Wiring of BMP180 ?

[vincen]

Hi

I bought these modules https://www.aliexpress.com/item/15969-F ... 8e01df20d4 but forgot to check if there were 3.3V or 5V: I have wired it on GPIO-4 (D2), GPIO-5 (D1) and also powered in 3.3V, it's declared as device in Devices tab but I have no vajue showing up in Web interface of ESP !!
I have read that pins of ESP are not 5V compatible but sensor being in I2C I guess there is no harm to power sensor board in 5V no ? as I2C are standard levels not depending of Power supply voltage on electronic right ?

Thanks for clarification

Vincèn

[LisaM]

pics or schematics of your setup please
Ps. i2c often needs a complete power-down (off and wait at least 10 secs!) to work...
ps2. 5v tolerant: https://ba0sh1.com/blog/2016/08/03/is-e ... -tolerant/

[vincen]

pics or schematics of your setup please

Here is config of my ESP:
and for wiring it's basic ! I wired the 2 i2C lines according to setup of ESP, and then wired GND and Vin to 3.3V !

Ps. i2c often needs a complete power-down (off and wait at least 10 secs!) to work...

Already done but not much success :(

ps2. 5v tolerant: https://ba0sh1.com/blog/2016/08/03/is-e ... -tolerant/

Thanks a lot for that so looks like it's safe :5 unhappy I discovered that I have no 5V pin to power my sensor Will need to find an other board with 5V power supply to check where is problem !

[LisaM]

Does the BMP show up on the i2c scan?
To which gpio's is the bmp's SDA/SCL connecten?
bmp setup details?

[Shardan]

First step:
Go to "Tools" and click "I²C Scan".

You should get back a list like this:

BMP180-I2C_Scan-Results.jpg (102.97 KiB) Viewed 11082 times

.
If you get a "No I2C devices found", the sensor does not answer.
In this case, check your cabling - a common mistake is swapping SCL and SCA or power cabling failures.
Check the "Hardware" settings too if the settings for I²C pins is correct.
(Should be set to GPIO4 & GPIO5).

Otherwise if the BMP180 is detected correctly please post a screenshot of the device settings.

Please remember that I²C is not suitable for long lines, it is limited to "In box connecting".
More info in the Wiki: https://www.letscontrolit.com/wiki/inde ... he_I²C_Bus

The sensor module runs quite well on 3,3V. The sensor itself has a voltage supply of 1,8...3,6V.
On the board avoltage regulator is placed (the black three-pinned part), so there is no matter
if you feed 3,3V or 5V als long as the power supply is reliable.
All BMP and BME sensors i use run on 3,3V without problems.

Due to the article LisaM linked the ESP can take 5V at the inputs under certain circumstances.
There are specific GPIO settings that will produce too much current over the GPIO and then
the ESP dies. As we usually don't control these settings i won't recommend to connect 5V to the GPIOs.
Anyways, as this breakout board has a voltage regulator, it will not push 5V to the GPIOs so no worries.

Regards
Shardan

[vincen]

Thanks a lot for detailed checking list and in fact I found the culprit (defective sensor board or sensor I changed it with an other new one and worked straight as soon as wired and it shows up now in i2C scan

Seeing that it looks like the DHT sensor family is really bad quality no ? 3° difference in temperature for 2 sensors side by side

[Domosapiens]

Shardan wrote an excellent Wiki: Basics:_Calibration_and_Accuracy
https://www.letscontrolit.com/wiki/inde ... d_Accuracy

With a BMP180 this "wet" procedure is difficult.

[vincen]

Shardan wrote an excellent Wiki: Basics:_Calibration_and_Accuracy
https://www.letscontrolit.com/wiki/inde ... d_Accuracy
With a BMP180 this "wet" procedure is difficult.

Put it in a dry condom as he indicates Considering it's 0.5° precision I guess I'll keep it with factory calibration

[Shardan]

It's not that easy with tolerances.

With the DHT it should (!) be more exact - it uses an DS18B20 internally that should (!) have a tolerance of +/- 0,5°C.
With +/- 0,5°C there is a maximum deviation of 1°C.... Anyways it's the rule as always: You get what you pay for.
I have same experiences with DHT22, especially with cheap ones from chinese market - tolerances of 10%.
DHT11 is worse, it has a resolution of 1° so deviations might add up to 2°C.

With the BMP/BME types it's same. I doubt that the sensor on the cheap breakout boards is really made by Bosch Sensotec
or Texas Instruments so the quality standards might be below the genuine ones.

On a second note ist this: From the "Bosch Sensotec" BME280 Datasheet:

Temperature measured by the internal temperature sensor. This temperature value depends
on the PCB temperature, sensor element self-heating and ambient temperature and is typically
above ambient temperature.

At least BME280 and Si7021 have a self-heating to recondition the humidity sensor.

I can confirm this. The sensors i use all have a deviation of +3 ...+3,5°C.
I took a lab thermometer for reference and just subtracted the difference in the formula field.
Not a really exact calibration but for homecontrol it's suitable.

Regards
Shardan

[Ton_vN]

Performance on I2C-interfaces may benefit from placing pull-up resistors for the SDA- and SCL-lines

[TD-er]

[....]
At least BME280 and Si7021 have a self-heating to recondition the humidity sensor.

I can confirm this. The sensors i use all have a deviation of +3 ...+3,5°C.
I took a lab thermometer for reference and just subtracted the difference in the formula field.
Not a really exact calibration but for homecontrol it's suitable.

For this reason, I changed some code (and merged the code for BME280/BMP280) in this pull request: https://github.com/letscontrolit/ESPEasy/pull/481
One of the changes was to change the parameters to allow for maximum amount of sleep time to prevent the sensor to heat up. This may result in 1.5 - 2 degrees of heating up by just sensing continuously.
And still the sensor heats quite easy by placing it inside a small box with the ESP8266 and some other components.
This seems to be a constant value, given the positioning of the box, since the amount of energy converted into heat is constant.
But subtracting this constant value from the temperature is not enough. You also have to compensate the humidity, which is a lot more complicated.