NanoVNA V2: Female Calibration Kit vs. Port Extensions 

instablaster.

It's a USB power bank that I use to run the NanoVNA V2. There is a spot on the board for a 1s lipo with a balance style plug, but I never installed one, so I just use external power.

The calibration numbers are 0,1,2,3. They refer to the calibrations that were all done on screen in the early sections of the video. Each section title states what was done for the calibration, and there is a section of the video showing the exact process for each. The spreadsheet was showing that all the measurements from each calibration were different except for cal 2 and cal 3 which were pretty much the same. The colors are all labeled by calibration in the plot legends. The magnitude and phase of the S-parameter (Scattering Parameter) S11 are being analyzed in the excel sheets to see if they would be the same. Scattering parameters are the actual vectors that a VNA (Vector Network Analyzer) measures. The core concept to understand here is that of the measurement plane. The measurement plane is the point in your cables/connectors where you perform your calibration. If the calibration is performed where you connect the device you will measure, then your measurements should be correct. If not, your measurements will be off by some phase if the load is matched to the characteristic impedance of the transmission line between the measurement plane and the DUT or some magnitude and phase if the load is unmatched to the transmission line characteristic impedance between the measurement plane and the DUT. Using an SMA (M) to SMA (F) cable on port 0 would allow you to use the male calibration kit, but it's the same as just using the stock cable with the SMA F to F adapter. You will need a way to connect to your DUT at the SMA F connector where your measurement plane lies. The point of this video was to demonstrate that at high frequencies (L-band and above) the difference between using and SMA adapter with the stock calibration kit and a calibration kit with the correct connectors is substantial, and the built in time delay function to de-embed the circuit or embed it depending on your perspective is inadequate if applied haphazardly by eyeballing it. For microwave frequencies the calibration kit provided by the nanoVNA V2 makers is inadequate if you don't intend to connect the DUT to exactly the plane of calibration. You might get a reasonable S11 magnitude measurement anyway (equivalent to VSWR mathematically), but the vector aspect of the measurement will be lost which can be seen in the huge phase differences between approaches.

@@RobertResearchRadios I think that I still don't quite understand what the phase measurement is. Is it between the forward and reflected voltages in the s11 measurements? Why is the phase waveform shaped like a sawtooth? I was wondering if you could just measure the characteristics of the adapter and then subtract it out of your readings. I would like to find a good set of instructions for the features of the vna v2. Thank you for your help.

The phase is the phase of the s11 complex number (the angle between the real and imaginary components). S11 describes the power incident at port 1 as a result of the power transmitted from port 1. It's related to voltage, but it's actually a measurement of power available to a load given a particular normalization impedance (usually 50 ohms). The phase waveform is shaped like a sawtooth because the phase (think of it like a delay in time) continually increases as the signal propagates, but the display wraps it around the 180 deg mark to only span angles from -180 deg to +180 deg. If you are unfamiliar with s-parameters I really suggest taking a look at other tutorials specifically on what they mean, because this video was more about addressing the usefulness of a specific feature of the nanoVNA that explaining what was happening in that regard. w2aew's channel may be a good place to look, but i don't know a series that i would recommend offhand on scattering parameters. The wikipedia page is a little dense, but it also might be a good place to start. Measuring the characteristics of the network leading up the DUT and subtracting them out is a process known as de-embedding the DUT. It's perfectly possible with the right methods, but the purpose here was to see if the characteristics of the female to female SMA adapter could be estimated as just an electrical length (by time delay) using the time delay function of the nanoVNA. The reality is that the connector probably has some unaccounted for losses and the estimation method used (a purely experimental one) wasn't accurate enough to yield the right results. If you look at the DIY L-band LNAs video in this series I mention that we used a de-embedding technique in class to measure the s-parameters at the leads of the transistor, but that was successful because we accounted for losses in the microstrip transmission line on the board leading to the transistor. It was not as simple as applying a port extension by electrical delay (- or + as needed).

I understand your concern, but the real crux of what I was investigating here was "can the electrical delay feature be used to obtain nearly the same results as calibrating with a different measurement plane (e.g. female vs. male cal kit)?" The answer to that question appears to be no, but it may also be worthwhile to investigate your concern about deviations between calibrations with the same setup. Certainly, there is some measurement noise that will cause deviations, but I expect it to be in single digits of phase in degrees and hundredths of a dB in magnitude. If the deviation between calibrations of the device with the same setup were a phase difference on par with that between the cals in this video, the device would be basically useless for any meaningful measurements. If, for example, everytime i calibrate amd measure a patch antenna i get an impedance measurement rotated randomly within 90ish deg (1/2 smith chart) I'm not going to be able to use my S11 measurements to match to that antenna with any precision at all. I might well just guess at a matching with only a VSWR meter as opposed to a VNA at that point. Since, I've used the device for this purpose quite successfully, I conclude this is likely not the case.

There is no better way to calibrate. In the world of VNAs a SOL or SOLT calibration is required for each new measurement setup for best results. The nanoVNA (V1) automatically recalls Cal 0 on boot and some revisions of its firmware do interpolation if you say change the frequency sweep after the calibration, but this leads to users being uneducated on the subject of when a new calibration SHOULD be done vs. when it MUST be done given the firmware of the device. I personally don't feel the device needs any more updates as far as calibration goes. Regarding a better way to recall data, I'm not sure what you mean since there is currently no way to recall "data" as in measurements on the device. Though I would like a way to save sweep data points on the device without having to connect to a computer, so i could record an s-param file in the field and then extract it with a computer later.

Yes, the default firmware supports flipping the screen; I use this feature in some of the later videos for convenience.

Thanks for answering. I bought this device, it's a pity that when it turns off, it forgets that the screen is turned upside down

It should remember when you recall a calibration with that setting in it, but yes, the current nanoVNA V2 firmware does not auto-recall a calibration (and other associated settings) on boot. You could probably track down the devs and make a feature request, but I personally like it this way in terms of not recalling a calibration on boot. Maybe you can ask specifically that they recall the screen orientation on boot instead.

@@dimsonk6811 It does remember it if you save it into the bootup position (zero).