›

Metawear C battery level and remaining time

kspengkspeng
in General
Hi,

We are using Metawear C to monitor the accelerator. We want to estimate how long the remaining time will be by battery level. However, the battery level is quite unstable and inaccurate. Sometimes the battery level drops very soon, but it will return higher level after reconnecting. If we want to estimate the remaining time, how can we do? Can you provide the formula or spec how you calculate the battery level? Thanks

Kuo

Comments

  • MattMatt
    @kspeng

    The high discharge required to stream accelerometer data will tend to droop the battery voltage and battery level reading.  When disconnecting the load is reduced and the battery is able to recover.  Using the throttling settings to reduce data rate of the radio connection will improve battery life and improve the accuracy of the calculation.

    Battery life estimation is highly dependent on the use case -- which sensors are on and at what performance, how much time is spent offline vs online, etc.  To get the best estimation for an end product, we recommend using the raw battery voltage end point and "characterizing" the battery voltage in the product use case.  That is, inserting a fresh battery and using it exactly as it would be in the field, while monitoring the raw battery voltage until it has completely died.  Then a model for battery voltage to battery life can be designed.

    The existing model is a piece wise linear function that calculates the battery life from the discharge characteristics of the battery and the present cell voltage.

    Coin cell batteries are particularly sensitive to loading because they are weak cells.  Our advice would be to take a battery measurement only immediately after connecting to a device -- we are considering this and other methods for implementation in firmware to improve the quality of battery life calculation in the C model, which is more sensitive to loading.
  • kspengkspeng
    @Matt

    Thanks for your answering. We are working on this way. Now we are plotting the relationships among voltage, time, and battery level. However, the API is quite unstable. After I run the API or the sample APP for few hours, the API or APP disconnects automatically.

    For current result, we found that the linearity between voltage and battery level does not quite make sense. Above 3V, the battery level keeps 100%. From 3V to 2.9V, the battery level drops from 99% to 45% very soon. In the interval of 2.9V and 2.8V, the battery level is in the range 44% to 30% and lasts for a long time. The bottom line may be 2.5V and I can not get the exact value so far. 

    I have checked the battery V-A curve and now I am trying to do some inference to calculate the remaining time.

    If you have any update, please let us know. Thanks.
This discussion has been closed.