Esp-01 and dht11 and pullup resistor

I’m try to set up a ESP-01 with a dht11 board on it. The readings in HA are not reliable. Most of the time a got unknown. In the ESPHome log there is: unknown readings, check wiring (pullup resistor or pin)
But sometimes I have readings so nothing wrong on that side. The schematic of the dht board shows a pullup resistor of 4.7k.
Anyone seen this before?

If you are using a sensor that is mounted on a board you won’t need a separate pullup as the board will have one built in, might be worth checking it is there and soldered OK. I would check connections between boards or change the pin you are using on the ESP.

Do you have the sensor model defined or have you left it as the default “AUTO_DETECT”?

Nine times out of ten setting this to the actual sensor model solves communication issues.

BTW the DHT11 is a terrible sensor. The DHT22 is much better.

It’s now working Ok for a few hours after rebooting. Could this behavior comming from the dht11?

As it seems to be intermittent I would guess its got something to do with wiring. How is it connected? soldered or with jumper cables?

I would have it soldered, but that is assuming you have the tools and ability to solder, far more reliable that way.

No it’s a board plugged on the header of the ESP-01. No jumper, no soldering. However, it’s still working.

I have defined it.
sensor:

  • platform: dht
    pin: GPIO2
    model: DHT11
1 Like

esp01Temp
Is this the board you are using?

Correct, that’s it.

This module suffers very badly from the heat from the board and the ESP01. I would suggest to save yourself the frustration and pick a different sensor.

Any advise? DHT22 as tom_i advised, or is a BME280 possible?

I did run a BMP280 on an ESP01 using Tasmota so BME should work.

I’ve had great success with SHTC3 sensor modules that consistently provide temperature and humidity readings across many sensors. Have at least 8 of these, and they are all ±1 degree on temp and humidity. They are often priced much less than DHT22 modules that are much worse.

I’ve tested BME, SHT and DHT. The SHTs and BMEs self heat a lot. DHT22s are still my goto sensor.

Spiro, are using the two GPIO’s on the header, or take one extra from the chip.

I needed 2 pins( for I2C the sda and scl)to run BMP280 as well as 3V and GND. I used GPIO2 and GPIO0. Couldn’t keep sda and scl connected while programming. Normally I would avoid using GPIO0 but I didn’t know enough then to be worried. I cut off the sensor on the board and used dupont wires to connect between the yellow slots for 3V and GND to the 3V and GND on the Esp01. I would not have had the skill to wire directly to the chip.

I had 2 sets of 2 wires joined together. I used them first to program the ESP01 (like in the picture above)and then moved them later to power ESP01 and BMP280.
I had to use the old board to get me 3V from a 5 V power supply. It was a mess of wires. Now I would have just used an LDO linear regulator and put it on a small PCB with a power and GND line to keep it tidy. ESP01 pins dont sit well on PCB perf-board though.

Don’t waste your time with these boards.
They suck.
Not only do they read incorrectly due to self-heating, but they fail often.

My experience with them:
https://www.aaroncake.net/house/automation/sensors.asp#ESP8266DHT22WIFIMOD