Radio amateurs making measurement tend to just believe any figures they see. This leads to various wildly incorrect interpretations of the data. Antenna testing in particular is prone to all sorts of errors. To understand what readings mean, the errors must be quantified and investigated. This is an experiment carried out to compare two identical small transmitting loops. It was carried out on 22-April-2017. Each antenna was connected to a WSPRlite with new firmware and the results were analysed on DXplorer using a proprietary technique (not available to other users at this time). The antennas were about 8 metres apart and orientated so that each sat in the predicted null of the other.

Tests were carried out on 30m for about 3.5 hours. The results are tabulated below.

The average difference in SNR was 0.12dB.

The system was then set to 40m where tests were carried out for 90 minutes. The results are tabulated below.

The average difference in SNR was 0.17dB.

The suburban garden used for these tests is far from perfect as an antenna test range, but notwithstanding that, the results give some cause for optimism. This is all the more so as they duplicate a much larger study done by another tester. More testing is required.