We are range testing BLE using the DK-52 with an external dipole antenna connected via the u.fl connector. We are using the DK-52 + packet sniffer software on a laptop to observe RSSI as we move a nRF51822 smart beacon away. Tests are open field on tripods, so no external influences.
We see a significant drop in RSSI over the first few meters down to around -60dBm, and then a more gradual drop out to around 50m where we see around -90dBm. We also see a more "lumpy" response over distance than we would expect from a dipole.
We need to use RSSI to approximate distance from the receiver, and so the loss of the first portion of the receiver sensitivity -20dBm to -60dBm represents a problem with dynamic range. We are left with -60dBm to -90dBm to represent 2m to 50m distance which isn't enough resolution.
See attached graph with range test results.RSSI vs Distance.pdf
There doesn't seem to be a manual for the DK-52, but we think the u.fl connector is matched to 50ohm. As a dipole is 73ohm impedance, we are wondering if the lack of match could be responsible for the significant loss of RSSI from -20dBm to -60dBm over the first 2 meters distance away from the antenna?
If this is the case, can you recommend component values for DK-52 components L1 & C3 to match to a dipole?
Yes, the u.fl connector is placed in the 50 Ohm reference point. If the dipole has an impedance of 70 Ohm, there will be a mismatch.
Regarding the RSSI values: The RSSI values are given in dB, which is not a linear "scale". As a rule of thumb, adding 6dB to the output power doubles the range. The same applies the other way around. For range testing, each 6dB lower RSSI, means a double range. Therefore, the RSSI values will decrease a lot in the beginning and then slower and slower.