SMA pigtail cable and impedance measurements

[alftel]

Need community opinion. We have a supplier in China (Voton Machinery, voton.net.cn or rf-connectors.com) who we used in a past to procure SMA and RP-SMA connectors. This time we required SMA pigtails with UFL connector. After receiving cables and attching them to our device we noticed some sort of mis-behavior that could be explained only by impedance mismatch. We crafted UFL socket with soldered 50 Ohm 603 resistor and connected cable to a small antenna analyzer (hard to see on a picture but it is there). To our surprise readings were somewhat off the chart - between 2.4 GHz and 2.5 GHz (100 MHz span) impedance went from 34 Ohm to 61 Ohm, see picture below:

We had another two cables from other inexpensive sources, and they did produce much better results under the same test, even though they are mage with 1.13 cable, and ones that we received from Voton are made with 1.37 cable (supposed to be better). These are readings from another two inexpensive cables:

I wonder if it is usually OK to have this kind of impedance "jumps" and I am paranoiac, or there is definitely a problem with this particular product from this company (again, other two randomly picked cables from our lab bin did produce much better results as seen above). So, this is a cable spec:

Voton cable - SMA connector - UFL connector, 1.37 cable, 100mm long, test frequency 2.4 GHz - 2.5 GHz

Other no-name cable #1 - SMA connector - UFL connector, 1.13 cable, 180mm long, test frequency 2.4 GHz - 2.5 GHz

Other no-name cable #2 - SMA connector - UFL connector, 1.13 cable, 150mm long, test frequency 2.4 GHz - 2.5 GHz

Any opinion or shared experience will be greatly appreciated.

--- Updated Aug 10, 2020 ---

Just to be clear - we did procure SMA and RP-SMA board edge connectors from this source in a past and didn't have any problems (looks like). My understanding is that for 2.4-2.5 GHz terminating cable with 50 Ohm small resistor is suitable setup for a quick and dirty test that supposed to at least show cable behavior in impedance domain vs frequency (at least allows to relatively compare two cables side by side).

 

[volker@muehlhaus]

Can you show a sweep down to DC in the Smith chart?

 

[FvM]

Problem with the measurements is that each pigtail cable has a different length, respectively does different transformation of mismatch.
I fear, the simple instrument (AAI N1201SA) doesn't have a Smith chart display mode, but it can show X and R scan separately. Also the reference plane is at SMA port, not the pigtail connector unless a user calibration has been performed. No port extension function available in the software.

I would start with a resistance measurements of center conductor, shield and center to shield, also capacitance per length can be measured as a rough indication of correct Z0. Measure with SMA/UFL sockets connected as there may be a problem of contact finish.

 

[volker@muehlhaus]

I did a quick test here with cheap U.FL pigtain cables and a homemade 50 Ohm termination from a U.FL socket. Impedance up to 3GHz was 50 Ohm +/- 3 Ohm.

So your results are really bad in comparison. I agree to FvM that a low frequency check for resistance and capacitance might be worth a try. Also, double check your termination resistor value.

 

[alftel]

Thnaks a bunch to everyone for reply! Unfortunately my RF person is away together with equipment and AAI N1201SA indeed has nothing but a simple interface (such as Z, SVWR, S11, X), no Smith chart. The only thing that I attempted to compare was side-by-side cables from three different suppliers. Result presented by Volker (thanks a lot!) is exactly what I would expect from decent quality cable. My termination resistor is 49.9 Ohm/1%. I was a bit shocked by difference in readings. Would be interesting to re-run this test with Digikey or Mouser cable - not like one would expect them to be made "coneptually" different, but at least there some degree of certainty that part is of acceptable quality.

btw - few words about AAI N1201SA - I usually pop a hood of any gadget that I own, and even though this small gadget is considered "a chip made in China RF toy", inside are top quality Analog Devices components as RF part goes, and my RF partner at some point compared it's reading side-by-side vs Rohde & Schwarz and confirmed that it is a very decent unit for whatever it does.