Bandwidth is probably the most important question as some approaches are only good for narrow band.
Also, you should consider if the outputs need to have the same transmission phase...

Steve

Back in the day, companies would provide design notebooks, and even encourage engineers to have witnesses sign pages that revealed new designs (for consideration in patents). Some changes in patent law probably caused the death of that practice. I still like drawing with a #2 pencil on Ampad Quadrille paper. Like on this page

https://www.microwaves101.com/encyclopedias/h-tree-antenna-feed

It seems archaic that people actually write stuff down, but I still do. Sad to admit this but I write on the nearest scrap of paper, like an opened envelope... if it is important I will enter it somewhere in computerland. If it sits around long enough I will forget what I was thinking and it will be thrown away. My handwriting is so poor I often cannot make out what I was thinking!

Most engineers don't use tablets, because they need to carry around a laptop to process data.

I have no real expertise in apps, but sharepoint (word, powerpoint), google docs, helpdesk... there are many ways to document designs. Collaborative tools help speed up the design process but are open to malicious intent or accidental misuse, or VPN issues. There is an entire cottage industry that will tell you what the best process is, even though the proponents have never designed anything useful in their lives.

I would be interested in hearing from other people on this topic.

Steve

This might be milking mice, but the 2.4mm will have slightly less loss than the 1.85mm....
Steve

Thanks to you both

I think I will stay away from trying to calibrate to that interface, and somehow model the adapter and subtract it off as an S-parameter file or maybe just a linear model. I am no expert on time domain but it seems like when you gate out a reflection the improvement sometimes too good to be true... This is not for device characterization, more like an acceptance test. A good design should have plenty of room for error, right?

I think about an acceptance test at a previous employer where phase tracking was needed over wideband. The operator was trained to misalign the GPO until it met the spec, then hit the save button... yikes.

Steve

I update the that page slightly, how does this look?

https://www.microwaves101.com/encyclopedias/microstrip#types

Steve

My experience with Microwave Office is that they just throw away the imaginary part of the impedance in transmission line calculations. Sorry I am no help!

Steve

Yole seems to be the most-cited microwave researcher

http://www.yole.fr/company.aspx

Of course, you have to pay for more than just a taste of the data! Other than that, I don't have any inputs. Except that 5G frequencies are very popular these days.

Steve

Can we agree to just call it quasi TEM?

https://www.rfwireless-world.com/Terminology/difference-between-TEM-wave-and-quasi-TEM-wave.html#:~:text=Quasi%20TEM%20wave%20mode%20exist%20in%20a%20microstrip%20line.&text=Hence%20in%20microstrip%2C%20wave%20propagates,referred%20as%20quasi%20TEM%20mode.

Thanks
Steve

I agree, you need a diplexer to split the signals, then operate on them independently

https://www.microwaves101.com/encyclopedias/diplexers

Steve

Tinus

Thanks for this post! I never considered the effect of surface roughness on propagation delay, but it makes sense. This deserves a new page on Microwaves101.

https://www.microwaves101.com/encyclopedias/surface-roughness-effect-on-propagation-delay

I personally think we should also start a page on "how printed circuit boards ruined my career".

Thanks

Steve

I am not an antenna designer.... but isn't part of antenna design creating some type of match closer to 50 ohms internal to the antenna? Or maybe just accepting that 3:1 or 4:1 mismatch is good enough?

I agree there that trying to realize 300 ohm in anything but twin lead. Then how will you get twin lead at all the various impedances indicated by the max flat transformer? Eight sections is going to be very long at this frequency band.

How about using a simple "twelfth-wave" transformer?

https://www.microwaves101.com/encyclopedias/twelfth-wave-transformer

Best of luck in any case, and thanks for engaging here!

Steve

Ben

Thanks
Funny how the data sheet cuts off at 19 GHz, good to know it works to 23 GHz! I check, it is "production" status. They have been obsoleting quite a few parts lately.

Steve

This site shows a used one for $250

http://www.aptecelectronics.com/WR-42.htm

Good luck!

Steve

By the definition of bandwidth, you can get 200%, in theory, but that would take a low-pass filter! It is just a convention.

If you have 10 GHz bandwidth centered at 8 GHz, you have 125%.
Steve

I think that the Millitech product line either originated at Hughes, or used Hughes diodes.

I found an old Hughes mmw catalog here

http://www.dudleylab.com/hughes%20mmwave%20products.pdf

Gunn diodes are on page 76

Maybe 17 means 170 mW?

Maybe searching on Hughes will help. Or contact Eravant... a competitor might know about this.

Good luck

Steve

I agree about recommending MWO, been using it for nine years. All good except the VSS product still confuses me, so I am gonna try out Genesys.

Have you looked at Microwave Office with Axiem? I was forced to switch from ADS, nine years ago, now I don't want to go back. It may be cheaper; I don;t get involved with pricing.

Does anyone have an opinion on what the best shareware is for a retired microwave engineer? Asking for a friend...

Steve

I have run into dispersion in coax, at low frequency. It has to do with skin attenuation, but my math skills are so rusty I gave up on solving for it. I started to write a page on it here:

https://www.microwaves101.com/encyclopedias/low-frequency-dispersion-in-tem-lines

Just remember, at low frequency, lines get electrically longer!

It is interesting that in Microwave Office this effect is NOT modeled in the coax circuit element. But it IS modeled in the stripline element! So you can fake a coax design with the stripline model if you don;t go to far outside the limits of the relative thickness of the strip versus ground plane spacing.

Once the center conductor diameter is large compared to skin depth the problem dies down quickly. Virtually every design above 1 GHz can ignore coax dispersion. The real killer will be at higher frequencies where the coax starts an unwanted mode.

https://www.microwaves101.com/encyclopedias/coax#TE11

Steve

We are going to need some expert help here... I have exhausted my limited knowledge of sources. My guess is that this is done in the digital domain, it would be hard to maintain 1 degree resolution using analog networks.

Where are all the smart people out there when you need them....

Steve

Weird that it was there for free

I downloaded a copy. One of these days I will try to organize some more material properties stuff and post...

Steve