RF transmission testing inside a Faraday cage

Hi Ralph,

For transmission testing, just a spectrum analyser type of tool can be used (with appropriate attenuation), and a coax would be connected from the 50 ohm output (i.e. not solder any antenna matching circuit). This provides the information on the power level without any antenna connected. No large enclosure is needed, because the entire transmitted signal goes to the attenuator/spectrum analyzer (the small transmitter board itself may have a shield, but it won't affect the measurement because the spectrum analyzer will only be sensitive to what arrives via the coax).

For antenna testing, a literally field measurement could be done; not super-accurate, but better than nothing, useful for comparing antennas (rather than precise absolute measurements), since an antenna chamber is (guessing) expensive even for most firms (most organizations pay external consultancies to do their antenna work for them, as far as I'm aware).
For the outdoor field method, any field could be used, e.g. large garden or park. For an example of the type of measurement possible, I attempted in a small garden using 2.4 GHz, here (near the end - see the section 'measuring radiation pattern': /products/roadtest/rv/roadtest_reviews/442/molex_24ghz_5ghz_ant_4

A large park should be way better though. 
I used a log periodic antenna to perform the measurement, and an SDR (not as accurate as a proper spectrum analyzer, but lower-cost). The log periodic antenna is low-cost from ebay.

Also, often WiFi/BLE chips have test modes, to simplify measurements, e.g. to just transmit at a spot frequency rather than across multiple channels, or to continuously transmit test patterns or data.

At a place where I once worked, we would use the outside field for testing (we had a license for that), for a different frequency band far lower than WiFi/BLE. It was handy for the occasional test. However all real testing was done with a spectrum analyzer or power meter, i.e. all done at 50 ohm, connected through coax.

One more approach (if SDR/spectrum analyzer is not available) useful for a more end-to-end test, is to simulate the transmission path loss only, by chaining loads of attenuators together, until you have a signal level realistic for the receiver, and then measure (say) received and demodulated signal quality, or bit error for digital transmissions, or any other metric, such as successful packets received (e.g. CRC being correct could be used). For such a test, again coax is used with 50 ohm connection from transmitter to receiver, but insert in a load of attenuation (realistically you may need 100 dB or more, and they usually come in various values up to 30 dB, so you'd need a chain of them).

Thanks Shabaz.

I'm trying to compare the transmission performance of a number of BLE boards in their normal use configuration, so I don't want to bypass the antenna.  On boards with an external antenna connection, doing a direct connection as you suggest would provide a good baseline.

I am looking at Direct Test Mode as it would allow me to automate tests, but it will require external control via UART.  I hope that I might get that working for a couple of boards.

Thanks Shabaz.

I'm trying to compare the transmission performance of a number of BLE boards in their normal use configuration, so I don't want to bypass the antenna.  On boards with an external antenna connection, doing a direct connection as you suggest would provide a good baseline.

I am looking at Direct Test Mode as it would allow me to automate tests, but it will require external control via UART.  I hope that I might get that working for a couple of boards.