Antenna showdown with my Nano VNA

RouterOS Wireless Networking
1

I just got the Nano VNA SAA v2 and started testing all the antennas I had.

First in line, the Asus one, the one that comes with the PCE-AX58BT. Shines the brightest on channel 11.
VSWR:
asus_log_mag.png
Log Mag:
asus_vswr.png
Impedance:
asus_impedance.png
Next, the Alfa 10 dBi panel, a beacon of hope on channel 1:
VSWR:
alfa_10db_vswr.png
Log Mag:
alfa_10dbi_impedance.png
Impedance:
alfa_10db_log_mag.png


Now time for the Mikrotik 6 dBi, the one I got with the hAP ax3. Marketed as the best in its price range.
VSWR:
Mikrotik_6dBi_vswr.png
Log Mag:
Mikrotik_6dBi_impedance.png
Impedance:
Mikrotik_6dBi_log_mag.png
Verdict: Best used as a Wi-Fi diet plan for your neighbor’s freeloading habits.


For the last contender, we have a 25 dBi Yagi I got from eBay. Touted as the zenith of antenna engineering, a beacon of hope promising the unattainable:
VSWR:
yagi_25dBi_impedance.png
Log Mag:
yagi_25dBi_log_mag.png
Impedance:
yagi_25dBi_vswr.png
Verdict: best used as a gift to your enemies.

I could not test 5 GHz as my VNA is rated only up to 3 GHz.

PS. If you want to buy the Nano VNA, make sure to buy a RP-SMA adapter or you won’t be able to connect your antenna.

2

I conducted some RSSI measurements with the antennas I have. The Yagi antenna has already been discarded.

  • Distance: 1.2 meters between the Access Point (AP) and the phone, with a clear line of sight. The biggest antenna is 0.16 meters so we are well further than the calculated far field of 1 meter (https://www.everythingrf.com/rf-calculators/antenna-near-field-distance-calculator).
  • The phone was positioned directly in front of the antenna.
  • AP: Mikrotik hAP ax3.
  • Phone: Samsung S7.
  • The phone has 1 RX and 1 TX chain, while the AP has 2 RX and 2 TX chains.
  • Measurements were taken on the phone using WiFi Analyzer and were averaged out.
  • Measurements on the AP were done using self-reported values in WinBox.
  • When measuring 5 GHz, the 2 GHz band was turned off, and vice versa.
  • All antennas were positioned at 90 degrees relative to each other.
  • “P” stands for phone, and “AP” stands for Access Point.

Here’s the table of measurements:

+------------+--------+---------+--------+---------+------------+-------------+
| Channel    | P-Asus | AP-Asus | P-Alfa | AP-Alfa | P-Mikrotik | AP-Mikrotik |
+------------+--------+---------+--------+---------+------------+-------------+
| 1 (2412)   | -54    | -40     | -51    | -36     | -62        | -51         |
+------------+--------+---------+--------+---------+------------+-------------+
| 6 (2437)   | -54    | -45     | -50    | -37     | -61        | -49         |
+------------+--------+---------+--------+---------+------------+-------------+
| 11 (2462)  | -52    | -39     | -53    | -38     | -60        | -50         |
+------------+--------+---------+--------+---------+------------+-------------+
| 36 (5180)  | -68    | -52     | -69    | -53     | -78        | -61         |
+------------+--------+---------+--------+---------+------------+-------------+
| 48 (5240)  | -76    | -53     | -63    | -44     | -73        | -52         |
+------------+--------+---------+--------+---------+------------+-------------+
| 52 (5260)  | -75    | -51     | -60    | -43     | -66        | -48         |
+------------+--------+---------+--------+---------+------------+-------------+
| 100 (5500) | -77    | -49     | -72    | -43     | -73        | -45         |
+------------+--------+---------+--------+---------+------------+-------------+
| 140 (5700) | -67    | -45     | -65    | -48     | -68        | -48         |
+------------+--------+---------+--------+---------+------------+-------------+

To-do:

  • Perform speed tests for each frequency and antenna, as RSSI may not fully reflect performance, especially at a distance.
  • Test different antenna configurations, such as positioning them parallel to each other, and at 30, 45, and 60 degrees.
  • Consider testing with two Mikrotik APs, using one as a client to eliminate the hardware and antenna of the client device as variables.

Notes:

  • Ultimately, placement and the environment will also influence the values, so I wonder if it makes any sense to seek “laboratory-grade” measurements.
  • These values alone cannot definitively determine which antenna is best, as they all have different gain values. For example, the Alfa 10 dBi was expected to perform the best given its higher gain.
  • This data can be useful in determining the best channel for a given antenna, but in-situ measurements are still necessary.

Now lets open a can of worms. I experimented with chain configurations. It seems that setting only 1 TX and 1 RX chain gives a loss on the 2 GHz band and a gain on the 5 GHz band.

+-----------+------------------+------------------+
|           | 1 TX 1 RX        | 2 TX 2 RX        |
+-----------+--------+---------+--------+---------+
| Channel   | P-Alfa | AP-Alfa | P-Alfa | AP-Alfa |
+-----------+--------+---------+--------+---------+
| 11 (2462) | -63    | -51     | -53    | -38     |
+-----------+--------+---------+--------+---------+
| 48 (5240) | -45    | -32     | -63    | -44     |
+-----------+--------+---------+--------+---------+

Well, this opens more questions than it answers.

3

When using polarized antennas (they all are) it’s important to perform measurements when polarization planes of both antennas match exactly.

If using 2 chains on one side and both antennas are at some angle (ideally at 90° angle), then they’ll both contribute to reception even if the other party only uses one chain (and antenna). This will to large degree elimitate the issue of polarization plane misalignment.
If both link partners are using single chain / antenna, misalignment causes big problems (which can explain the large signal level drop in your second table). So when doing these tests, make your best to align polarization planes … either by turning phone to find position with best signal levels or by aligning AP antenna (if it’s adjustable). It shouldn’t matter which end you’re using to display signal level, antenna polarization is the same for both Tx and Rx.

Measurements also depend on environment (e.g. amount of reflections off room walls, which can either improve or degrade signal levels) and the receiver’s DSP performance. There’s a reason for using testing chambers wgen measuring antenna performance, they are engineered to absorb stray RF energy rather than reflect it.