2008, 4:03am

Hi,
Folded mixers are useful if your concern is voltage headroom since you don't have all the elements stacked one over the other. But beeing a different circuit topology brings to other design trade-offs like:between Vdd and the transconductor is a current source, a resistor(both lower the headroom),an inductor(if it's not a bondwire the finite Q could be losses),a LC tank tunned at fo?. Also the switching pair is a PMOS so it should be taken into account it's suitability in function of your operating frequency and technology.
Hope it helps,

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Mar 2^nd, 2008, 8:53am

Hi,
Thanks for your reply.....
what my question was the design procedure of folded cascode mixer is same as the folded cascode except the switching transistor?

thanks...

Title: Re: cmos down conversion mixer
Post by didac on Mar 2^nd, 2008, 9:11am

Hi,
I apologize if my answer was unclear. The design procedure it's the same for sizing(gm,rload,bias,LO amplitude), but my point was that beeing a completely different topology you should consider the specific issues of the circuit.

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Mar 2^nd, 2008, 9:36am

Hi,
thanks a lot.......
Can I find any tutorial for this design procedure?

Thanks.......
Uday

Title: Re: cmos down conversion mixer
Post by didac on Mar 2^nd, 2008, 10:50am

Hi,
You can check:
http://weble.upc.es/rfcs/Material/mixer_chandra.pdf
http://www.zen118213.zen.co.uk/RFIC_Circuits_Files/MOS_Gilbert_Cell_Mixer.pdf
Or the chapter of Mixers of Thomas Lee Book:"The Design of CMOS RFIC".
Hope it helps,

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Mar 2^nd, 2008, 11:04am

hi,
thanks for ur reply.......
From the links you have sent,I understood that from the gain equation we can directly calculate the size of the RF transistors from the biasing current and transconductance,no need to find the poles and follow the design procedure of a cascode....Is it right?

Thanks......

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Mar 6^th, 2008, 10:47am

Hii,
In a gilbert mixer The LO transistor should be biased in triode or saturation and why?

Thanks

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Mar 31^st, 2008, 1:10pm

Hii,
In my mixer circuit in the LO ,I have used a pulse source and its peak amplitude is 250mV and source resistance is 100 ohm...so how to claculate the LO power in dBm?and what is maximum LO amplitude that can be used in mixers?

Thanks.......
Uday

Title: Re: cmos down conversion mixer
Post by udaykumarchintala on Apr 7^th, 2008, 10:21am

Hi,
I have designed a folded cascode mixer at 2.4GHz RF and LO is at 2.3GHz,I'm getting 1 dB compression point as 0.56dBm,for calculating IIP3 I have given two tones as 2.401G and 2.4G , while plotting the curves, both the fundamental and third order components are intersecting what could be the problem?I have not designed the matching circuits......will it cause this?

Thanks in advance....