nRF52840 receiver sensitivity specification discrepancy

RE: nRF52840 receiver sensitivity specification discrepancy

ketiljo — Thu, 22 Mar 2018 08:35:37 GMT

The Bluetooth specification have set of minimum requirements for the receiver. This is to ensure that the Bluetooth technology is only used on HW with good (enough) performance.

They say that the level when BER is measured to be 0.017% shall be < -70 dBm. We measure 0.01%, less than 0.017%, at our sensitivity limit, which is lower than -70 dBm. So all is good.

We define the sensitivity limit as the input level when PER = 0.01%

RE: nRF52840 receiver sensitivity specification discrepancy

dom — Wed, 21 Mar 2018 16:33:48 GMT

No, I think you are incorrect. The Bluetooth requirement is that actual sensitivity is better than -70 dBm for LE Uncoded PHYs. Actual sensitivity is defined by the Bluetooth spec as the receiver input level for which 0.017% BER is achieved for payload lengths >= 128 bytes.

My initial question (bolded above) still stands.

Here is the full sensitivity section in the bluetooth spec:

RE: nRF52840 receiver sensitivity specification discrepancy

ketiljo — Wed, 21 Mar 2018 11:09:57 GMT

The sensitivity limit is stated in the datasheet as for short packets, < 37 bytes, BER = 0.1 % and for long packets, BER = 0.01 %.

The Bluetooth definition is that BER shall be better than 0.017% at -70 dBm. This is not the same as the sensitivity limit in the datasheet.