I did a quick search and found a NASA paper from 1980 by an M.C. Bailey. Is it possible Bailey became Bagley through typographical error?

Can't say I do. A picture might help. BTW, when I search on your site I get no result.

I had a book used for college transmission lines class. At the time, I hated the book. When I got a job after graduation this became my go to book. All my books I used for work have all the important pages marked with post-it notes. Later on, I rescued old microwave books bound for the landfill. (I coveted a Bode Feedback book where I worked but couldn't 'take' it!)

I know you can sell books but the value is low. I am considering using them for cremation starter kit for my books when I go.

Just a thought. A sneak path seldom considered is the DC supply lines. Signal from the output after 75 dB of gain can sneak back to the input if isolation is less than 75 dB. The 3 dB pad could be just enough to bring a little gain margin back in to prevent oscillation. I suspect an isolation (rf out to dc at gain block 2 to rf in to dc at gain block 1 ) of about 72 dB. (You can measure it with a spectrum analyzer and a good bias-tee. A quick check could also involve distinct power supplies for the two gain blocks.

Z= Zo x sqrt(mu_r/epsilon_r). Skin depth also decreases with increasing mu_r which in turn increases resistivity (rho= L/A and A decreases with mu_r).

Thanks UE but how do they do it? Teachable moment.

I have had good experience using a plater familiar with the Enthone process. The process roughens the surface so the plating can stick better than smooth surface. I have investigated all sorts of methods; vacuum deposition, conductive paint, plasma torch, conductive additive in plastic pellets, etc.

Here is my combline diplexer design (in plastic). Plating kind of sucks inside the resonators but this was the first one and design improvements improved the process. https://www.epner.com/capabilities/plating-on-plastics/

That "classic" uwave finish definitely did NOT apply for narrowband high-Q filters. Nickel was a big BIG no-no. We could even see detrimental effects (higher insertion loss) from the use of organic brighteners in the silver plate. I once evaluated many finishes applied to a standard cavity. The results are as you would expect; silver is best and things degrade rapidly with any changes. Silver is better than aluminum (bare) which is better than chromate conversion coated aluminum. Ultimately, I guess I cannot determine the meaning of your question.

It is an interesting and tough problem. There is a fine line between measuring the near fields accurately and distorting the very field you are trying to measure. The goal I suppose would be infinite impedance for E-probe and infinite admittance for H-probes. Any materials you are trying to measure will also alter the response of the probes. Like I said, tricky business.

Please provide dimensional information. Is relative mu (permeability) of NbTi equal to one? A non unity value would increase impedance. Here mu_r=1.22 would give the result you are seeing.

I guess I don't know but is SiC difficult to machine? I would suggest looking at shops that double disk grind pucks and ferrites for isolator/circulator manufacturers.

It seems to me if you are worried about shorted resistors after a solder operation that you have a design and/or process problem. Fix those and testing becomes less important. Just sayin'.

(And by the way, solder shows extremely well on xrays.)

Geez, we just printed identical resistors to nowhere adjacent to the "real" resistors and measured those. Did this for space products. The ohms per square of the resistor material is set and stabilized with a high temperature bake. The alternate resistors verify the process.

You can achieve negative group delay in a medium with anomalous dispersion (like seawater). This just means the signal arrives prior to the carrier (group velocity > phase velocity; what makes it anomalous). Media with high insertion loss can also yield negative group delay such as just outside a bandpass filters passband.

Not sure what you mean by topology? If you mean Tee or Pi type you may be surprised that most are neither, relying instead on a 2D slab of resistor material (which ohms out properly as if it were Tee or Pi). The big trick is broadbanding techniques especially when trying to ground the shunt. I would need to find an old paper to explain this type of attenuator. Madengr should know.

YIG is a very expensive single crystal. Yes the sphere would resonate but to "see" it at 6 feet would probably require gain (dc power, cost, etc.). I guess cost in quantity >~$40.

Buy and plate standard screws. However, when you make the cover (or whatever) with the threaded holes use an oversized tap to ensure a fit after cover plating (i.e. adjust the taps pitch diameter). This is how we did it in combline filter world. You can get almost any p.d. you want. I stripped a 2-56 hole and had it tapped up to 3-56 to save an expensive device. A good designer can figure out the appropriate p.d. depending on screw size and plating thickness. In contrast, screw manufacturers probably wouldn't make nonstandard screws unless you were willing to buy maybe 100,000 of them.