RG58C/U cable on 50 Ohm load with VSWR flactuations

[cplia]

Some really basic measurement experiment with my MiniVNA Tiny analyzer.

After careful calibration of the VNA, I wanted to check how the input impedance up to 500MHz of a 4m long RG58 cable, terminated on a 50 Ohm dummy load varies. To my surprise I noticed that the SWR response was not flat at all. Actually it was fluctuating so much that I've repeated calibrating my VNA again and checking connectors for short/open circuits etc. Still same response. (response attached on this post). Dummy load is relatively accurate up to 1GHz. When I connect the same dummy load directly to the VNA port the SWR is pretty flat, so obviously something is happening with the cable. But, the cable is brand new, checked for open/shorts as well - all, seemed ok!

Is it really that even minor imperfections along the coaxial cable, the dummy load or the connectors that make the SWR/Zin jump up and down to that extend? I was expecting a flat response at least at the low end of the frequency spectrum.

73,
Kostas

Setup:
|VNA| <-> |4m RG58| <-> |50 Ohm load|

 

[Aussie Susan]

Looking at the specs for RG58, the signal velocity is about 65.9% the speed of light or approx 2e8m/s.
Therefore the 4m of cable will be resonant every 2e8/4 Hz or about 49.4MHz.
If you look at your graph, you will see that there are dips in the SWR at about event 50MHz (actually just a little bit less which the above would tell you is correct).
Therefore what you are seeing is the way the cable reacts when the frequency is a multiple of 1/2 the resonant frequency.
The gradual decrease in Z would also be expected as the frequency increases due to the capacitance of about 60pF/m.
Susan

 

[cplia]

Ok, I got your point now. The dips appear every about 25MHz or so, which is near the λeff/2 of the cable as you mentioned, a condition which is known to force the value of the load (i.e. 50 Ohm) to 'appear' at the input.

So, these fluctuations on the response clearly indicate that the system isn't exactly 50 Ohm uniform from the input port to the load across the measured frequencies.

Kostas

 

[FvM]

So, these fluctuations on the response clearly indicate that the system isn't exactly 50 Ohm uniform from the input port to the load across the measured frequencies.

Something is completely wrong with the measurement. I don't see more than 1.2:1 SWR in a similar setup.

The huge impedance ripple shows a system far off from impedance matching, either a defective VNA or dummy load, wrong calibration or connector mounting fault.

 

[cplia]

I really can't figure out what's going on with this wide SWR flactuation across the spectrum.

I have repeated the measurement again this time, but with a 1.5m long RG174 cable. One end of this cable is attached on the VNA port (MiniVNA Tiny), the other end is terminated on a 50 Ohm Diamond load (specs up to 1GHz) through a SMA to UHF-F connector. Attached are the results of this simple measurement up to 500MHz.

The weird think, is that when I am checking the SWR with the 50 Ohm load directly attached to the VNA port (without any cable), the response is flat, which is an indication of proper calibration with the 50 Ohm load.
Again, as before with the 4m long RG58 cable, the RL(dB) across the frequency, shows some deeps every λeff/2.

I can only conclude the the RG58 and RG174 cables don't have a uniform impedance value along their lengths and this is pronounced especially above the VHF band?

 

[FvM]

I can only conclude the the RG58 and RG174 cables don't have a uniform impedance value along their lengths and this is pronounced especially above the VHF band?

If we can sort out defective dummy load, there are still the points mentioned in post #4. Unlikely that it's the cable, I think.

 

[cplia]

If we can sort out defective dummy load, there are still the points mentioned in post #4. Unlikely that it's the cable, I think.

After many hours of measurements, I finally discovered that the adapters I've used to connect the VNA with the cable (SMA-UHF) were the cause of some serious impedance mismatch that was generating all these terrible ripples on the VSWR response. The extend of the problem was so scary that I've decided to write an article about this on my blog ( **broken link removed** ), so people can be aware of this 'innocent' but disastrous problem. It was really frustrating to figure out this cause, although now obvious,, but finally it came to an end. I now feel that I want to throw in the bin all my adapters and get new ones from trustworthy suppliers.