My first RF: directional coupler. Impressed but issues at DC

[240RS]

For the first time I measured 'forward' and 'reflected' power. I took this component from Mini Circuits:
https://eu.mouser.com/datasheet/2/1030/ADC-20-4_2b-1700033.pdf

I am interested measuring at HF from 2 to 30MHz. I used a signal generator sweeping from DC to 30MHz at 0dBm and as load I used a 50 Ohm termination.

I soldered in metal housing with three SMA connectors and was blown away to see the result (this being my first RF soldering). Attached a screenshot of the measurement:
- blue line: loss at the main line (less than a dB)
- purple line: coupled signal in forward direction(-20dB)
- green line: coupled signal with reversed connections of the signal generator and 50 Ohm load

Seeing the purple line so flat from 2-30MHz impressed me.

The green line I interpret as the sum of "Coupling" and "Directivity" from the datasheet. But please let me know if that is correct. It is around -44dB and the sum doesn't really match then. If I take off the load, the green line coincides with the purple line which I interpret as "all power is reflected".

But here comes my DC issue: The real HF signal I want to measure is superimposed on a couple of volt DC and a 1kHz 24Vpp signal. I had not noted the datasheet that states that DC signals are shorted and I could measure that indeed. See the second screenshot for the measurement at low frequencies.

I was hoping I could just pass the DC, 1kHz and RF in the main line and then measure the RF signal and its reflections. I trust it is safe to assume I selected the wrong component. So my next question is: Is there a component that would just pass DC and lets me measure forward and reflected RF power at HF frequencies?

Another question I have is what how the mini-circuits part names work. This one had 20-4+ in it. I understand the 20 being the coupling, but what does the 4 and the plus sign mean?

[SuzyC]

The +Suffix identifies RoHS Compliance. See top right corner of datasheet .pdf
The '4' indicates a specific packaging/model id.
You can isolate the ground of the 'DC connection grounded' just by inserting a HF smd coupling capacitor.

Whether you intend to multiplex DC power on your connections will only affect the DirCoup if the current is high enough to affect the magnetic characteristics the core used to make the DirCoup, else the DC cur be so much as to cause the small diam wires used to overheat and shortout/open circuit, otherwise your measurements without introducing a DC component should hold steady when multiplexing DC. Good Luck with that.

Testing with/without DC bias will show if  you are getting the DirCoup device to work correctly.

It is usually considered bad design practice to multiplex power on a RF circuit connector because of electrolysis resulting in failures of connectors, as well as the above caveats.

[langwadt]

The +Suffix identifies RoHS Compliance. See top left corner of datasheet .pdf
The '4' indicates packaging/model id.
You can isolate the ground of the 'DC connection grounded' just by inserting a HF smd coupling capacitor.

assuming it looks something like this https://upload.wikimedia.org/wikipedia/en/e/e6/Transformer_directional_coupler.svg
the 50R fwd/rev loads needs AC coupling too

[240RS]

Not sure if I understood whether to put a capacitor between the directional coupler and ground. I put a 1.5nF between the GND terminal of the directional coupler and true ground and ended up with a bit of a mess as pictured:
- blue trace is main line and has attenuation now. Short to ground is gone.
- purple and green are not that useful anymore.

Did I do as intended and did the idea fail? Is there another type of directional coupler that would be better suited for the multiplexed use case (not my idea, it is a powerline communication standard used for vehicle charging)?

Thanks!

PS: I didn’t have SMD capacitors, just these through hole ones. SMD significantly different for RF?

[langwadt]

try 100nF, both on ground and the fwd/rev port

[240RS]

try 100nF, both on ground and the fwd/rev port

Yes, will try. Could you explain what to connect to what?

[langwadt]

capacitor in the ground and in series with the 50R resistor and the 50R scope

[MathWizard]

Not sure if I understood whether to put a capacitor between the directional coupler and ground. I put a 1.5nF between the GND terminal of the directional coupler and true ground and ended up with a bit of a mess as pictured:
- blue trace is main line and has attenuation now. Short to ground is gone.
- purple and green are not that useful anymore.

Did I do as intended and did the idea fail? Is there another type of directional coupler that would be better suited for the multiplexed use case (not my idea, it is a powerline communication standard used for vehicle charging)?

Thanks!

PS: I didn’t have SMD capacitors, just these through hole ones. SMD significantly different for RF?

So how expensive is that little box and connectors ? It looks nice.

[240RS]

It was an old directional coupler with smaller than SMA connectors on the other side from eBay. Adaptors for that were really expensive so I added two SMA’s to this side (I could not open the back half) and put in the micro circuits coupler. Which is wrong because of DC short (adding 100nF capacitor will violate my other constraint of <1us rise/fall at 1kHz square with 1000Ohm source and load impedance).