Testing/Teardown of a 4-way 500 MHz to 6 GHz Wideband Wilkinson Power Divider from Werbel Microwave

In this video, I tested a 4-way Wilkinson power divider/combiner from Werbel Microwave (D-4066) and took a deeper look at the broadband Wilkinson divider construction techniques.
Werbel Microwave: werbelmicrowave.com/
werbelmicrowave.com/product/w...
werbelmicrowave.com/product/s...
00:00 Overview, specifications
05:06 Insertion loss, isolation testing
10:35 Testing beyond the specifications
12:32 Teardown, conclusions

[@mikkolankinen3330]

I don't think there are any additional features on the other side. The design looks like a typical microstrip Wilkinson. Microstrip structures require a solid ground on the other side. Those pieces of wire soldered on the line looks like a tuning stubs used to adjust performance.

[@TonyLing]

Brilliant

[@cambridgemart2075]

Very interesting, just slightly spoiled by reflections of your lights on the VNA display making it difficult to read.

[@davidgrey943]

Hi, Kerry there seems to be a dry joint on Ports 3&4 where the connections comes through from the bottom of the board. I am comparing between the Ports 1&2 where the connections from the bottom of the board I hope this observation has prevented a fault later on.

[@Manf-ft6zk]

I also realized it but it seems that this wire is partly well connected what might be sufficient. I also look at the fine adjustments (cuts) at the last loops in front of the output. They are identical for the output ports but they are different from the shape of the first divider, what might well make sense. When those fine adjustments are relevant then I wonder about the wires at the through connections, they are hanging over quite a bit coupling with the following lines.

[@ElektronikLabor]

Thanks for the demonstration. I line the layout of the PCB. I wonder how much of it is based on similation and how much on experience of the layouter

[@KerryWongBlog]

Don't know for sure in this case, but nowadays most of the RF PCB designs rely heavily on simulations.

[@ElektronikLabor]

Am I misunderstanding the specifications or should the unit in the first line be "MHz" instead of "GHz"?

[@Bernie_the_Bolt]

Yes, it should be 500MHz or 0.5GHz

[@pchunter168]

👍

[@JouMxyzptlk]

Maybe you have an X-Ray friend so you can look on the other side that way?

[@zyghom]

I am very surprised you did not tear it down properly - that must have been first time on this channel. very disappointed I am now ;-(

[@KerryWongBlog]

Sorry, didn't take it out as I did want to desolder the SMA connectors.