" Old hands " Thermal imaging project.

[Grumage]

The floor is yours Chet. 

[Chet K]

Thanks Grum

here any and all input towards developing an inexpensive thermal imaging tool for members here [and elsewhere] to use in experiments

is welcome and appreciated

a tidbit to start


https://hackaday.io/project/6416-raspberry-pi-thermal-imaging

https://hackaday.com/tag/thermal-imaging/

[ion]

Try this, may be "good enough"

http://www.ebay.com/itm/Thermal-camera-imaging-software-for-FLIR-cameras-may-work-for-other-models-EU-/301542483062?hash=item4635552476:g:Cu4AAOSw8d9UuC~K

"If not satisfied, keep the software and ask for refund - no problem!"

Worth a try

[Chet K]

I agree
I believe we will find many options to explore here

[TinselKoala]

http://www.eevblog.com/forum/thermal-imaging/

[Slider2732]

The demo video was very interesting.
It showed the limits of resolution, but also that it does have some resolution.
64 x 16 pixels I think, shown as a strip across the middle of the image.

Being I2C as it's communications protocol, it likely will return values per 'pixel'. From which, the program running on the host microcontroller could switch on/off the heating element for the project in the other thread

[Chet K]

Mark sent me this on the thermal imaging "job"

http://www.ebay.com/itm/MLX90614ESF-BAA-000-TU-ND-Infrared-Thermometer-Module-IR-Sensor-for-Arduino-/142234923646?hash=item211ddcb27e:g:MzUAAOSw-0xYbh9g

http://www.instructables.com/id/Thermal-Camera/?ALLSTEPS

[Slider2732]

Indeed Chet,
The MLX90614 is available for just $5.17 ! ($9.49 from USA suppliers) and has 4 pins on it. 3.3V, Ground, SCA and SDL.
That's the I2C protocol, they are intended for hooking up to microcontrollers such as the Arduino or Raspberry Pi. An Arduino Pro-Mini will run at 3.3V, even if it has a 16MHz crystal, just connect to the RAW Pin. That negates the need for level shifting.
As such, they work similarly to a thermistor, but, if the Instructable is followed, the use of a couple of servo's can create a scanning method. That scanning can be likened to an old CRT TV and will form an image based on the readings from moving.
Not ideal, but $100's cheaper than FLIR thermal imagers.


This is also a very interesting project, using the same sensor:
http://www.instructables.com/id/Build-a-Thermal-Flashlight-Light-Painting-with-T/?ALLSTEPS
He puts RGB LED's on the front of his flashlight, to show the temperature differences where the flashlight illuminates.

[ion]

Depending on how accurate and robust you want the end product, many factors will need to be considered when designing an optical pyrometer and temperature control "system".

Before redesigning the wheel, I would familiarize oneself with current optical methods for temperature control, then also consider that you will be driving a transformer load which can be tricky if you want smooth phase angle control vs "bang bang" on-off control methods.

http://www.ebay.com/itm/Phase-Angle-SCR-Driver-Board-with-Current-Limit-for-large-SCRS-Inductive-loads-/162023259453?hash=item25b956bd3d:g:OuQAAOSwAvJW-~JO

Coding a good PID loop or fuzzy tuning algorithm can also be a bit tricky.

Fortunately the temperature control component to do this with PID as well as autotuning has been  available off the shelf for $10 to $15 USD.

http://www.ebay.com/itm/SSR-220V-Digital-PID-Temperature-Controller-Thermocouple-K-Dual-LED-Display-/222371957216?hash=item33c666c9e0:g:8RgAAOSwEzxYSnXJ

You need also to consider that if the optical device is not well focused on the target or if someone bumps the optical input device away from the target, the temperature control loop will be fooled into thinking the temperature is just low and will need more power. This can be dangerous and will require special handling in the coding to prevent large accidental over temperature problems. If the scanning method is used, it will need mechanical control as well as a method to keep the controlled area of interest as the target area, some auto tracking method. IMO, a contact measurement method might be less costly, safer and easier to implement, although the non contact scanning method can have some advantages at the sacrifice of absolute accuracy.

I spent a good portion of my adult life designing such uP based controllers as well as contact and non-contact measurement methods, and indeed there are pits ahead to be avoided, therefore requiring some forethought. It can be done, and yes it can be done with Arduino and other platforms if you've got lots of time on your hands for coding and debugging the mechanical scanning interface as well as the temp control methods.

Just purchasing and mating a IR sensor to an Arduino is the easiest part.

Edit: Looking over the data sheet on the MLX90614, it seems like a good choice for the task in it's present conception.

[Chet K]

The Flir one for android,  240 odd dollars gets a very useful tool ,quick review below


https://www.youtube.com/watch?v=QZ-RIhiBtpE

and there is a wonderful fellow at EEV forum doing some open source work to expand this resource [THX to TK for mentioning ]

http://www.eevblog.com/forum/thermal-imaging/question-about-flir-one-for-android/275/

http://www.eevblog.com/forum/thermal-imaging/question-about-flir-one-for-android/275/

[ion]

Using a standard camera with a few mods:

http://www.instructables.com/id/infrared-digital-camera---the-real-way/

http://www.wrotniak.net/photo/infrared/

and an interesting $10 dollar camera hacked to IR

https://www.youtube.com/watch?v=IdJgxNVVrs4

and more:

https://www.youtube.com/watch?v=1oGN0tVLeiw

Maybe not high quality FLIR but hey!