Happy New Year!
Bluetooth Low Energy has come a long way in 2012; there are now multiple smartphone models on the market supporting Bluetooth 4.0 (including APIs), single-mode chips have been on the market for almost 1.5 years and we expect to see a lot of new, exciting Bluetooth 4.0 products to be shown at CES next week.
My prediction for 2012 is that this is going to be the year of Blueooth Low Energy end-products finally hitting the store shelves. By the end of the year, I also expect Bluetooth 4.0 to be a pretty standard feature on new smartphones. In short; it will surely be an exciting year!
On another note, our local rivals over at Nordic Semiconductor started the new year by writing an article on coin cells together with Energizer (http://www.eetimes.com/ContentEETimes/Documents/Schweber/C0924/C0924post.pdf). In my view, it is much more thorough than the short article they published a year ago (I wrote a response to this here: http://e2e.ti.com/blogs_/b/bluetooth_low_energy_blog/archive/2010/11/09/bluetooth-low-energy-and-coin-cells-iii.aspx), and therefore much more interesting.
Their approach is a bit different from the one we used to create our coin cell whitepaper (http://www.ti.com/litv/pdf/swra349); we did all our testing on our own, and tested many different batteries from different manufacturers, while Nordic have cooperated with Energizer, and have more detailed data for their cells. Overall, this means the two articles complement each other well.
In my opinion, the data from the Nordic article supports our main conclusion: effective capacity of a CR2032 coin cell is pretty much the same in a BLE application, regardless of whether the peak current is 15 mA or 30 mA (Nordic measured 10 mA and 30 mA). Looking at Figure 8 in the Nordic article, you can see that C (30 mA) and D (10 mA) both have around 200 mAh effective capacity. Remember that in our study, we saw 10-20% differences in effective capacity between cells from the same manufacturer and the same manufacturing batch. We didn't test for higher currents, but from the Nordic data, it is obvious that as you get to 50 or 80 mA peaks, the effective capacity of a CR2032 is severly degraded.
But peak current is just part of the picture. We didn't have the time to look at effects of different peak widths and repetition times, but the Nordic article provides some more insight into this. As shown in Figure 4, the battery voltage will continue to drop under a continous load due to chemical effects. These chemical effects can be ignored in typical BLE scenarios, as the peaks are so short (few ms). If the active periods are longer, however, this effect becomes very significant. The reason Bluetooth Low Energy is so well suited to coin cell application is not just about peak current, but also that the active periods are so short. Many other RF protocols have much longer active periods, and this makes them less suited to coin cell operation even if the peak currents were identical.
Best regards,
Karl