CC2650: increase the power consumption

I will check internally to see if this is a known issue. 

I checked with QA and we have not done any changes in the CC2650 production during this period and hence a change in power consumption is not expected.

- "They measure and plot the power consumption of the entire product before shipping the application.": 

- Do they just measure current or do they also measure other parameters as part of the production test? If they also measure other parameters, how is the variation on these over time?

- Is the production test always done with the same parameters (temperature, settings, firmware, same setup)?

- Has it been done any changes to the design?

- Could you give more details on the swap test? Was two devices selected, one from the start of the production ("A") and one of the latest (B") and the CC2650 from "A" was placed in "B" and the other way around? And the power consumption increases in "A" after the swap and decreases in "B" after the swap?  

Thank you for your reply. I will ask the customer in detail, so please wait a moment.

Hi TER-san,

I will talk with customers about measurements and swaps next week.

“We have not done any changes in the CC2650 production during this period”:
I told to cusotmers this, they had an additional question.

DS has only TYP values, but do you have variance, maximum, minimum data?

Sorry to bother you again. I will definitely reply next week about measurements and swaps.

Regards, Rei

We have some data on the expected variation be the variation is expected to be random meaning that when you do measurements on different chips the expected is that the results will be following a Gauss curve and not a linear increase over time.

1 mA current variation between two random sample is expected  but not that the current consumption increase slowly over time.  

Thank you for everything. I told this information to customers.

- Do they just measure current or do they also measure other parameters as part of the production test? If they also measure other parameters, how is the variation on these over time?

Customers measure other parameters, but haven't arranged them yet. However, customers are also making another CC2650 prduct, and that product doesn't seems to increase.

- Is the production test always done with the same parameters (temperature, settings, firmware, same setup)?

Temperature is unknown (indoor), same firmware and settings.

(RF Differential,internalbias, DCDCON, Zigbee mode ,5dBm)

- Has it been done any changes to the design?

No change

- Could you give more details on the swap test? Was two devices selected, one from the start of the production ("A") and one of the latest (B") and the CC2650 from "A" was placed in "B" and the other way around? And the power consumption increases in "A" after the swap and decreases in "B" after the swap?

Attaching B to A's board increased the current consumption. When A was returned to the A substrate, the current consumption decreased. It has not been fully tested for swap because it has not been tested on the B board. Also, the difference in current consumption in the above test is about 1mA, which is within the error range you said.

As a result of arranging this situation, we think that CC2650 is not bad.

Thank you for your cooperation. We are currently requesting the customer to measure the chip itself and check the peripheral circuits. I will keep you updated.

Regards, Rei

" However, customers are also making another CC2650 prduct, and that product doesn't seems to increase.": Meaning that they only see an increase in current  over  time on one product but not on a different product that uses C2650 from the same production batches? 

If that is the case it indicate something on the board.