For a battery driven application I am considering using the DA14580. The plan is to only turn on the part (from full power down) once a day over a 4 year period. In trying to get as much life span out of the battery as possible I would like to push the operation of chip below 2.5V. The datasheet clearly states that cold boot is not possible below 2.5V due to OTP corruption. It goes further to say that it is not recommended to operate below 2.35V. This raises several questions:
1. Can the DA14580 boot from an external memory below 2.5V?
2. I know the BLE stack is contained in ROM. Can the ROM operate reliably below 2.5V, 2.35V?
3. When operating below 2.35V is performance of the chip compromised in any way?
4. When operating at low voltages in buck mode, does the current draw increase? If I assume constant power I calculated the following curve. Is this accurate?
mA Vbatt
5.10 3.0
5.16 2.9
5.36 2.8
5.56 2.7
5.79 2.6
6.02 2.5
6.28 - 2.4
6.56 2.3
6.87 2.2
7.21 2.1
7.57 2.0
Thanks for any enlightenment you can provide.
嗨Markhillig,
During a cold boot, certain trim and calibration values are read out of the OTP and saved in registers. It is feasible to read these register values out during production (while the voltage is sufficiently high) and store them in your external memory. When you boot-up in the field, you would then read these settings out of you external memory and save them in the appropriate registers. The issue only pertains to the OTP so no other blocks are affected. Again, this method is cumbersome, but is feasible.
You might actually want to consider using boost mode instead of buck mode. Silver Oxide batteries (button cells) have a better performance thanyour typical Lithium battery. A silver oxide battery will also have a larger energy density (mWh per volume) and they hold their nominal voltage almost all the way to the end. In boost mode, you do not have to worry about the OTP voltage.
I am not sure how you calculated the consumption figures. First of all your numbers seem to capture the peak currents and those are not a good indicator of average power consumption. Secondly, the peek current at 3.0V is just below 5mA (Which includes everything from loss in buck regulator, micro controller and receiver opened). But remember, that a BLE chip is at sleep most of the time - in your case that would be deep sleep which is about 550nA @ 3V of 1.2uA @ 1.5V. At these current levels, the timers are running and will allow the daily wake-up.
Thanks for the reply.
我们一个re using a proprietary battery chemistry and are currently defining what the battery performance characteristics need to be. The reason I listed the peak currents is we are trying to balance instantaneous battery sourcing capacity with overall energy density. They are counter balanced when mixing different battery chemistry. Boost mode would require too much instantaneous sourcing capacity (>10mA) for our application and we have no room for charge storage in the electronics. We understand the "average" current consumption calculations very well. In buck mode, per the datasheet, the peak current is around 5.1mA (RX) at 3.0V.
Hello,
here some current consumption data I had: DA14580-01 in buck mode @ 2.0V; 2.4V; 2.8V.
Antenna terminated with 50 Ohm load.
Vbat3V = 2.0V: Tx pk = 6.35 mA ; Rx pk = 7.2 mA ; Iavg = 0.96 mA
Vbat3V = 2.4V: Tx pk = 5.25 mA ; Rx pk = 6.0 mA ; Iavg = 0.85 mA
Vbat3V = 2.8V: Tx pk = 4.75 mA ; Rx pk = 5.3 mA ; Iavg = 0.74 mA
Best regards, BB_Dialog.
谢谢你的信息!这个问题的答案4. Do you have any thoughts on questions 1-3?
Hi markhilliq,
I think that mhvid_dialog provided most answers in his reply.
In general, below 2.35V, th OTP can not be read anymore.
So, if your application is not using OTP (but e.g flash to boot from), the DA14580 can be used below 2.35V in buck mode.
Best regards, BB_Dialog
If you are using external memory, there is option to set the voltage to 3V in production test, read the XTAL trim values and store them in FLASH.
BR JE_Dialog