https://www.w0qe.com/SimSmith.html

The “RF Demo Kit” for testing NanoVNA has no information about the component values, which make it less useful for testing and practical demonstration of a VNA. I used my Der EE DE-5000 LCR-meter to find the values, as this instrument can measure L, R and C even when mounted in a complex circuit. Also this instrument use Kelvin-measurements, with 4wires to the probe tips. Inductance, capacitance and resistance in the test leads will therefore not have any influence on the test results. My “Demo Kit” is the type with a 6.5MHz BSF and 433MHz BPF. There is another version too on the market, with 10.7MHz BPF, and probably other values for the components. The results are below:

LPF30: 195pF/0.8uH/195pF, HPF100: 170nH/C/170nH (not able to get the value for C here, but I wouldn’t be surprised if it is 100pF), 7: 100pF, 8: 760nH, 9: 100pF/50 ohm, 10: 225nH/100pF, 11: 100pF/220nH/50 ohm, and 12: 50 ohm/220nH/100pF.

This is a version of the RF Demo Board. This one uses SMA connectors The RF Demo board uses a 0.032″ board.

Besides the well known SOLT calibration method (or TOSM in Rohde & Schwarz terms) there are a number of other ways to achieve a full two-port calibration of a vector network analyzer (VNA). One of them is the thru-reflect-line calibration (TRL). As its name implies, TRL uses a through standard, a reflect standard, and one or more transmission lines of appropriate length for the intended frequency range to determine the error terms of the underlying error model, which is somewhat different than the model commonly used for TOSM (but can be mapped to the traditional and more comprehensive error model if some additional quantities, the switch terms, are known).

In TOSM, the match standard is instrumental for calibrating to system impedance. This role is played by the impedance of the line(s) in TRL.

TRL has the drawback that since the phase shift of the line standard must be within certain limits, it only works over a limited frequency band, and it cannot be sensibly used with very low frequencies (but can be extended to low frequencies by other means).

The TRL board

The most simple way to manufacture a test fixture for components like SMD capacitors, inductors, ferrite beads, etc., is to realize it on an ordinary PCB that can be produced with one of the many inexpensive PCB pool services. This allows putting the necessary TRL calibration standards on the board as well. Certainly this does not yield the most accurate test fixture, in particular if the board is ordinary FR4 and one saves on an impedance controlled service, but it has the advantage that the components are tested in their natural environment.

The TRL board that I have designed can be seen in the above picture. It contains the following features for calibration, verification, and testing:

A thru standard for TRL
A 50 Ohms match standard
A 25 Ohms mismatch standard for verification
An open standard for verification
A short standard, used as a reflect standard for TRL
A line of 45 mm length, used as a line standard for TRL
A series fixture for 0805 components
A shunt fixture for 0805 components

The board is designed to be usable up to about 2 GHz (it fails to achieve that, see below).