TMP007- does ir temp sensing work well?

Breakout boards, sensors, other Adafruit kits, etc.

Moderators: adafruit_support_bill, adafruit

Please be positive and constructive with your questions and comments.
Locked
tomo21
 
Posts: 17
Joined: Sun Apr 14, 2013 10:35 pm

TMP007- does ir temp sensing work well?

Post by tomo21 »

Hey guys, was hoping some people who have used the TMP007 could tell me a little bit about it. I have one right now and I have it up and running no problem, but the temperature keeps drifting all over the place. Also, is that 4 second delay in the sample code necessary? Is there something significant about getting 16 samples for stability or something? Any comments appreciated!

Thanks!

tomo21
 
Posts: 17
Joined: Sun Apr 14, 2013 10:35 pm

Re: TMP007- does ir temp sensing work well?

Post by tomo21 »

For reference, I'm pointing my sensor at an aluminum water bottle full of ice water, which I'm measuring with an Oakton Infrapro 6, and it's telling me that the surface is around 60 degrees F. Here's the output of my TMP007:

Object Temperature: 39.25*C
Object Temp Farenheit: 102.65
Die Temperature: 11.94*C

The room temp is about 78F (25.5C). TMP007 readings make no sense to me. So, my question is more or less does this thing work at all, or do I have a bad one, or am I doing something wrong? I have the sensor about 1/8" from the surface of the bottle. Thanks.

User avatar
Franklin97355
 
Posts: 23910
Joined: Mon Apr 21, 2008 2:33 pm

Re: TMP007- does ir temp sensing work well?

Post by Franklin97355 »

Could you post clear pictures of both sides of your board showing any soldering you have done and the connections to it?

User avatar
egutting
 
Posts: 297
Joined: Wed Nov 14, 2012 12:57 am

Re: TMP007- does ir temp sensing work well?

Post by egutting »

If I'm remembering correctly, the sample every 4 seconds is the highest conversion rate. You can change the conversion rate to anything from 1 to 16 (time of .26 seconds to 4.01 seconds.) Also, you have to remember that the TMP007 has a Field of View that senses all radiation in front of it (and weighs it depending on where it's at) Theta(90 degrees) is 50% of the energy and Theta(135 degrees) is about 90%. So if your glass of water is in front of a really hot monitor, then it could be sensing the temperature of the monitor (or anything else for that matter) and including it in the calculation.

You can check out the data sheet here: http://www.adafruit.com/datasheets/tmp007.pdf

Are you doing this on an Arduino or Raspberry Pi? Let me know and I can help tell you how to change the sample rate (conversion rate)

tomo21
 
Posts: 17
Joined: Sun Apr 14, 2013 10:35 pm

Re: TMP007- does ir temp sensing work well?

Post by tomo21 »

Pictures of setup: http://imgur.com/LpCDmpm,UndwvTO,JEq0Ss6#0

Hard to see, but I'm connected to 3.3V and GND on the Arduino.

egutting, thanks for the advice. Hadn't got that far down in the datasheet. Kinda glazed over at the boltzman constant stuff :). So if I 'put some blinders on' the sensor and constrain the FOV to just the water bottle, I'm assuming I"ll get some better results? Also, I didn't see anything mentioned about an 'ideal distance' for measuring temperature. What seemed to work best for you? I'll take a look at the conversion rate stuff. This thing is only accurate to 3 degrees C (which isn't great), so the more accurate, the better. 4s per sample is more than adequate for this application. I'll look into that on the datasheet. Thanks.

tomo21
 
Posts: 17
Joined: Sun Apr 14, 2013 10:35 pm

Re: TMP007- does ir temp sensing work well?

Post by tomo21 »

Also, have you used the TMP006? The only real differences I can see are 1)price at volume (TMP006 wins) and 2)an 'integrated math engine', which I have no idea what it actually does.

User avatar
egutting
 
Posts: 297
Joined: Wed Nov 14, 2012 12:57 am

Re: TMP007- does ir temp sensing work well?

Post by egutting »

As far as the blinders, yes you can limit the field of view by using some. (I think TI has a video on their website talking about that subject. I can't find it right now, but I know I watched it last week or so.) I haven't actually deployed mine in the field as I'm still working on a mounting solution (going to be using it for my weather station as a cloud sensor). Also I've not tested it against another IR temp sensor as I do not have another. When I aimed it at myself, or the computer monitor, or a glass of water, it seemed to report temps that were "reasonable".

If you want to change the sample rate add one of these variables:

Code: Select all

TMP007_CFG_1SAMPLE 
TMP007_CFG_2SAMPLE 
TMP007_CFG_4SAMPLE 
TMP007_CFG_8SAMPLE 
TMP007_CFG_16SAMPLE 
....to your begin statement.
ie:

Code: Select all

 if (! tmp007.begin(TMP007_CFG_4SAMPLE )) {
Serial.println("No sensor found");
while (1);
and change your delay to correspond with the datasheet. For the above it would be:

Code: Select all

 delay(1010); // 1.01 seconds per reading for 4 samples per reading
I do not have the TMP006. A buddy of mine does and I think it works for him, but I'm not as versed on the differences between the sensors....Updated means better to me :) (I know this is not always true, but I went with the TMP007 as it was just being offered by Adafruit when I was looking into IR temp sensors)

tomo21
 
Posts: 17
Joined: Sun Apr 14, 2013 10:35 pm

Re: TMP007- does ir temp sensing work well?

Post by tomo21 »

I made a little blinder thingy and 3D printed it. I'm still testing it, but here's what it looks like.
http://imgur.com/jxDfoUH,cggeKdV,4GEIdbD#0

If someone wants .stl files for it let me know. It was pretty easy to make in tinkercad. It interfaces with a 70mm diameter bottle pretty nicely.

I'm still trying to wrap my head around what changing the sample rate will do, but once I find out what sample rate works best for me, I'll implement it with your code snippets, thanks.

Is there a way to calibrate the sensor to a specific material? I think the emissivity of the material being measured impacts the amount of IR radiation it emits, so the sensor should be calibrated to the material for accurate readings.

tomo21
 
Posts: 17
Joined: Sun Apr 14, 2013 10:35 pm

Re: TMP007- does ir temp sensing work well?

Post by tomo21 »

...found a calibration guide. It's calibrated for something with an emissivity of .95 out of the box. Anodized aluminum is .9, so I might just roll with that for now.
http://www.ti.com/lit/ug/sbou142/sbou142.pdf

tomo21
 
Posts: 17
Joined: Sun Apr 14, 2013 10:35 pm

Re: TMP007- does ir temp sensing work well?

Post by tomo21 »

UPDATE: The blinders have helped with consistency, however, the sensor still isn't very accurate. Measuring off the aluminum surface was giving me garbage, so I put some black tape over it and I got some more consistent readings. I have some ice water in the aluminum water bottle again.

TMP007 on black tape: 58F
Oakton Infrapro 6 on black tape: 49F

Perhaps calibrating will help this offset?

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

Return to “Other Products from Adafruit”