Heat sensing, with IR diode?

Guest
Hi

I am way out here on deep water, but here it goes:

I need to detect a change in temperature on a black metal surface. Let's say that there is no other option than a contactless method. The sensor is inside a cabinet, so high attenuation to outside light

So looking at IR thermal radiation.

I do not need an absolute measurement, just a detection if the temperature rises fast

An IR diode has wavelength below 1um. Heat is primarily at over 1.5um

Example of cheap IR diode:

http://www.everlight.com/file/ProductFile/201407061702402207.pdf

Could one just use a standard IR diode to measure temperature shift (measurement range from 40 degrees to 100 degrees C)

(I have an absolute temperature in the sensor)

Cheers

Klaus
 
On 26/08/2019 09:46, klaus.kragelund@gmail.com wrote:
Hi

I am way out here on deep water, but here it goes:

I need to detect a change in temperature on a black metal surface. Let's say that there is no other option than a contactless method. The sensor is inside a cabinet, so high attenuation to outside light

So looking at IR thermal radiation.

I do not need an absolute measurement, just a detection if the temperature rises fast

An IR diode has wavelength below 1um. Heat is primarily at over 1.5um

Example of cheap IR diode:

http://www.everlight.com/file/ProductFile/201407061702402207.pdf

Could one just use a standard IR diode to measure temperature shift (measurement range from 40 degrees to 100 degrees C)
Click to expand...

Only if it is going to reach 800K or more. You would be much better off
cannibalising a PIR burglar alarm sensor and tweaking it a bit.

https://www.rapidonline.com/Catalogue/Search?tier=PIR%20Sensors

They are optimised for detecting movement or change so you might need to
interpose a chopper or do radical surgery to the sensor assembly.

Or find a supplier of the sensors for non-contact thermometers.

--
Regards,
Martin Brown
 
Use a thermopile. Hamamatsu sells them, and they're super easy to use. Pyroelectrics are okay if you chop them, provided you know where the signal is coming from when the chopper is closed--it'll be either the chopper temperature, the detector board temperature, or somewhere in between.

A thermopile comes with a built-in thermistor and it gives you a DC reading of the temperature difference between the sensor and the scene.

Cheers

Phil Hobbs
 
On Monday, August 26, 2019 at 7:18:18 AM UTC-4, pcdh...@gmail.com wrote:
Use a thermopile. Hamamatsu sells them, and they're super easy to use. Pyroelectrics are okay if you chop them, provided you know where the signal is coming from when the chopper is closed--it'll be either the chopper temperature, the detector board temperature, or somewhere in between.

A thermopile comes with a built-in thermistor and it gives you a DC reading of the temperature difference between the sensor and the scene.

Cheers

Phil Hobbs
Click to expand...

I assume you missed this part :

Let's say that there is no other option than a contactless method.
 
pcdhobbs@gmail.com writes:

They are optimised for detecting movement or change so you might need to
interpose a chopper or do radical surgery to the sensor assembly.

Probably both. PIR motion sensors use a vertically split sensor
connected differentially, with a segmented Fresnel lens to cast about
a dozen images, so that tall narrow warm (or very cold) objects create
an AC output as they move. The AC is detected and rectified to control
the switch. (It's a pretty slick scheme actually.)

It's simple to pull off the lens, but less simple to defeat the
differential action without damage. A little bit of aluminum tape
covering one half might work.
Click to expand...

A chopper close to the elements can block the PIR elements in sequence
to create AC signal. If you know the phase, you can measure if the
target is warmer or cooler than the detector, too.

Also check the element geometry; dual element or quad with 2 outputs
are much easier than quad with single output. Duals are also cheaper.

But the thermopiles you mentioned are way easier.

--
mikko
 
They are optimised for detecting movement or change so you might need to
interpose a chopper or do radical surgery to the sensor assembly.
Click to expand...

Probably both. PIR motion sensors use a vertically split sensor connected differentially, with a segmented Fresnel lens to cast about a dozen images, so that tall narrow warm (or very cold) objects create an AC output as they move. The AC is detected and rectified to control the switch. (It's a pretty slick scheme actually.)

It's simple to pull off the lens, but less simple to defeat the differential action without damage. A little bit of aluminum tape covering one half might work.

Cheers

Phil Hobbs
 
Whoey Louie <trader4@optonline.net> writes:

On Monday, August 26, 2019 at 7:18:18 AM UTC-4, pcdh...@gmail.com wrote:
Use a thermopile. Hamamatsu sells them, and they're super easy to
use. Pyroelectrics are okay if you chop them, provided you know
where the signal is coming from when the chopper is closed--it'll be
either the chopper temperature, the detector board temperature, or
somewhere in between.

A thermopile comes with a built-in thermistor and it gives you a DC reading of the temperature difference between the sensor and the scene.

Cheers

Phil Hobbs

I assume you missed this part :

Let's say that there is no other option than a contactless method.
Click to expand...

It is contactless! (optical).


--

John Devereux
 
On 8/26/19 8:32 AM, Mikko OH2HVJ wrote:
pcdhobbs@gmail.com writes:

They are optimised for detecting movement or change so you might need to
interpose a chopper or do radical surgery to the sensor assembly.

Probably both. PIR motion sensors use a vertically split sensor
connected differentially, with a segmented Fresnel lens to cast about
a dozen images, so that tall narrow warm (or very cold) objects create
an AC output as they move. The AC is detected and rectified to control
the switch. (It's a pretty slick scheme actually.)

It's simple to pull off the lens, but less simple to defeat the
differential action without damage. A little bit of aluminum tape
covering one half might work.

A chopper close to the elements can block the PIR elements in sequence
to create AC signal. If you know the phase, you can measure if the
target is warmer or cooler than the detector, too.

Also check the element geometry; dual element or quad with 2 outputs
are much easier than quad with single output. Duals are also cheaper.

But the thermopiles you mentioned are way easier.
Click to expand...

Yeah, you could do that if you can get the chopper close enough. It
would work better as well, since the DC bias on the sensor output isn't
well defined--it's basically V_GS of a cheap small JFET, which varies
all over the place.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
On 8/26/19 10:10 AM, Arie de Muynck wrote:
On 2019-08-26 10:46, klaus.kragelund@gmail.com wrote:
Hi I am way out here on deep water, but here it goes:

I need to detect a change in temperature on a black metal surface.
Let's say that there is no other option than a contactless method.
The sensor is inside a cabinet, so high attenuation to outside
light

So looking at IR thermal radiation.

I do not need an absolute measurement, just a detection if the
temperature rises fast

An IR diode has wavelength below 1um. Heat is primarily at over
1.5um

Example of cheap IR diode:

http://www.everlight.com/file/ProductFile/201407061702402207.pdf

Could one just use a standard IR diode to measure temperature shift
(measurement range from 40 degrees to 100 degrees C)

(I have an absolute temperature in the sensor)

Cheers

Klaus


Look for thermopile sensor like:
http://www.ti.com/ww/eu/sensampbook/tmp006.pdf (obsolete)

or "Non-contact Infrared Temperature IR Sensor": MLX90614 $4 at
https://www.aliexpress.com/item/32899618353.html
Click to expand...

These are the ones we use:

<https://www.hamamatsu.com/us/en/product/optical-sensors/infrared-detector/thermopile-detector/index.html>

Cheers

Phil Hobbs
--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
On 8/26/19 9:48 AM, George Herold wrote:
On Monday, August 26, 2019 at 4:46:13 AM UTC-4, klaus.k...@gmail.com wrote:
Hi

I am way out here on deep water, but here it goes:

I need to detect a change in temperature on a black metal surface. Let's say that there is no other option than a contactless method. The sensor is inside a cabinet, so high attenuation to outside light

So looking at IR thermal radiation.

I do not need an absolute measurement, just a detection if the temperature rises fast

What is 'fast'? The pyroelectric IR sensors I worked with in the
distant past were kinda slow... milliseconds.
(But they may make faster ones these days.)
Click to expand...

Macroscopic objects aren't going to heat up any faster than that anyway.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com
 
On Monday, August 26, 2019 at 4:46:13 AM UTC-4, klaus.k...@gmail.com wrote:
Hi

I am way out here on deep water, but here it goes:

I need to detect a change in temperature on a black metal surface. Let's say that there is no other option than a contactless method. The sensor is inside a cabinet, so high attenuation to outside light

So looking at IR thermal radiation.

I do not need an absolute measurement, just a detection if the temperature rises fast
Click to expand...

What is 'fast'? The pyroelectric IR sensors I worked with in the
distant past were kinda slow... milliseconds.
(But they may make faster ones these days.)

George H.
An IR diode has wavelength below 1um. Heat is primarily at over 1.5um

Example of cheap IR diode:

http://www.everlight.com/file/ProductFile/201407061702402207.pdf

Could one just use a standard IR diode to measure temperature shift (measurement range from 40 degrees to 100 degrees C)

(I have an absolute temperature in the sensor)

Cheers

Klaus
Click to expand...
 
On 2019-08-26 10:46, klaus.kragelund@gmail.com wrote:
Hi

I am way out here on deep water, but here it goes:

I need to detect a change in temperature on a black metal surface. Let's say that there is no other option than a contactless method. The sensor is inside a cabinet, so high attenuation to outside light

So looking at IR thermal radiation.

I do not need an absolute measurement, just a detection if the temperature rises fast

An IR diode has wavelength below 1um. Heat is primarily at over 1.5um

Example of cheap IR diode:

http://www.everlight.com/file/ProductFile/201407061702402207.pdf

Could one just use a standard IR diode to measure temperature shift (measurement range from 40 degrees to 100 degrees C)

(I have an absolute temperature in the sensor)

Cheers

Klaus
Click to expand...

Look for thermopile sensor like:
http://www.ti.com/ww/eu/sensampbook/tmp006.pdf (obsolete)

or "Non-contact Infrared Temperature IR Sensor":
MLX90614 $4 at https://www.aliexpress.com/item/32899618353.html

etc...

Regards,
Arie
 
On Mon, 26 Aug 2019 16:10:48 +0200, Arie de Muynck
<no.spam@no.spam.org> wrote:

On 2019-08-26 10:46, klaus.kragelund@gmail.com wrote:
Hi

I am way out here on deep water, but here it goes:

I need to detect a change in temperature on a black metal surface. Let's say that there is no other option than a contactless method. The sensor is inside a cabinet, so high attenuation to outside light

So looking at IR thermal radiation.

I do not need an absolute measurement, just a detection if the temperature rises fast

An IR diode has wavelength below 1um. Heat is primarily at over 1.5um

Example of cheap IR diode:

http://www.everlight.com/file/ProductFile/201407061702402207.pdf

Could one just use a standard IR diode to measure temperature shift (measurement range from 40 degrees to 100 degrees C)

(I have an absolute temperature in the sensor)

Cheers

Klaus


Look for thermopile sensor like:
http://www.ti.com/ww/eu/sensampbook/tmp006.pdf (obsolete)

or "Non-contact Infrared Temperature IR Sensor":
MLX90614 $4 at https://www.aliexpress.com/item/32899618353.html

etc...

Regards,
Arie
Click to expand...

Here's another one:

https://www.adafruit.com/product/1748

Omega makes high-end stuff, $200 and up.

Handheld IR thermometers are cheap on ebay or Amazon.
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top