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.

A question on balun tuning

Status
Not open for further replies.

wizardz

Member level 2
Joined
Aug 5, 2006
Messages
52
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,286
Activity points
1,740
Typical on-chip transformer balun needs tuning with capacitors to minimize loss in their use in actual circuit.

Then, how about Machand Balun, which is based on coupled lamda/4 transmission lines? I read papers that indicate that typical insertion loss( S21 and S31) of planar Marchand Balun are about -5dB. Is it possible to tune these baluns to reduce loss (e.g. with capacitors)?

Thank you!
 

You probably mean Marchand balun. This balun is based on splitting the power, so the minimum theoretical insertion loss cannot be less than 3 dB, Practical loss for well designed and built balun will be about 3.5 dB. You cannot tune it out by capacitors. I never see balun tuning with caps. Balun is relatively broadband device and must be well matched. The loads must be resistive, balun does not like any reaactance at any of ports. Therefore you need to make loads as resistive as it possible, but not to try to tune balun. Of course tunung is possible for very narrow band balun, when reactance practically does not change across the band. But it is narrow band match. Or do you use word tuning for matching?
 

Thankyou RF-OM for your insightful help.

Sometimes, on-chip transformers can be designed as three port devices to provide one single end signal and two balanced signals with 180 degree phase difference. In real CMOS RF circuits, these three port transformers are tuned with MIM capacitors to minimize loss at the cose of bandwidth.

As transfomers are lumped elements, they have the benefit of small size.

Then, can such kind of tuned tranformers replace transmission line baluns if the we do not need broad bandwidth?
 

I have not so much experience to design on-chip balun, but I believe the basic principles are the same. As I know three-port balun is always single-ended on one side and single-balanced on another. For three port device it is impossible to have two balanced signals on the other side of the balun. The balanced output signal can be represented by two vectors that are equal in amplitude and opposite in phase (180 degree phase shift). The amplitude and phase imbalance are important balun parameters and should be kept minimal. I cannot tell you about MIM capacitors because I never use them, but I will ask about it our IC designers. May be they know what it is and why this caps are used. In lamp element designs I never saw or used capacitors to tune the balun. For me it is very strange. Any reactance in balun circuit are parasitic and must be avoid.
 

I've read couple of papers on this topic.

Although on-chip transformers with tuning capacitors have input/output impedace as rectance, they sometimes also work as part of impedance matching circuits such as LNAs or mixers.

It seems that they are transformers with special connection set up that provide such fuctions as baluns, although I doubt they are real baluns
 

I talked to the head of our IC department and as it was expected he told me the same as I already told you. There is only one case when reactive elements can be connected to the balun: matching circuit. This elements together with balun and the rest of the circuit must provide resistive loading for balun. Probably such elements can be tunable but it is not balun tuning, rather it is matching circuit tuning in oder to convert circuit impedance to balun load resistance.
 

and, I have one question,
now in my design, the foundry provides several model of transformers, stacked and interleaved. and they provide the performance parameters.
I wonder how to verify the performance?
in my bench, one port of the primary is connect to 50ohm port (that is the case I will use) and the other connect to gnd

the secondary is connect differentially to two 50ohm port.
could this work? the simulation results seems to be somewhat inconsistent with foundry.
 

To bigworm,

It is not clear to whom you are asking your question. If me, sorry I cannot answer because I am not IC designer.
 

RF-OM said:
I have not so much experience to design on-chip balun, but I believe the basic principles are the same. As I know three-port balun is always single-ended on one side and single-balanced on another. For three port device it is impossible to have two balanced signals on the other side of the balun. The balanced output signal can be represented by two vectors that are equal in amplitude and opposite in phase (180 degree phase shift). The amplitude and phase imbalance are important balun parameters and should be kept minimal. I cannot tell you about MIM capacitors because I never use them, but I will ask about it our IC designers. May be they know what it is and why this caps are used. In lamp element designs I never saw or used capacitors to tune the balun. For me it is very strange. Any reactance in balun circuit are parasitic and must be avoid.



yeah,you are right.i saw this kind of design solution.

content your talking is very useful!


best wishes
yours le yao
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top