Connecting a DS18B20 Temperature Sensor to the ESP32

目录

After getting the ESP32 connected to WiFi and Bluetooth in the previous post, next I want to hook up some sensors to play with. Let’s start with the most common temperature sensor, the DS18B20. Its advantages are digital output, multiple devices on a single bus, and decent accuracy.

1. What You Need

• One ESP32 development board
• A DS18B20 temperature sensor module (or the bare sensor)
• One 4.7kΩ resistor (for pull-up)
• Some jumper wires

2. Wiring

The DS18B20 comes in a three-pin package. Wire it up as follows:

DS18B20	ESP32
VCC	3.3V
GND	GND
DATA	GPIO13

Note: a 4.7kΩ pull-up resistor must be connected between DATA and 3.3V; otherwise the OneWire bus is unstable and devices often can’t be read.

        4.7kΩ
3.3V ───\/\/\/───┬──── DATA (GPIO13)
                │
DS18B20        ─┴────
VCC ──────────────── 3.3V
GND ──────────────── GND

Photo of the actual wiring:

3. Code to Read the Temperature

MicroPython ships with the onewire and ds18x20 modules — no extra installation needed.

from machine import Pin
import onewire
import ds18x20
import time

ds_pin = Pin(13)
ds_sensor = ds18x20.DS18X20(onewire.OneWire(ds_pin))


def read_temperature():
    roms = ds_sensor.scan()
    if not roms:
        print('未检测到 DS18B20 设备，请检查接线')
        return None

    ds_sensor.convert_temp()
    time.sleep_ms(750)  # 等待转换完成

    temps = []
    for rom in roms:
        temp = ds_sensor.read_temp(rom)
        if isinstance(temp, float):
            temps.append(round(temp, 2))
    return temps


while True:
    temps = read_temperature()
    if temps:
        print('当前温度:', temps, '°C')
    time.sleep(1)

Code notes:

• scan() scans the OneWire bus for devices and returns each device’s 64-bit ROM address.
• convert_temp() sends the temperature-conversion command to all devices.
• The DS18B20’s default 12-bit precision conversion takes about 750ms, so sleep_ms(750) is a must.
• read_temp(rom) reads the temperature of the specified device.

4. Identifying the Sensor Model

Running the command below reads the ROM address and family code:

from machine import Pin
import onewire, ds18x20, binascii

ow = onewire.OneWire(Pin(13))
ds = ds18x20.DS18X20(ow)
roms = ds.scan()
for rom in roms:
    print('ROM:', binascii.hexlify(rom), 'Family:', hex(rom[0]))

Output from the one I have on hand:

ROM: b'2824bb020002249f'  Family: 0x28

Family code 0x28 means it’s a DS18B20. If it were 0x10, it would be the older DS18S20.

5. Actual Readings

Serial output:

当前温度: [32.19] °C
当前温度: [32.19] °C
当前温度: [32.12] °C

The thermometer reading as photographed:

This temperature is room temperature plus the residual warmth from my fingers pinching the sensor, so it reads a bit higher than normal room temperature.

6. Common Issues

• Device not detected: first check whether the pull-up resistor is connected and whether the jumper wires are firmly plugged in.
• Temperature occasionally jumps to 85°C: this is the DS18B20’s power-on reset default; it usually means the conversion hadn’t finished when you read it. Lengthen the sleep_ms time to 1000ms.
• Multiple sensors: scan() returns multiple ROMs; convert_temp() triggers all devices at once, then just read_temp(rom) one by one.

7. Uploading to the Blog Database in Real Time

blog-api already provides a temperature write endpoint, so no backend code is needed. The API details:

• Write: POST https://your-api-domain.com/api/v1/temperature/realtime
• Auth: header Authorization: Bearer sk_xxx
• Request body:

{
  "value": 26.56,
  "timestamp": 1723968000000,
  "device_id": "2824bb020002249f",
  "source": "esp32-ds18b20"
}

7.1 Creating an API Token

Log into the blog-api admin console and create an API Token with write permission, or just use curl:

curl -X POST https://your-api-domain.com/api/v1/admin/api-tokens \
  -H "Authorization: Bearer <你的JWT>" \
  -H "Content-Type: application/json" \
  -d '{"name":"ESP32温度上传","permissions":"write"}'

Put the returned sk_xxx into api_config.py:

API_BASE_URL = 'https://your-api-domain.com'
API_TOKEN = 'sk_你的Token'
UPLOAD_INTERVAL_MS = 300000  # 5 分钟

7.2 The Upload Script

Core logic of temp_uploader.py:

import network
import time
import urequests
import ujson
import ntptime
from machine import Pin
import onewire
import ds18x20
from wifi_config import SSID, PASSWORD
from api_config import API_BASE_URL, API_TOKEN, UPLOAD_INTERVAL_MS

DS_PIN = 13


def connect_wifi():
    wlan = network.WLAN(network.STA_IF)
    wlan.active(True)
    if not wlan.isconnected():
        wlan.connect(SSID, PASSWORD)
        while not wlan.isconnected():
            pass
    print('IP:', wlan.ifconfig()[0])


def sync_ntp():
    try:
        ntptime.settime()
        print('NTP synced')
    except Exception as e:
        print('NTP sync failed:', e)


def read_temperature():
    ds_sensor = ds18x20.DS18X20(onewire.OneWire(Pin(DS_PIN)))
    roms = ds_sensor.scan()
    if not roms:
        return None, None
    ds_sensor.convert_temp()
    time.sleep_ms(750)
    rom = roms[0]
    temp = ds_sensor.read_temp(rom)
    rom_hex = ''.join('{:02x}'.format(b) for b in rom)
    return round(temp, 2), rom_hex


def upload_temperature(value, device_id):
    url = API_BASE_URL + '/api/v1/temperature/realtime'
    headers = {
        'Authorization': 'Bearer ' + API_TOKEN,
        'Content-Type': 'application/json',
    }
    # 当前 MicroPython 的 time.time() 使用 2000-01-01 为 epoch，
    # 需要加上 1970-01-01 到 2000-01-01 的秒数偏移，转换为标准 Unix 毫秒时间戳
    unix_timestamp_ms = (time.time() + 946684800) * 1000
    payload = {
        'value': value,
        'timestamp': unix_timestamp_ms,
        'device_id': device_id,
        'source': 'esp32-ds18b20',
    }
    response = urequests.post(url, headers=headers, data=ujson.dumps(payload), timeout=15)
    status = response.status_code
    response.close()
    return 200 <= status < 300


def main():
    connect_wifi()
    sync_ntp()
    while True:
        temp, rom = read_temperature()
        if temp is not None:
            print('Temperature:', temp, '°C')
            upload_temperature(temp, rom)
        time.sleep_ms(UPLOAD_INTERVAL_MS)


main()

Once running, the ESP32 reads the temperature every 5 minutes and uploads it to blog-api’s realtime_temperature_records table.

7.3 Verifying the Data

Public query endpoints:

# 最新一条
curl "https://your-api-domain.com/api/v1/public/temperature/realtime?user_id=1"

# 最近 7 天历史
curl "https://your-api-domain.com/api/v1/public/temperature/history?user_id=1&days=7"

Just replace user_id with your actual user ID.

8. Showing It Off

After the data was flowing, I added a “Current Hangzhou Temperature” widget to my blog homepage; clicking the temperature pops up trend charts for the last 1 day and last 30 days. The image below is the live effect, screenshotted just this morning:

9. Video Demo

Here’s a video demo of this tinkering session:

10. Next Steps

I could add a temperature card to the blog frontend that calls /api/v1/public/temperature/realtime to show the current ambient temperature, plus a line chart calling /history for the 24-hour trend — and that would make a complete wireless thermometer dashboard.

文章作者：阿文

文章链接： https://www.awen.me/post/51fea752.html

版权声明：本博客所有文章除特别声明外，均采用 CC BY-NC-SA 4.0 许可协议。转载请注明来自 阿文的博客！