Prior to the measurements I calibrated each VNA using the terminations that are included with each - SMA's for the nVNA; N-types for the FieldFox. I had to use two additional adapters (not necessary with the nVNA) to connect the filter to the FieldFox. These probably added 0.1 to 0.2 dB to the readings on that piece of equipment.
The device used as the Guinea pig was the ADS-B (1090 MHz) bandpass filter from FlightAware. Its advertised specifications state that it "passes 980MHz - 1150MHz. Impedance is 50 ohms and insertion loss is <2.5dB".
Here are the results of each VNA:
.As can be seen, the nVNA holds its own quite well for a $60 device. The FieldFox costs about $12,000.
The differences between the two don't show up until 40dB down. At that point, and below, the nVNA indicates a broadening of the skirt not seen on the FieldFox until 55dB down.
This is due to the limited dynamic range of the nVNA compared to the FF. I can't really think of a ham-related application of a VNA where this limit of dynamic range would matter. If we're looking at filters or trying to determine phasing of how to feed an antenna with multiple driven elements, 40 dB is plenty of dynamic range to see what we need to know.
Also, this is a difference in dynamic range of 15 dB at 1 GHz. It's likely to be less - maybe much less - at HF frequencies where most of us hams live and play. I'll find out soon and post the results.
Needless to say, I am very impressed at the nVNA's performance. Even if it cost several times more, it would still represent a fantastic value.
In future postings, besides the HF dynamic range comparison mentioned above, I plan to test the device's accuracy as a TDR and and will also look at several ways of accessing the data and controlling the nVNA.