All the sensors! - Which device?

[neuer-Lehrer]

Hi,

I wanted to use as many sensors with phyphox as possible.
And for that I wanted to get a good phone capable of it.

According to the phyphox database, there are only few devices which have all sensors
(they are on top, when you click Humidity (rarest sensor) "Available" twice).
https://phyphox.org/sensordb/

On another website I found a list, which has a bit more smartphones with humidity sensor.
https://www.phonegg.com/list/302-Cell-Ph...ity-Sensor

E.g. Samsung Galaxy S5 has a humidity sensor on that website.
https://www.phonegg.com/phone/4565-Samsung-Galaxy-S5
SM-G900T (one version of Samsung Galaxy S5) in phyphox database has no humidity sensor available.

-------------------------------------------------------------------------------
- Questions:

How reliable is the phyphox database?
In case of SM-G900T (Galaxy S5) there is a sample size of 2
- does it mean 2 people have submitted their information about their Galaxy S5s?

Is there a realistic chance, that a device can use its sensor in phyphox
even though the database says something else?

And very important, how reliable are those rare sensors, temperature and humidity,
for example in the Galaxy S4?
+- 0,5° Kelvin and +-2% humidity would be accurate enough for me.


If you have some positive experience with all sensors in one device, please share!


All the best,
Felix

[SJS]

Hi Felix,

as far as I know, the Samsung Galaxy S4 uses a Sensirion SHTC1 which covers a humidity measurement range from 0 to 100% and a temperature measurement range from -30 to 100 °C with a typical accuracy of ±3% RH and ±0.3 °C and thus meets your requirements.

However, you should keep in mind that the smartphone itself produces heat (e.g. from the cpu/gpu) and you produce heat as well, if you hold it in your hand or keep it in your pocket. Thus, smartphones might need some time to rest before you use them to measure temperature. Personally, I wouldn't use the smartphone itself to measure temperature, because unless you constantly monitor and limit gpu/cpu activity to a minimum, you'll never (really) know how it has affected your values.

Based on my personal experience I can recommend a Galaxy S4 as a "phyphox-phone" in combination with a SensorTag (CC2650). This would give you the possibility to take measurements with both devices at the same time or compare them to one another. Moreover, both devices have the advantage that they are quite popular, so that many used devices may be available and could be purchased at a low price.

Best wishes, Sebastian

[neuer-Lehrer]

Thanks a lot for the answer!

--------

Just in case, someone knows answers to my other questions, I will quote them:

"How reliable is the phyphox database?
In case of SM-G900T (Galaxy S5) there is a sample size of 2
- does it mean 2 people have submitted their information about their Galaxy S5s?

Is there a realistic chance, that a device can use its sensor in phyphox
even though the database says something else?"

[Dominik Dorsel]

Hey,

In case of SM-G900T (Galaxy S5) there is a sample size of 2

- does it mean 2 people have submitted their information about their Galaxy S5s?

Yes thats right! Here is some more Information you can find on the database webpage:

This is the number of devices that contributed to this entry in our database. Note that while each user can submit their data multiple times, we only take into account the submission with the lowest standard deviation. Hence, the sample size corresponds to the actual number of physical devices that have been tested for this entry.

But keep in mind we dont know how accturate our users follow the test-conditions. Maybe someone discovered the database experiment on a train journey and submits significant worse data due to vibrations.

E.g. Samsung Galaxy S5 has a humidity sensor on that website.
SM-G900T (one version of Samsung Galaxy S5) in phyphox database has no humidity sensor available.

I think you will not be able to use the temperature/humidity sensor with phyphox. If I unterstood Sebastian (Staacks) right phyphox is scanning for a sensor which is named "temperature" (or something like "temp" is in the sensors name). Unfortunately some sensors are just named by there manufactures name. For example if there is a BME280 sensor (temperature, humidity, pressure) and it is named like "BME280" or "Bosch BME280" phyphox will not recognize the temperature sensor. This naming problem is a consequence of the fact that these temperature sensors are not designed to measure the enviroment temperaure but the device temperature as a hardware safety feature (prevent overheating).
I also fully agree with Sebastian (SJS) that a internal temperature/humidity sensor is not that useful (except you are interested in the smartphones temperature while charging or different power modes in general). A bluetooth low energy sensor as the sensortag or a diy arduino project will suit much better.


Best wishes

[neuer-Lehrer]

Thanks!

All my questions have been answered!

Thanks Sebastian!
Thanks Dominik!

[neuer-Lehrer]

It took some time, but now I have a used Samsung Galaxy S4
(GT-I9515 with Android 5.0.1).

Problem:
Humidity and Temperature aren't listed as sensors in the app.
But in the device info of phyphox (i) those sensors are listed.

What do I have to do to use the humidity and temperature sensor in phyphox?
- Upgrade Android?

[Jens Noriʇzsɔɥ]

Humidity and Temperature aren't listed as sensors in the app.
But in the device info of phyphox (i) those sensors are listed.

There are no experiments included for these sensors by default. Have you tried to add a simple experiment with (+)?

Please note that the (internal) temperature value might have less to do with the outer reality.

[neuer-Lehrer]

Thanks, Jens!
Add an experiment is the way to go
(unfortunately I can't rename it to humidity-measurement e.g.).

I am aware of the internal temperature being measured.
My results right now are around 1 to 2 °C warmer than the real environment.

[Jens Noriʇzsɔɥ]

(unfortunately I can't rename it to humidity-measurement e.g.).

You could just newly set it up…

Full flexibility you could have by using phyphox.org/editor or by “manually” editing the .phyphox file. Note that you could send/save experiments anywhere by saving its state.

I am aware of the internal temperature being measured.
My results right now are around 1 to 2 °C warmer than the real environment.

My bad: I have not noticed that it has been already mentioned multiple times in this thread.

[neuer-Lehrer]

Nice advice, thanks again!