DXplorer.net opens up a new way to measure antenna performance. 3dB is the smallest change that is generally considered as being audibly detectable on a standard communications circuit and so I wondered how well DXplorer.net would work at detecting this change under controlled conditions.

Two antennas were used for this experiment: 

• a doublet, 43m long and 10m up matched with an MFJ-993B
• a Chamelon P-Loop "magnetic loop antenna" set up indoors close to a south-facing window.

The tests were carried out on 40m. 

The WSPRlite on the P-loop was set to 200 mW output power. The WSPRlite on the doublet was set to either 10mW (-13dB) or 5mW (-16dB). Note that as the antennas have different radiation patterns the relative dB measurements are not a true measure of the differences between the antennas however for this experiment it will not matter.

A total of 10 x 24 hour tests were run, 5 tests were at 10mW and 5 tests were at 5mW. The mean DX10 differences between the antennas was recorder for each test period. Both WSPRlite test transmitters were set to a 50% repeat rate.

A typical 24 hour graph is shown below.

TBA

The following tables shows the Mean and Maximum DX10 metric shown as a percentage of the maximum possible DX10 score. The yellow columns are the differences between the antennas.

Figure 1. Test results at 5mW

Figure 2. Test results at 10mW

Discussion

Figure 1 (5mW) shows that the Mean difference was always negative and the Max difference was either negative or equal.

Figure 2 (10mW) shows the the mean difference was mostly positive or equal while the Max difference was either equal or negative.

Comparing the tables the difference between 5mW and 10mW is clear.

Conclusion

The DX10 mean metric can detect a 3dB difference in power levels during 24 hour test periods. It would be hard to measure this difference across various wave angles and distances in any other way.