Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

single-balanced vs double-balanced mixers

Status
Not open for further replies.
S

Superowl

Junior Member level 2
Joined
Nov 24, 2004
Messages
23
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
226
This question is about single-balanced mixers vs double-balanced mixers (MOS transistor mixers):

1) Which has more conversion gain and why?
2) Which has better IP3 and why?
3) Say both is at a certain RF freq. With LO freq at a slightly lower freq. What would be the effect (on conversion gain and IP3) of decreasing the LO freq even more?
4) An added source inductance would have what effect on the conversion gain and IP3 for both?
 

double *passive* mixers have the highest IP3. For active MOS mixers, I have read conflicting arguments.

Why do double-balanced active MOS mixers have higher IP3 than single-balanced?
 

The advantages of a double-balanced mixer over a single balanced mixer are increased linearity, improved suppression of spurious products (all even order products of the LO and/or the RF are suppressed) ant the inherent isolation between all ports.
The reason for increasing linearity is easily understood; the incident power is simply shared amongst at least twice as many active components, thus increasing the signal handling capability.
The disadvantages of double balanced mixer are that they require a higher level LO drive and increase complexity.
The conversion gain is sensitive to the unloaded Q of the balun/inductors. More inductors or baluns, less conversion gain you have (higher insertion loss).
The roll-off frequency of the inductors/baluns affect the conversion gain, if the LO is close or far from RF signal.
Usually a tuned source inductance improves the conversion gain.
 
This is interesting - I was reading this paper (I attached it): it talks about why a single-balanced mixer has higher IP3 (pg 114) - let me know your thoughts!

Follow up the IF freq question:
Case 1) RF=4GHz, LO=3.95 GHz
Case 2) RF=4GHz, LO=3.5 GHz
Which has better conversion gain and which has better IP3 and why?


It was also interesting when you said a tuned source inductance could optimize the conversion gain. Could you elaborate a little more on that (how/why)?

Thanks!
 

The IP3 and conversion gain of the mixer will depend mainly of the transconductance amplifier.
Increasing the source degeneration resistors Rs, can increase the linear input voltage range, but source inductance can also provide beneficial degeneration.
The inductance values required would be too large for RFIC implementation (depend by frequency range), thus sometimes we are stuck with the resistors (they will add noise).
The load resistors could also cause gain compression if the voltage swing at the drains is large enough to cause the output to clip under large signal drive conditions.
 

So regarding the follow up the IF freq question:
Case 1) RF=4GHz, LO=3.95 GHz
Case 2) RF=4GHz, LO=3.5 GHz
What you're saying is, both should have roughly the same IP3 and conversion gain? (or were you talking bout something else)

And I still don't quite understand how a tuned source inductance provides an optimal conversion gain.
 

A doubled balanced mixer is said to be much linear due to the ability of this topology to suppress the image signal. Usually a degenerated resistor is used to increase the linearity of the transconductance stage but a high value of resistor results in the reduction of gain and increase in thermal noise.

Regarding the tuned value of inductor issue, I think in this case the inductor is tuned to oscillate at a resonant frequency of interest ( which is the IF) with the parasitic capacitor to filter out undesired spurious signal

Rgds
 

Yea, I know what a resistor does to the circuit - that's why I'm NOT interested in resistors. Rather, my question is about INDUCTORS.

And inductors on conversion gain.

I know that source inductance generally worsens the conversion gain.

BUT, I've found that there's a certain small inductance value that if chosen right, can actually INCREASE the gain.

Any insights as to why that's so? - that's my question.

Thanks very much!
 

Everything is related to matching of impedances:
In common-source topology, the source degeneration inductor (Ls) introduces a real part into the input impedance seen looking into the gate. This impedance is used to match the amplifier to the source impedance, Rs.
The maximum conversion gain is, when for an input power match Rin must be equal to Rs
The real part of the input impedance, Rin is given by:
Rin= (gm/Cgs)*Ls
where
gm = transconductance
Cgs = gate-source capacitor,
Ls = source inductor.
 

    S

    Superowl

    Points: 2
    Helpful Answer Positive Rating
Thanks that helps. Can you give me a paper or a source where that came from? (The real part of input impedance.)


Just to make sure:

Rin is the resistance looking from outside into the gate (LO).

Rs ... does that refer to the INTERNAL resistance of the source of the transistor? Because I don't have any external source resistors.
 

I am talking about internal source impedance.
To understand better what is going on, you have to see this stage as an amplifier (or a driver) for your mixer.
In the design of a LNA, or Mixer driver stage there exists an optimal source impedance to provide the maximum power transfer into the transistor and thus the maximum gain.
Probably is hard to find mixer papers that describe this situation, but definitely you will find for amplifiers. Do a web search for this.
Everything that I specified is from my experience.
 

Thanks. I was wondering about the Rin= (gm/Cgs)*Ls equation. Is this derived or mentioned in some literature or book, or did you come up with that equation yourself?


Also, the conversion gain of a single balanced vs double balanced mixer: The double balanced has only 1/2 the gm, does that mean it only has half the gain? (what about Rout?)
 

I was wondering about the Rin= (gm/Cgs)*Ls equation. Is this derived or mentioned in some literature or book, or did you come up with that equation yourself?


It seems quite the same with the Rin of LNA, I guess the derivation is the same too.
You can find it in papers about LNA quite easily.
Click to expand...
 

Really? I'm not finding any papers or sources that has the Rin of Mixers OR LNA ...


Also, the conversion gain of a single balanced vs double balanced mixer: The double balanced has only 1/2 the gm, does that mean it only has half the gain? (what about Rout?)
 

Well, if u just check IEEE papers about LNA of Goo or the cmos rfic book of T. H. Lee, you would find it.
You can make a simple caculation to get it.
Just use the most simple mos model that got only Cgs and gm.
I don't think the double balanced mixer just have the 1/2 gain.
As Rout, which part do u mention here?
 

A single balanced with a load resistor of 500 ohms at each transistor.
A double balanced with a load resistor of 500 ohms at each transistor.

What is the relation of the equivalent output resistance for the two? (same, or double has twice the output resistance?)
 

Hi

I noticed that Noise Figure was not one of the parameters? Is it not important or do you assume that higher conversion gain automatically means better NF (which is common but not always)?

A.T.B
 

it's important, but just focusing on gain and IP3 now.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top