Ask Harlan, June 29, 2006

From: Jean Almira, Eudyna Devices Asia Singapore

Dear Harlan,
I would like to ask your advice regarding directional couplers. I am confused about backward wave couplers and forward wave couplers. I have read some articles: waveguide couplers in the forward direction (forward wave couplers) and microstrip/stripline couplers are backward couplers. I understand this is the rule of thumb, but I am not sure why? What are the main differences between forward wave couplers and backward wave couplers? Your advice is much appreciated.

Dear Jean,
While it is common that most waveguide couplers are forward coupling and most coaxial couplers are backward, it is possible to build either type in either transmission line. The direction of coupling is related to the phase relationship of the coupling structures. A good discussion of this with examples and math equations is in Chapter eight of Microwave Engineering-Passive Components, P. Rizzi, Prentice Hall, 1988, ISBN#0-13-586702-9.

From: Chin Keat Sim

Dear Harlan,
I am relatively new to UWB technologies. Could you recommend any reference books?

Dear Chin,
There are two books that may be useful to you:
Ultra-wideband Radio Technology, K. Siwiak, Wiley, 2004, ISBN#0-470-85931-8 and
Ultra-wideband Radar Technology, J. Taylor, CRC Press, 2001, ISBN#0-8493-4267-8.

From: George Vendelin

Dear Harlan,
Can the phase noise of an oscillator drop below kTB? It is very close right now.

Dear George,
Since kTB is the noise floor, I do not believe anything can drop below it. If so, how would you measure it?

From: Miroslaw Siemczyk, Polkomtel S.A.

Dear Harlan,
We are developing a software radio deployed on DSP. I am trying to locate a radio part; for example, a receiver for UMTS band 2 GHz. Could you please direct me to some manufacturers selling these types of modules ready to implement in measurement equipment?

Dear Miroslaw,
I am sorry, but we do not recommend specific manufacturers. You may be able to find such a module in the active components section of our buyer's guide, but I doubt that you will find a receiver module that you can use directly.

From: Constantine Kakoyiannis, National Technical University of Athens

Dear Harlan,
I am currently designing PIN diode switches for antenna feed networks at 2.45 GHz, and I would like to elaborate a bit on the TTL driver issue (my question was triggered by a related one that appeared previously): A. TTL drivers consume power that is precious in an energy-aware design. I have properly designed and laid out the diode circuit, and I have also provided access to a bias pin that is capable of supplying the specified voltage and current bias. Is it absolutely necessary to use a TTL driver to bias such diode circuits? B. The design mentioned above uses a single supply (VCC = 1.8 V). Therefore, it would be too "expensive" to generate negative bias voltages within it. In order to bias a series PIN diode in reverse mode, the best I can do is ground the bias pin. Would you suggest that a PIN diode with a zero DC voltage drop across it is sufficiently reverse-biased?

Dear Constantine,
No, you do not have to use a TTL driver. There are more efficient drivers available such as CMOS. Grounding the bias point will work at very low power levels, but will not give you the best isolation.

From: Richard Ney, CNRS/CETP

Dear Harlan,
We need a ferrite limiter for protecting the receiver of an X-band solid-state radar receiver. The maximum recovery time should be less than 20 ns for an 80 W peak power and 50 W average power. What manufacturers provide such ferrite limiters?

Dear Richard,
We do not recommend specific manufacturers; however, any of the companies in our buyer's guide producing ferrite components should be able to help you. As a technical point, I do not believe that a ferrite limiter can recover in only 20 ns. I suspect that you may need a TR-diode limiter. In that case a tube manufacturer will be your best choice.

From: Jon Breukel, Parliament of Victoria

Dear Harlan,
I sleep within 10 feet of a DS820 microwave/PIR intrusion detector that emits microwaves 24/7 at a frequency between 9.9 to 10.687 GHz. Is this level of exposure safe or should I seek to disable the sensor while in close proximity during the night?

Dear Jon,
Intrusion detectors of this type are limited to output power of less than 10 mW. The effective power falls off as the square of the distance from the antenna. These power levels have been shown to have no effect on humans. I would not worry.

From: Nandita Bhatt, ADE

Dear Harlan,
How can the insertion loss in BAW delay lines at X-band be improved? Can you tell me of any manufacturer making BAW delay lines up to 12 GHz other than Teledyne?

Dear Nandita,
Insertion loss is a function of materials, bandwidth, delay time and frequency. There is no single solution to reducing it. The best source of trade-off information will be the manufacturer. I am sorry, but we do not recommend specific manufacturers. You may find additional sources in our buyer's guide.

From: Anonymous

Dear Harlan,
I am currently determining specs for a stripline assembly from the 1960s, in order to make more. There are no written specs so I am just using test data to make the specs. I am confused by some of the mechanical aspects of the box, and am not sure what type of connectors are best for this type of assembly (I know what I need in terms of frequency, but the interface of the connectors to the board is something I am not sure of). You will probably be pleased to know that I have purchased your book, Stripline Circuit Design, and am waiting for its delivery. In the meantime, are there any other resources you can direct me to about practical things relating to stripline? I have read many, many IEEE articles, etc., but they focus on equations and theory, not on practical aspects of manufacture. I intend to have it manufactured, but I need to be able to spec the quality of my design first. Hope you can point me in the right direction.

Dear Anonymous,
It is impossible to answer your question since you do not describe the assembly, its purpose, original manufacturer, frequency, bandwidth, materials or original connectors. Since you wish to remain anonymous, I have to question your motives. If you will identify yourself and provide more specifics, I will try to help.

From: Renzo Basili, AMI

Dear Harlan,
I work in the field of calibration for power sensors (using the system 2 of the tegam). I run into a problem when I am trying to measure SWR of a thermistor mount hp p486 (waveguide band p) with a vector analyzer (VNA 360) not having the kit of calibration in guide. Could you provide any guidance to my problem?

Dear Renzo,
Yes, you will have a problem. A simple solution will be to add a wavequide adapter to the VNA and then use the standard SOL calibration at the output of the adapter.

From: EH Chong, Northeast Precision Sdn. Bhd.

Dear Harlan,
I am interested in pursuing the drop-in isolators business, as our current capability is able to produce casing only. Could you provide any advice on how to start this business from mechanical to complete module as finished goods with test?

Dear Mr. Chong,
The drop-in circulator is not a simple mechanical device. It is a complex electrical component with both microwave transmission line and magnetic circuits. Your first step should be to hire an EXPERIENCED microwave ferrite circuit engineer. Unfortunately, there are not many of them around. The market is very competitive. You will also need a good applications/sales engineer as well. Personally, I would not pursue it.

From: Stephen Morley, Delta Electronics Mfg. Corp.

Dear Harlan,
I hope all is well with you these days, we have not spoken since our brief discussions at M/A-COM. I have a question that I hope you can help me with. Actually, I would guess that the answer to this query would be of interest to a large number of RF/microwave designers.
Often product requirements require microwave design engineers to look at various alternative polymer compounds for use in new designs. In the past Teflon has been a material of choice for many good reasons and its electrical properties are well known; however, mechanical properties are often required in a structure that cannot be satisfied with Teflon. As designers view materials such as Ultem 1000 (polyetherimide), PEEK (poly ether ether-ketone), Noryl (polyphenylene Oxide), etc. (I have an extensive list), we find that there is very little data on dielectric constant and dissipation factor at microwave frequencies for many materials that have other useful properties for microwave applications. Our modern design tools can do a great job at structure simulation, only if the electrical properties are available. The companies that manufacture these materials have characterized them for all of the properties that are important to mechanical engineers, but we electrical types are left out in the cold. Often the best data that I can find is dielectric constant at 50 Hz of maybe 1 MHz.
In your vast experience, have you found any good source(s) for this information?

Dear Steve,
Your problem is similar to one that we faced in the early days of microwave PC board materials. Most manufacturers do not have the capability to make elecrical measurements and the quantities that our industry uses do not warrant the investment to acquire the capabilities. Some work has been done at NIST, but your best bet may be to buy or build a test chamber and make the measurement yourself. That is what we did for stripline materials in the early days.