RF and electronics blog LA3PNA

NanoVNA linearity

Over the past few months, considerable work has been done with the NanoVNA, an excellent compact vector network analyzer that performs well up to around 300 MHz. However, the available dynamic range is somewhat limited, comparable to that of an HP8410 at 10 GHz. The following measurements were performed to evaluate an original NanoVNA-H4.

One challenge encountered was when measuring amplifiers. For a portion of the amplifier’s range, there was clear evidence of compression. But was it the VNA, the amplifier, or some other factor causing it? This led to the development of a setup to measure the compression point of the NanoVNA.

The setup included the NanoVNA, calibrated using a SOLT (Short, Open, Load, Through) procedure, with the shortest possible coaxial cable to ensure a 0 dB S21 response. A proper amplifier (Mini-Circuits ZFL-1000H) was used, terminated with a 6 dB attenuator, followed by a programmable attenuator. The amplifier’s output was measured at a single frequency while adjusting the attenuator. This ensured that the input level to the amplifier and the impedance remained consistent across all measurements, minimizing inaccuracies.

A Python script controlled the process, automating the reading of the NanoVNA, adjusting the attenuator, and taking new measurements. The NanoVNA’s power mode was set to “AUTO” during these measurements, as it was found to provide more consistent power compared to manual output settings.

The results are as follows:

The data is plotted with the attenuator value on the X-axis and the measured output on the Y-axis. At HF frequencies, the results were satisfactory. It is generally safe to measure an amplifier with up to 10 dB gain connected directly to the VNA, but for amplifiers with higher gain, either the VNA’s output power should be reduced, or an external attenuator should be used on port 2. Above 200 MHz, there was little change in the compression point until around 300 MHz, where the VNA begins using the third harmonic for detection. As expected, this resulted in a lower compression point.

The disappointing part occurred above 600 MHz, where the VNA relies on the fifth harmonic. Here, the compression point decreased with increasing frequency. Interestingly, there was a specific power level at which the reported output was -24 dB. This value was consistently observed, and after multiple tests with different amplifiers and attenuators, it was determined to be an actual response of the NanoVNA, not a measurement artifact.

For those using the NanoVNA to measure amplifiers, it is recommended to proceed with caution when dealing with amplifiers that have high gain. Be sure to adjust the VNA’s output power or include external attenuation to avoid distortion in the measurements.

Transverter interfacing – Flex

I like operating on the VHF and UHF bands those few openings we have here in the north. With my new Flex radio, I decided that the time was in for designing a whole section of transverters to cover the bands 70MHz-1296MHz.

The criteria for the transverters should be that they should be “State of the art”, with preformance that are such that the HF transceiver used as the driver will be the limiting factor. Noise figures in the sub 3dB range and power around 20W, should be suitable for driving the power amplifiers.

The transverters will all be selected by the radio, via the ethernet connection in the transverter. When the transverter has ethernet connection, it connects to the transceiver and subscribes to status packages that tell somthing about what frequency the transceiver listens to. If this frequency equals the operating frequency of the transverter, then the transverter is enabled and the transverter is connected into the IF bus, and the PTT from transceiver is enabled.

The IF bus has to mechanicaly disconnect from the IF bus when the transverter is not in use. This is done so that the input attenuators in the transverters not in use does not load down the bus line, and allows this to function as a RF selected bus. 10MHz frequency reference goes from a distribution amplifier to each of the transverters and ethernet goes from each transverter to a ethernet switch.

The ideal approach would be if the radio had one connection for each transverter, and a SCU (that means 15 connections for the setup here), so some switching is needed. The downside to the current approach is that its going to be impossible to transmit on one band while receiving on another. So for now, satelite operations are going to require some creativity.

Curve tracer christmas calendar 24 – NPN Transistor

Last day this year will be a regular BC547 transistor. I hope you have liked this little advent calendar, its been fun to figure out all of the devices to measure.