Android Thermal -1 | Mobile Phone Heating

Shubham Garg
4 min readMay 3, 2023

--

What differentiates a good device from the rest?
What do a potential customer expect when purchasing a new device?
Answers can be many but we can ultimately narrow it down to one word: Performance
We are likely to buy the device which gives best performance assuming the price range of device in consideration is same.

While Performance itself is very broad word, in context with the smartphones we can categorize in 3 topics:

  1. Thermal Performance: Device heating
  2. Application Performance: Touch response, application launch time
  3. Power Performance: Battery backup, standby power.

Performance can be attributed to the 2 major part: Hardware and Software.

Let’s go by an example:

Between a 60hz refresh rate display vs 120hz refresh rate display. The scrolling experience (display performance) will be better in 120hz display.

Another example is while the fast charging reduces the overall time required to fully charge the device it deteriorate the thermal performance as fast charging generates more heat.

Similarly application performance depends on the processor capability.
But while hardware play an important role at a in a given range different smartphone vendors uses almost similar hardware components. Hence software becomes deciding factor for best smartphone.

In this blog we will try to go in depth for some performance customizations.

This was the short preface for the conversation we are going to have about Android Customizations. In this blog we are going to follow the questions and answer writing style.

Let’s start with the Thermal Performance part.

Before going in depth in thermal performance we need to understand 2 important points.

  1. In the smartphone there is no hardware to do the active cooling (direct cooling) so it is necessary to implement some software solution to do passive cooling.
  2. In order to implement some software solutions we need to understand the source of the heat and what action can be taken to reduce the heat.
Simple block diagram for thermal module

Now that we have a layman understanding of the thermal module. Let’s dig down and ask ourselves some questions :

  1. How we are going to measure the temperature?
    : through thermal sensors.
  2. How many sensors we need to accurately gauge the temperature?
    : Some may answer 1 some 2 or 3, to get an estimate let’s ponder the answer of next question.
  3. What are the major components responsible for generating heat?
    :CPU & Battery (Main), Some secondary power draining units like flashlight, Camera (Indirectly cause heating)
  4. Does sensors gives the temperature the user is actually feeling?
    :No they only provides the temperature of the component they attached to.
Board diagram showing the battery temperature sensor.

Below is a table of a smartphone temperature at front and back at different areas while charging for 30min.

A device under thermal camera while using camera for a long time

While we use thermal material to even out the temperature spread in device and better cooling, still due to different back material like glass, plastic, metal or ceramic and different usage like camera, charging.
The device temperature is different in different areas on front and back.

Now with all these question and some data in our pot let’s start cooking and come to some conclusions.

We have around 2 or 3 thermal sensors. 1 at the battery and 1 at the CPU.
While the CPU temperature can go up to 55C and the battery temp around 45–52C . The actual temperature the user is feeling at the front or the back of the smartphone may be different.
So to take correct action we need to do some math.
We get the user temperature which by finding the perfect coefficients of a linear(majorly) or quadratic(sometimes) functions of the CPU Battery and some Virtual sensors (mix of CPU and Battery).
Example:

Back temperature = a1*Btemp + b1*Ctemp + c1
Front Temperature = a2*Btemp + b2*Ctemp + c2

We find the coefficients which is generally done by Hardware engineer by either Machine Learning or by hit and trial since the coefficients are related to the thermal material used and the back material.

The final temperature on which the decision is made is generally the maximum of back temperature and front temperature.

Final temp = max(Back temperature, Front Temperature)

Now we have covered the measure temperature part we will move to decision maker in next blog.

Stay connected and drop a comment or msg in case of any doubt.

--

--