Optimizing Advertising Power NRF52840

[robsrick]

Hi I'm trying to optimize the advertising current on my NRF52840EK. I'm fairly happy with the results I'm getting so far but I'd like to check and see if there's any other reductions I can make. 

This graph shows a zoomed in advertising pulse recorded using PPK2 and the next is a zoomed out view. The idle current between pulses is 7uA and the average is about 20uA.

Below is the power calculator with the correct interval which shows quite different results from my measurements:

{
  "chip": "1",
  "voltage": "3",
  "dcdc": "on",
  "lf_clock": "lfxo",
  "radio_tx": "0",
  "ble_type": "adv",
  "ble_int": "1000",
  "tx_size": "0"
}

so my questions are:

1) The calculator says i should be able to achieve 2uA idle but i'm getting 7uA. Any tips on where to look for this extra consumption?

2) The total average is listed as 11uA and i'm getting 20uA so even accounting for the difference in idle current of 5uA that's still an additional 4uA from advertising.

          2a) The peak current in the calculator is about half what i'm measuring (7mA vs 14mA) why could this be?  

          2b) the Duration of the advertising pulse is about 3.5ms in the calculator where i'm measuring about 6ms. Why could this be? 

I'm wondering if my settings in the calc aren't matching up to reality. For example i'm not really clear what my tx payload should be set to or how i work it out? 

below are some screenshots i pulled from nRF connect mobile:

Thanks for your help!

[Kenneth]

1) The calculator says i should be able to achieve 2uA idle but i'm getting 7uA. Any tips on where to look for this extra consumption?

Leakage current of some sort; gpios, level shifters, other circuitry etc. This current is temperature dependent, but I assume you are in room temperature.

2a) The peak current in the calculator is about half what i'm measuring (7mA vs 14mA) why could this be?  

The calculator show average during an event, so I expect there may be some variations here when measured on actual hardware.

2b) the Duration of the advertising pulse is about 3.5ms in the calculator where i'm measuring about 6ms. Why could this be? 

That is likely the reason why you measure higher current. There is an extra transmission period, likely because there is a central in proximity that is performing an scan request and receive a scan response:

Kenneth

[Kenneth]

Also I expect that you should increase TX payload to 30 in your case in the calculator:

[robsrick]

thanks. so i guess if i reduce the tx payload i'll reduce the current? 

the link to the generic access profile pdf in the app is broken but i guess this is the full data being sent through on a scan? can you link to an active copy of that pdf?