Ask Harlan, September 28, 2006

From: Qing Sun, University of Manchester

Dear Harlan,
What is the relationship between coupling factor and frequency in couplers and the bandwidth of a coupler as a function of even/odd mode impedance?

Dear Qing,

The bandwidth of a coupler is not a function of coupling factor and even/odd mode impedances. It is related to the type of coupler. For quarter-wave coupled line couplers the normal acceptable bandwidth is one octave (2:1) regardless of coupling factor. This bandwidth can be increased to as much as 25:1 by adding additional sections and cascading coupling sections. There are a number of design examples in Chapter 5 of my book Stripline Circuit Design, H. Howe, Artech House, 1974, ISBN#0-89006-020-7.

From: Ahmed Boutejdar, Universittat Magdeburg

Dear Harlan,
I would like to know how to calculate the coupling matrices of 2 DGS in metallic ground and 1 DGS in top.

Dear Ahmed,

We have seen a number of papers recently with DGS structures and good results; however, none of them have explained how to calculate the characteristics. They have been created using computer synthesis or by empirical techniques (design by fooling around).
If anyone has actually done calculations that are usable by others, we would welcome a paper for possible publication in Microwave Journal.

From: Yong-Sub Lee

Dear Harlan,
Can you provide a prospectus of the power amplifier device market (for example, LDMOS, GaN, SiGe, etc.)?

Dear Yong-Sub,

I am sorry but we do not do market research so I cannot answer your question. There are companies that specialize in market studies; however, their reports tend to be very expensive ($5000 to $20,000) and we do not buy them.

From: Hyun-Chul Park, Shungkjem Kwan University

Dear Harlan,
What is the non-linearity of HBT, VBIC models? How do they work at high power?

Dear Hyun-Chul,

While there are generic models for most devices, they are only as good as the device parameters provided by the manufacturers or measured by the user. There is a good section on device modeling in RF and Microwave Semiconductor Device Handbook, M. Golio, CRC Press, 2003, ISBN#0-8493-1562-X.

From: Kim Sang-Gyu, Integrated Technologies

Dear Harlan,
To realize UWB mobile communication, I am working on a UWB filter, but I am having difficulty with the group delay characteristic. What is the acceptable group delay value?

Dear Sang-Gyu,

I do not know of any general specification for UWB group delay. The major problem appears to be with delay spread due to multi-paths. There is an extensive discussion of delay spread in Ultra-wideband Radio Technology, Siwiak & McKeown, Wiley, 2004, ISBN#0-470-85931-8.

From: A. Kumar, Microwave Associates

Dear Harlan,
What is the physical meaning of even and odd mode impedance of a coupled stripline? Somewhere the terms common and differential mode impedances are mentioned. Are these terms the same as even and odd mode impedances? If not, what exactly is the difference?

Dear Mr. Kumar,

The even and odd mode impedances are the impedences between coupled lines for an unbalanced transmission system such as stripline. Common mode and differential mode impedances are related to balanced lines such as two wire lines. There is a discussion of even and odd mode impedances and methods of calculating them in Chapter 4 of my book Stripline Circuit Design, H. Howe, Artech House, 1974, ISBN#0-89006-020-7.

From: Majid Haghighat, Isfahan University of Technology

Dear Harlan,
Have you seen or read any articles that demonstrate an application of microwave in gas distribution pipelines to avoid freezing of natural gas in cold areas?

Dear Majid,

I have not seen anything written on this subject. Since heating requires the absorption of energy and pipelines tend to be very long, it is unlikely that sufficient energy could be distributed to do the job without creating a hazard.

From: SH Wong, Agilent Technologies

Dear Harlan,
In a spectrum analyzer and a peak power meter, one of the key specs is referred to as video bandwidth. Why is it called "video" bandwidth?

Dear Mr. Wong,

Video bandwidth is the final information bandwidth after downconversion from the input RF. Why it is called "video" probably dates back to early usage

From: Peter Karuga, ITACS

Dear Harlan,
I am interested in calibrating an open area test for radiated EMC testing, 30 to 1000 MHz. I have only one set of calibrated antenna in the frequency range and some homemade dipole. Could you please let me know how I can use the homemade dipole and the calibrated antenna to calibrate the site? Do I need to calibrate the dipole beforehand and if yes how can I achieve that?

Dear Peter,

You will need an additional test antenna, which can be uncalibrated. Set up your measurement with the calibrated and test antennas and then substitute the homemade dipole for the calibrated antenna and compare the results.

From: Mark Kennard, Kencom Electronics

Dear Harlan,
Could you please explain the result of using the incorrect material for PCB manufacture when building RF amps in the 900 MHz band? How many different types are there, and which one is the most suitable for 900 MHz use?

Dear Mark,

Any of the commercial materials will work at 900 MHz. The most popular materials include Woven Teflon Fiberglass, Microfiber Teflon Fiberglass, Ceramic filled Teflon and FR-4. FR-4 is usable at 900 MHz, but not all of the manufacturers control the dielectric constant. It is the lossiest but is also the least expensive and can be processed by any PCB house. If space is a consideration, the ceramic filled materials have a higher dielectric constant but are the most expensive. My personal preference is the glass filled Teflon products.