IR Photodiode project

[shane ren]

The overall aim of the project is:
(1) Amplifier design and calibration of a non contact thermal sensor
(2)Generation of thermal heat maps
(3)Resolution testing
(4)integration of motors, non-contact thermal sensor measurments and control of thermal source.

So the overall aim of the project is to design and construct a system capable of producing a thermal map of different heat sources. So there will be a heated base with various amounts of peltiers which all have different temperatures, a stepper motor will then work its way all around the heated base. The non contact thermal sensor will be applied to the stepper motor to record these temperatures. My part of the project is the non contact thermal sensors. I have been out of the educational system for three years and I am severely struggling with the practical element of my course which is really holding me back from passing my final year. Any help would be greatly appreciated, I have decided to use a IR photodiode.

 

[jpanhalt]

What temperature or IR wavelength are you interested in measuring? Do IR photodiodes work at the wavelength/temperature in which you are interested?

 

[dknguyen]

You said you have decided to use an IR photodiode. I have to ask if you've really investigated whether it's actually suitable or whether you just chose it because it has "infrared" in the name.

Normally, thermopiles are used for non-contact temperature measurement rather than IR photodiodes because thermopiles measure long-IR which is heat and photodiodes only measure near-IR. Long-IR is actually radiated heat. This is what temperature guns use and doctor ear thermometers.

If you do decide to look at thermopiles, this is the best one for your purposes, as far as spectral qualities are concerned, that is actually available in low quantities:
https://ca.mouser.com/ProductDetail/Measurement-Specialties/G-TPCO-031?qs=sGAEpiMZZMueJGT/0PGc46Y5IhofrXIbTLoMw%2b48BxI=

It's also available from Digikey but is far more expensive there, especially if you need 10 or more.

That doesn't mean there aren't also others also available that might also be cheaper. What makes this one different is it comes with an 8-14um filter which rejects the effects of moisture in the air and lets it work more effectively over long ranges which may or may not be of concern to you (matters most when humidity is high or when measuring over long distance). This filter is usually a special option normally only available if you custom order thermopiles in large quantities. I spent about 8 years searching and waiting for a thermopile with this filter that didn't need to be purchased in quantities of 100+ before this one became available.

What might not be desirable for your application is the rather wide FOV of this particular thermopile if you want to be far away and measure the temperature of a small object.

Don't forget thermopiles measure RELATIVE temperature (usually relative to their own temperature).

 

[jpanhalt]

I was thinking along the same lines. The term "thermal map" made me wonder whether the TS meant something more along the lines of an image. FLIR sells its microbolometers on a breakout board (Lepton, https://store.groupgets.com/products/l3). I believe it is also available from Digikey and Mouser. I have not seen the Seek sensor advertised separately; although, the entire Seek camera for smartphones/tablets is about the same price.

Just FYI, while looking for a thermal camera, I came across a new start-up company whose approach is quite a bit different. Apparently, it has developed a material that changes its RI/reflectivity when exposed to long wavelength IR. It then bounces ordinary visible light and detects the image from that. So far as I could tell (about 1 month ago), it is not for sale yet.

John

 

[dknguyen]

Just FYI, while looking for a thermal camera, I came across a new start-up company whose approach is quite a bit different. Apparently, it has developed a material that changes its RI/reflectivity when exposed to long wavelength IR. It then bounces ordinary visible light and detects the image from that. So far as I could tell (about 1 month ago), it is not for sale yet.

John

So like some kind of magic mirror? My thoughts immediately went to whether this would work on vampires or not.

 

[jpanhalt]

Close. Maybe better for ghosts.

I'll see if I can find it again and post the link as an edit.

EDIT:
I found a description. However, the date is earlier than I recall. Twelve years seems a bit long. The term I recalled was "red shift" being used in a different manner than in astronomy.

These links are to the same article. I wasn't sure the second link to the pdf would work for everyone:
http://www.sciencedirect.com/science/article/pii/S0961129006717641#fig2
**broken link removed**

This old advertisement might be of interest too:
http://www.securityinfowatch.com/pr...ical-wavelength-converter-for-thermal-imaging

 

[shane ren]

Sorry for the delay, I was not getting notified about these replies. Thank you very much for all your help. The peltier is going to produce temperatures within the range of 0C - 95C max. So when I am checking photodiodes and IR led's I am looking at the operating temperature. Would I need to look into its wavelength too? The sensor will be placed pretty close to the heat source, so less than a 100cm definielty. I also cannot use thermpiles as somebdy else in the class has already chosen that. My lecturer advised me to use an IR photodiode or a thermoelectric sensor.

Would this analog sensor be of any use to me if I was to copy it?
https://maxembedded.com/2013/08/how-to-build-an-ir-sensor/

 

[dknguyen]

I also cannot use thermpiles as somebdy else in the class has already chosen that. My lecturer advised me to use an IR photodiode or a thermoelectric sensor.

I question your lecturer right there. Not to mention that a thermopile IS a thermoelectric sensor so I have no idea what your lecturer is going on about. There are only so many methods to measure temperature, let alone non-contact methods. You can't just force one group to use a much more difficult or unfeasible method (or some other method that doesn't even exist) just because another group has already chosen the only straightforward or feasible method. Does your lecturer even know what he's talking about?

I'm just going to come out and say it: As far as I am aware, you can't use near-IR to measure temperature. All IR photodiodes measure near-IR. Look at the photo below. The person is the same darkness as their surroundings, sometimes darker. That's definitely not temperature. Politely points these things out and ask your lecturer to be more specific. I suspect he knows less than he's letting on.
https://www.nickspiker.com/ilwp/wp-content/uploads/2016/05/Jill.jpg

I would definitely challenge him on that or go a different route on this project if possible. It sounds like an unfeasible, if not impossible task. Not all IR is heat. IR just means redder than red. For heat you need long IR and photodiodes don't do that. Thermopiles do.

The next most feasible is completely unfeasible and is pyroelectric sensors. But that's very complicated and involved because pyroelectric sensors don't measure DC temperature like thermopiles do. They measure AC temperature so you have to have a spinning wheel with slots or something that modulates the exposure of the pyroelectric sensor to the heat source. Once you get all that done, , then you have to calibrate the thing since the readout will depend on the speed you are chopping at and the temperature of the object. From what I've seen, single element pyroelectric sensors are also very very expensive $100+ and difficult to get ahold of. It seems they almost always have to be special ordered. The cheap pyroelectric sensors that you can buy easily and cheaply are for motion detection and have two elements so you can't use them for temperature measurement. Pyroelectric sensors are also very sensitive to wiring noise and require careful design.

Also, is there a language barrier here? You said you are looking at operating temperature, but that is just how hot the sensor can get. It's not the same as how hot the sensor can measure. For example, you can use hold a heatgun in your hand and use it to measure something that is thousands of degrees.

 

[shane ren]

Do you reckon the use of thermopiles would be alot easier to go about? I could talk to him tomorrow about it. All I want to achieve is a pass in this project, so a straight forward working project would be good enough for me. I cannot seem to find anything on IR photodiodes anywhere which is really starting to stress me out.
No theres no language barrier, but because of the fact that I have been out for so long, everything has seemed to have left my head, so I feel like a beginner again haha.

 

[jpanhalt]

That photodiode will not detect the temperature range you are interested in.

Wavelength vs. temperature for a blackbody is well defined. Here is a chart:https://www.icc.dur.ac.uk/~tt/Lectures/Galaxies/Images/Infrared/Regions/irregions.html

0°C (ca 273°K) to 95°C(368°K) will emit light in the micron range -- roughly 5 u to 25 u (probably closer to 14 u).

 

[jpanhalt]

You could try a thermopile. Or get one of the very cheap, non-contact thermometers at Harbor Freight and scavenge its parts.

 

[dknguyen]

Do you reckon the use of thermopiles would be alot easier to go about? I could talk to him tomorrow about it. All I want to achieve is a pass in this project, so a straight forward working project would be good enough for me. I cannot seem to find anything on IR photodiodes anywhere which is really starting to stress me out.
No theres no language barrier, but because of the fact that I have been out for so long, everything has seemed to have left my head, so I feel like a beginner again haha.

A LOT easier. With a thermopile you pretty much just hook it up to a non-inverting op-amp and make sure the op-amp has very low bias currents.

Pyroelectric sensors are a project unto themselves (and a nightmare for parts sourcing). And near-IR won't do what you want.

You could try a thermopile. Or get one of the very cheap, non-contact thermometers at Harbor Freight and scavenge its parts.

His lecturer said he can't because another group was already using thermopiles. FFS. Ridiculous to the max.

 

[jpanhalt]

Here's a link to an online calculator: https://www.ajdesigner.com/phpwien/wien_equation.php

50°C = 9 micron wavelength. Frequency is the usual C/wavelength

 

[shane ren]

I have sent an email asking about the use of thermopiles so I should know by tomorrow afternoon. Fingers crossed. I will have a total of 6 weeks to assemble the project and produce a thermal map, hopefully it is all possible. I will be using labview as my hardware program, are any of ye familiar with this? if so is it definitely possible to achieve a thermal map using this? I also have the option of matlab.

 

[dknguyen]

I have sent an email asking about the use of thermopiles so I should know by tomorrow afternoon. Fingers crossed. I will have a total of 6 weeks to assemble the project and produce a thermal map, hopefully it is all possible. I will be using labview as my hardware program, are any of ye familiar with this? if so is it definitely possible to achieve a thermal map using this? I also have the option of matlab.

If you ultimately want photodiode materials for whatever reason, I can send you some when I get off work. They really are hard to find or are rather obscure. It's mostly scraped together from a few application notes, a slide lecture, and a big book about optical systems that I found online. The book has a lot of black-box analog design mumbo jumbo that's beyond both our depths, but you can scrape some things from parts of it. It gives you a good idea of what you need to do for the simplest photodiode amplifier. A photodiode amp is actually a really good and enlightening project on it's own...just not for temperature measurement.

I will be using labview as my hardware program, are any of ye familiar with this? if so is it definitely possible to achieve a thermal map using this? I also have the option of matlab.

Used both. No longer remember anything about either.

 

[dknguyen]

My photodiode materials.

In the book, Chapter 18, Front-Ends is probably the most useful. It is the full length version of the document "Photodiode Front-Ends".

Also, "Transimpedance Amplifier" is the general term you need when looking up photodiode amps.

 

[schmitt trigger]

dknguyen:
If it were possible to give you three or four stars to thank you for sharing these incredibly valuable materials I would.
The book is absolutely priceless. Not only does it provide a lot of very good technical information, it does it with the right amount of philosophy and humor.

In this respect, I particularly liked the J.W. Garner quote on the first chapter:

An excellent plumber is infinitely more admirable than an incompetent philosopher. The society which scorns excellence in plumbing because plumbing is a humble duty and tolerates
shoddiness in philosophy because it is an exalted activity will have neither good plumbing nor good philosophy.
Neither its pipes nor its theories will hold water.
—John W. Gardner†

 

[shane ren]

I agree, that is a very enjoyable piece to read which contains a lot of useful information. Thank you very much for going out of your way for me there, really appreciate it. My lecturer told me today that I may have the option of using a thermopile. I am nearly half tempted to go ahead with the photodiode and see if I can tackle the problem by using an IR led with it.

 

[jpanhalt]

You will not get a thermal mapping over the required temperature range using ordinary IRED's or IRED detectors. Of course, you will have something physical to turn in. It won't work, but you knew that going into the project. Maybe the professor will give you a pass for effort. Comment: Sometimes students don't appreciate the flak a professor gets at some schools when a student is failed. Thus, the professor is conflicted between being honest and keeping his position. Our social networking has made it easier for students to take advantage of that situation. Some might say, no harm, no foul.

If you really want to try something that might work in theory, you could coat the Peltier device with a substance that reversibly changes color with temperature. Then scan the images in visible light and use Matlab to convert the colors to a temperature map. If you can't paint the unit, then you might be able to use a flexible film; however, stress from flexing might change the colors too (think of a liquid crystal display).

The trick will be to find the "thermochromic" liquid crystal paint with the right temperature range.

This might help: https://www.thermometersite.com/technology

Source: cited above
Thermochromic liquid crystals, similar to liquid crystal displays, are color changing crystals that undergo a reversible thermally induced color change. Traditional TLC’s are microencapsulated and coated on a black backing. As they approach their temperature rating they pass through the colors of the spectrum in sequence from tan to green to blue before turning back to black. They are calibrated to display green at a rated temperature. TLC’s can possess acute thermal sensitivity, detecting temperature changes as small as 0.2ºF and are available from -22⁰F to 248⁰F. LCR Hallcrest has developed and offers a single color liquid crystal that transitions from black to color without going through the colors of the spectrum, making it easier to read.

If something like that appeals to you, I suggest that you do a more thorough search for such coatings. They may even be made with an elastomeric vehicle so they can be painted on, allowed to "dry", then peeled off easily when finished.

John

 

[shane ren]

Thanks very much John, never heard of that before. Sounds very interesting to experiment with. But like yu said earlier, I think what the lecturer wants to see is hard experimental work put into the project, so even if my project isnt able to perform the way I want it to, well then thats okay. As long as I can show meaningful ways in which I tried to overcome all these problems. By having the method which you just informed me about sounds mind blowing, I really think they would like to see would that benefit the sensor in anyway. I know photodiodes might not be the right tool for the job, but I am looking forward to investigating into these in further detail. Thank you very much to John and Dknguyen for all your help, really would not be half as comfortable as I am about the project now. I also have a nice few bits to look back on and to inseert into my project. Keep up the good work and once again THANK YOU.