NRF52 transmit power measurement

[Karthikeyan]

Hi,

 How to measure tx power of the nrf52 BLE? Is there a proper/recommended way to measure it?

Regards,

Karthik

[ketiljo]

Use the functions in the radio test example in the SDK. Setup a constant carrier and measure the output power with a spectrum analyzer. Alternatively, you can setup a constantly modulated carrier and measure the channel power using the spectrum analyzer.

[Karthikeyan]

Am doing an automation project, where I don't want to use spectrum analyzer. Is there a way we can measure without it, also cost effectively?

[AmbystomaLabs]

Functionally you can measure the power with a power meter (about $2,000) but you will have to have a spectrum analyzer anyway in automated test to measure channel accuracy.

A spectrum analyzer is not considered a calibrated power measurement tool. They do have to be calibrated regularly to keep their error down to reasonable levels and to identify hardware issues but their accuracy even with a fresh cal is maybe +/-2dB and will be worse over the whole operating band.

In a test setup the spectrum analyzer is calibrated in situ with a calibrated power meter. In this manner you can accurately measure power over frequency.  Most spectrum analyzers have relatively poor clock accuracy.  Normally you provide a 10/13MHz GPS reference.  Then you end up with an auto test tool that can accurately measure power and frequency.

Care needs to be taken in how you couple to the DUT. We have been assuming a properly matched 50ohm test port on the DUT.  Most designs do not have a test port.  Then you have to measure to a coupled antenna. This brings with it the added complexity of:

1. calibrating your test antenna and path loss.

2. Shielding from ambient RF noise.

3. Making sure you are not performing a near field measurement.  Near field coupling is highly influenced by the antenna matches on both ends of the test setup and small differences that mean little in the field can produce huge changes in measured power output.

And, then cabling brings with it more problems.  All test setups are done with semi-rigid coax. Inexpensive flexible coax has poor phase/amplitude response when flexed.  With flexible coax one day you have passing devices, the next day 50% failure because someone pushed a cable on the bench.

If you look on the devzone some people have proposed using an nRF as the test equipment and then using RSSI as the power measurement tool.  RSSI is horribly inaccurate.  Maybe +/-3dB to +/-6dB and there is nothing you can do about it since it is not based on a calibrated power measurement device. Plus you still have no way to measure channel accuracy.

[ketiljo]

Look into power detectors, something like this: https://www.minicircuits.com/pdfs/ZX47-60+.pdf

[Karthikeyan]

Regarding power meter (about $2,000), which one do you suggest?