Temperature differences

[darryl]

Me again

I am experimenting a little and now have the ds18s20 on the temp/solar/hum and a seperate ds18s20 sitting right beside each other.

Sensor 0 is reading 20.12 C
Sensor 1 is reading 23.00 C

Note the difference between the two is larger than the +/- .5 C in the specs. Any idea what could be causing the difference?

 

[ericvic]

Darryl,

When you hooked up the DS18S20 did you tie the VDD pin and the ground pin together? They need to be tied together if you are not providing external power.

Eric

 

[FrankMc]

Hi Guys

I also use the DS18S20 with my own hardware and when testing (only a few inches apart) i noticed similar findings to yours...but if i tied both sensors together i found that they were always reading the same temp ....??

HTH
Frank

 

[darryl]

I'll try powering the sensor to see what happens..


Is the ds18s20 powered in the hum/temp/solar sensor?

 

[ericvic]

No it is not. But it does have VDD tied to ground.

Eric

 

[darryl]

I've tried powered and parasitic and i've tried 2 different chips. The onboard ds18s20 always reads about 2 deg C higher than everything else.

I also tied vdd and gnd together when i didn't provide external power as per your first post (it doesn't seem to make a substantiative difference in the readings though).

 

[darryl]

bump.

Any clues to the difference?

 

[ericvic]

I don't have any ideas why it would be reading higher. The sensor on the board is hooked directly to the 1-Wire network with no external power so it should be connected just like the stand-alone one. Could the board itself be getting warmed up by some external means (sunlight, sitting on something warm, etc...)? I could see where the thermal mass of the board could cause the temp sensor to warm up a bit.

Eric

 

[darryl]

I have the hum/temp/solar sitting on top of an empty cardboard box with no direct sunlight. I have the standalone (now powered) sitting approximately 3 inches away on the same box. I'm reasonably sure there is no thermal transfer anywhere that could account for the 2 deg C difference.

-Darryl
P.S. I'm not attempting to question the quality of your device, i'm trying to make sure i'm not doing something overtly dumb as well as try and see if anyone has seen variances like this on ds18s20 chips.

 

[ericvic]

Darryl,

I never got the feeling that you were questioning the quality of our products and I hope I'm not coming off defensive. I'm just throwing out possibilities.

I can't think of anything else unless you somehow got a DS18S20 that wasn't calibrated correctly.

Maybe someone else has some other ideas.

Eric

 

[huggy59]

I have not used these particular sensors, but I have seen significant differences in the 1-wire sensors I have used (2-4 degrees F), and can't account for it other than the sensors are not calibrated as closely as one would like. I decided to correct it in software for the critical ones.

Is it possible to harm them when soldering so that the calibration is off?

 

[ericvic]

I think if you left the soldering iron on the joint too long you could damage the component but normal soldering temps and times shouldn't cause any problems.

Eric

 

[darryl]

Not to beat a dead horse but when I got home today I checked my sensors and they now have a 4 1/2 degree C difference and are a handful of inches apart.

I'm totally confused that in all the research I did prior to buying 1-wire hardware I was always under the impression that they were fairly reliable and accurate.

 

[ericvic]

That is rather strange. I could see them being 1 - 1-1/2 degrees celcius apart because of the 0.5 degree resolutiuon and accuracy. I will do a bit of research and see if I can come up with an explaination.

Eric

 

[darryl]

Hey Eric. I mounted anther ds18s20 inside the humidity case (connected to the screw terminals). It reads basically the same as the other ds18s20 i have sitting right beside the case. The only thing I can figure is the ds18s20 that was mounted inside is out of calibration by 2 deg C.