ESP8266 Community Forum

- Sat Feb 06, 2016 11:45 am #40570 Perhaps I have lead everyone down the wrong path, I should not be so hasty.
In anycase my issue is not resolved but I am a lot wiser.

Chip IDs: 0x55 BMP180, 0x58 BMP280, 0x60 BME280

Chips I received were BMP280 when in fact they should have been BME280.

The bottom line is that I also have temperature errors with the real BME280 and that in fact slews all other values.
I2C timing is within specs as per datasheet with VDD & VDDIO & 3.3V and <1mv noise.
I have multiple devices on my I2C bus and all others are bahaving as expected.

So on the end of the day I have not contributed a solution.

- Sun Feb 07, 2016 1:47 pm #40632 I think I found my problem. I build small modules with 3.3V power supplies on them in which I could plug in the ESP-01 modules. I made them quite compact, since I wanted to place them around the house. The header I plugged the BME280 into was close to the ESP-01 module and I suspect picking up some warm air currents heated by the nearby chips. There might even have been some heat conduction along the leads themselves, I have no way to be certain.

On the arduino and digilent (chipkit) boards I put the BME280 breakout on a small breadboard and had 10cm wires making the connections (as well as a level converter in between). Plenty of distance away and the readings were fine. I mimicked this by stacking headers on my ESP-01 module PCBs high enough so my BME280 breakout board was about 5 cm above the PCB. It then gave me readings within about 0.5C of the mercury thermometer. I found a tiny (< 1mm diameter) thermistor I could place close to the BME280 chip to confirm this.

So I suggest checking where you put your BME280 carefully and make the same mistake I made.

- Fri Mar 11, 2016 7:03 am #42893 I have the same problem with the DHT22 and the BME280 sensor.
As soon as I bring the sensor to near to the ESP8266 board the temperature is rising
and humidity is falling.
If I bring the sensor a bit far from the ESP8266 module the problem on the DHT22 is almost gone.
The problem is on a ESP-12e module much bigger compaired to the ESP-3 module .
If i add on my adapter board additional silver wire I get at least
a better GND and VCC to all components and to the sensors. This helped a bit.
Secondly, longer distance between sensor and ESP module helped also on the DHT22.
On the BME280 helped only the distance a bit.
If I test the BME280 on a Arduino board the reading is ok.

My asummption is:
You need a very clean VCC and GND, no ripple no noise.
The Sensor get to much HF from the ESP module.
The heat from the module has no impact.

To overcome all of this problems I designed a new PCB with very big GND and VCC layer and
ferrite pearls to all GPIOS and to the sensors VCC and GND.
I hope I get my new board in the next weeks to continue my tests.

Let you know my results soon.

Regards

Lorenz

- Fri Mar 11, 2016 10:46 am #42905 I had scoped out my signal lines (I2C) and power lines. My ESP8266 board was not noisier than the arduino based boards. I doubt it is noise related. However, as I pointed out, I did put a thermistor next to the BME280 and confirmed the elevated temperature coming from some nearby air currents (presumably off the warm processor).