I’ve got my miniVNA Tiny+ this summer and start making test with it. A VNA is a Vector Network Analyzer. Behind this dark name, this is a tool able to analyze the radio response of your circuit and ensure your circuit is tuned to the right frequency. From the data measured by this tool you can find the right adaptation circuit to match your central frequency.
For more information about RF circuit matching, you can read this other post.
The miniVNA Tiny+ is a low cost VNA solution less than 300€ covering frequencies from 1MHz to 3Ghz. These frequencies are good for most of IoT need : LPWAN around 868MHz and Bluetooth at 2,4GHz. The steps are 10Hz for a large precision. With two ports you can measure S11 – power transmitted and received over the same port – and S21 – power transmitted from port 1 and received over port 2.
The miniVNA Tiny+ is an usb solution working with a computer connected to and a software running on the computer. The software is based on Java and may support different Operating Systems.
The VNA is provided in a nice plastic case but the documentation is limited to a single A4 paper sheet with a link to download the software. So you have to follow this link to getting started. http://www.mculab.com/eastchip/minivnasoft.asp
At this point you get a second link to a JAR file to download.
The miniVNA+ uses a FTDI chip to communicate with the desktop over a serial transmission. You need to have the driver installed on your machine. The good news is that FTDI is really standard, common and existing in most OS.
This link is related to old version of the software. With my JVM 1.8 the jar terminated with an error.
The source of the software and the documentation is the VNAJ project. You can obtain it following this link : http://vnaj.dl2sba.com and download the version 3.1.19 here http://download.dl2sba.com/vnaj/3.1.19/
The next step is to calibrate the VNA, first of all you need to ensure the VNA Tiny is selected in the Analyzer / Configuration menu. Then go to the calibration menu and create a calibration. You have 3 steps that may be active :
- open (nothing connected on DUT & DET)
- 0 Ohm where Short is connected on DUT
- 50 Ohm where the unamed terminator is connected on DUT
You can increase the number of calibration step in the menu Analyzer/Info. Basically it is corresponding to the number of calibration point over the full spectrum in mode 2. In mode 1 you can do a calibration on a given range of frequency for a better precision. The mode 1 need to edit the file CalRanges_miniVNA Tiny.txt located in ~/vnaJ.3.1/preset
This is what I choose to do for a calibration focus on 865-876 Mhz by editing the file with the following content :
1000000 864999999 1000 1 865000000 875999999 5000 2 876000000 3000000000 1000 1
Note : the extra blank line is required. Format is :
Freq Start*Freq Stop*Sample*Oversampling
This calibration can be saved to be reused later.
The calibration can be done by directly putting the Open / Short / 50 Ohm termination on the miniVNA connectors but this will not take into account the cable and soldering net impact. I was expecting no big impact but in fact it seems there are some. So personally I do the calibration with the target connectivity in place.
The following graph is the result obtained on an antenna with a calibration made directly on the miniVNA connector (signal is not like a V) The following graph is the same antenna with a calibration made at the connector end
Signal is more like the expected V ; I’m hopping the result is more realistic than the previous one.
You need to connect the DUT output to the antenna connector and perform a sampling:
Select a frequency range to analyze, you can save it on the Preselection list.
Then click on Simple Sampling green button on bottom to get the data.
By changing the speed to a negative number you will have an oversampling.
The precision of the data sampled is depending on the speed choice and the window size. Larger is the window, more data will be read (more frequency step will be captured)
Then you get the data from the VNA and you can display the different element measured.
On the next picture you can see a result presenting the SWR Standing Wave Ratio for the system under test. A Marker is set on 868MHz
The SWR is 9:1 the following table gives an idea of the corresponding power loss:
With a such SWR the energy output to antenna is about 35% only… 65% has been lost due to non-matching.
You can also use RL indicator where the equivalent value is represented in dB.
To correct this matching, you need to calculate a matching circuit.
Calculate a matching circuit
To get the matching circuit corresponding to your design, you can read from the graph the impedance at the matching frequency with the two measures :
- Rs for the real load impedance
- Xs for the imaginary load impedance
Then you can enter these value in a tool like these one
- Analog Device RF Impedance Calculator tool
Indicating the target impedance and the target frequency. They give you different matching circuit you can use.
You can fin more tool with google search key : “antenna impedance matching network calculator”
A really good tool for optimizing your matching and easily see the direction your smith chart will evolved when changing the matching circuit is Atyune, a free and powerful tool.