DHT22... bad sensor?

Breakout boards, sensors, other Adafruit kits, etc.

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
Locked
arpinkr
 
Posts: 1
Joined: Tue Mar 12, 2013 7:24 pm

DHT22... bad sensor?

Post by arpinkr »

I just bought a DHT22. I hooked it up and ran the serial test sketch included with the DHT library. The connection was working fine and I was getting readings. The temperature started at about 20C (which is the actual temperature of my house) but quickly began to increase and stabilized around 34C. The sensor felt slightly warm to the touch. When placed in front of a fan, the temperature decreases back to near 20C. I suspect I have a bad sensor with a short leading to some self-heating? After a couple hours, the sensor stopped communicating altogether. I think I got a bad sensor. Anyone have thoughts or things I haven't tried?

User avatar
adafruit_support_bill
 
Posts: 88093
Joined: Sat Feb 07, 2009 10:11 am

Re: DHT22... bad sensor?

Post by adafruit_support_bill »

Very strange. If it was communicating, you must have had it connected properly. I'd have to agree it is some sort of internal short.

Contact [email protected] with a link to this thread and ask for a replacement.

User avatar
hagrd
 
Posts: 1
Joined: Fri Jun 06, 2014 10:27 am

Re: DHT22... bad sensor?

Post by hagrd »

I've been experiencing what I'm considering self-heating as well with the DHT22 sensors. On a breadboard or when connected through pin-headers on a PCB the effect is less severe, although definitely present, than when the sensor is soldered into the PCB directly. I haven't found anything warning against soldering the sensor to a PCB and I have used a low-heat soldering temperature just in-case. With the sensor soldered directly to a PCB I see temperatures 3-4 degrees F higher and when plugged in through pin-headers the heating effect is lower, between 1.5 and 2.5 degrees F high. I've tested a number of these sensors and it seems that forced air circulation is the best way to get correct temperatures in any case.

In another test, I've tried powering the DHT22 through a digital output pin on my microcontrollers (Teensy 3.1) and have gotten good results, although still high temperatures. The setup is as follows: PCB with DHT22 soldered on and the device is not in an enclosure, i.e. open to air. The microcontroller performs a number of tasks in each loop (about one minute per loop) including reading the DHT22. One second before asking the sensor for humidity and temperature readings, a digital pin powers up the DHT22, then 100ms after the tasks the digital pin is set to low and the DHT22 is turned off. Even with the DHT22 soldered to my PCB the readings are only about 1.5 degrees high, which is a big stride forward without forced air.

If anyone has experienced this problem, found a solution, or know of a forum post I haven't seen yet please let me know.

User avatar
adafruit_support_mike
 
Posts: 67454
Joined: Thu Feb 11, 2010 2:51 pm

Re: DHT22... bad sensor?

Post by adafruit_support_mike »

Self-heating has no relation to the way you soldered the connections or what the device is connected to. It's a side-effect of sending current through the components.

Any current flowing through any resistance converts some amount of energy to heat. You don't usually get much heat at low current levels, but relative humidity is calculated in terms of air temperature, so the DHT22 has to keep its thermal footprint as small as possible.

If you're getting self-heating, it means you're sending too much current through the device.

The effects you're seeing don't sound like self-heating though, especially when you explicitly limit the amount of time the DHT22 receives power.

I have to ask though: what kind of calibrated temperature sensor are you using for comparison, and where are you taking that measurement?

User avatar
adafruit_support_bill
 
Posts: 88093
Joined: Sat Feb 07, 2009 10:11 am

Re: DHT22... bad sensor?

Post by adafruit_support_bill »

Also, what other components are on the PCB? If there are other power consuming components close by, the temperature near the surface of the board could easily be several degrees above ambient.

User avatar
mattfield
 
Posts: 1
Joined: Wed Oct 08, 2014 6:16 am

Re: DHT22... bad sensor?

Post by mattfield »

I had the same problem..I applied 2 fixes...
1. I realized my bread board was sat in front of a stream of hot air from my PCs ventilator....
2. I hadnt put the 10K pull up resistor.So I put one in.

Not sure which one fixed it, but it fixed it...

Locked
Please be positive and constructive with your questions and comments.

Return to “Other Products from Adafruit”