Forgive my initial response as I was not familiar with the current numbers published within the WICED Smart Sleep/Deep-Sleep Explanation and Techniques Design Guide
This is a newer document that I am not as familiar with...
I am checking with the author of that document now to confirm both the operating voltage used during those measurements and whether or not an external DC-DC was leveraged. At this point I am assuming these tests leveraged a different test setup as compared to the datasheet.
Related to this topic, it's important to understand the following about the current values provided in both datasheets (SoC and Module).
1.
All datasheet values apply to the SoC alone, which is implemented as raw die inside of the SIP module. So in addition to chip current, there would also be some added current from the other internal components inside the SIP module (EEPROM, Xtal).
2.
As stated in the datasheets for both SoC and SIP module:
Note: With an optional DC-DC (90% efficient) converter at 3V, the current between the battery terminals (shown in Table 5) are about 50% lower than their nominal values at the default operating voltage (1.62V).
So both 3V and an external DC-DC are ideal circumstances.
No DC-DC, then expect higher current, up to 50% more than what's published.
Today, there is no SIP module with external DC-DC capabilites, but that will change with the BCM20737L, which will be released most likely in the upcoming quarter.
2a.
3V is also optimal for lower current since reductions in supply voltage yield increases in current.
3.
As boont mentions, the TAG3 is not an optimal tool to measure current for a variety of technical reasons (i.e. bad GND plane, lacks a means to minimize current seen from periperhals, etc.). With that said, there is a new TAG board due to be released in Q3 that addresses these shortcomings and customers will be able to use it for accurate current measurements.
