DIY-Thermocam V2: A do-it-yourself thermal imager

maxbot:
my problem is that i dont know how to configure the sensor without this evaluation board (it costs about 250 dollar -.-)
can somebody help ?

I just came across this at sparkfun http://www.sparkfun.com/products/9813($50). Not sure if this is the same thing?

Question.. How do you think the IR sensor sold at sparkfun - MLX90614ESF-BAA - would work in your project, you specify a MLX90614ESF-DCI. I already own the one from sparkfun and was wondering if I could get away with using it?

Thanks for posting your project.

Jumbar:

maxbot:
my problem is that i dont know how to configure the sensor without this evaluation board (it costs about 250 dollar -.-)
can somebody help ?

I just came across this at sparkfun http://www.sparkfun.com/products/9813($50). Not sure if this is the same thing?

Question.. How do you think the IR sensor sold at sparkfun - MLX90614ESF-BAA - would work in your project, you specify a MLX90614ESF-DCI. I already own the one from sparkfun and was wondering if I could get away with using it?

Thanks for posting your project.

As per datasheet, xCI models are the best one, as they have the narrowest FOV and an internal "temperature gradient compensation system" which allows further narrowing.

By the way, reaplacing the BAA by the DCI should be easily possible.
Does sparkfun have resellers in europe?

AWOL:

because it wouldn't work; ordinary glass mirrors don't reflect heat (long IR).

What about surface-coated mirrors?
Glass doesn't transmit the IR well, but if it comes off the metal layer directly?
Old laser printers and scanners can be a good source of these.

Sorry about the lag in response time - I didn't see this for some reason.

I don't think a front-surface mirror would work, either; from the research I did, I never ran across anything about using a front-surface mirror, but then again everything was about the sensors and such, using special long IR transmissive lenses and elements made from something a bit more exotic than usual (thus a part of the higher price). I'm not really certain though; perhaps something to look into (if this could work, then you could potentially have a much higher scan rate, up to the limit of the sensor itself - which I understand isn't "high speed", but maybe it could be made better - hmm, I wonder what peltier cooling would do for speed?)...

:slight_smile:

Older longer wavelength thermal imagers used germanium optics (read "expensive"), but internally used mechanically-scanned facetted mirrors to produce the image across relatively few sensors.
Maybe the mirrors are simply polished metal.

How about:
http://cgi.ebay.co.uk/20mm-25mm-Gold-Plated-Reflection-Mirror-for-CO2-Laser_W0QQitemZ160387764155QQcategoryZ26219QQcmdZViewItemQQ_trksidZp4340.m263QQ_trkparmsZalgo%3DSIC%26its%3DI%252BC%26itu%3DUCI%252BIA%252BUA%252BIEW%252BFICS%252BUFI%26otn%3D15%26pmod%3D400077108529%26ps%3D63%26clkid%3D8039527281106968300
?

Lovely, thanks for sharing!

I think a huge improvement will be to use a digital micromirror device capable of reflecting IR (if exists) and then go for the approach described here

http://dsp.rice.edu/cscamera

i.e. implement a single pixel camera

Hello,

I'm currently starting to build this, the sensor is on it's way. It's not -DCI but MLX90614ESF-BCF - they didn't have DCI in stock right now. But I'm trying with this and see how it goes.

One helping thing for scanning speed would be to limit the "focus"-area after the first initial scan. As I'm mostly interested in monitoring "heat buildup" during my electronics tests (for example electric vehicle conversion motor & controller), I could easily point the areas that interest me and limit the scanning around them. So to me, it would be nice to have first scan in low resolution and after that "focus" the scanning on the most heated spots either automatically or manually. That way we could get much higher scanning rate for the areas we are interested in and much lower scan rates for "background" areas.

I'm certainly going to try this when the sensor arrives.

With best regards from Finland,

Henkka

The Cheap-Thermocam is now three times faster than before thanks to adjusting the sensors internal settings !
Additionally, I improved the software and plan to build a portable version with LCD Touchscreen, CMOS camera, SD card and lithium battery. It will be standalone from any pc as the thermal image can be viewed on the display (resolution 320x240 for the display, on the pc its highter) and later analyzed with the computer software.

@martin_bg: Thanks for the information. I currently try to understand how the compressed sensing theory is working, its not that easy :slight_smile:
@oh6kft: Have a look at my post at the other thread: Problem with Reading/Writing to the EEPROM of a MLX90614 - #6 by maxbot - Sensors - Arduino Forum . WIth the 10 degree version of the MLX90614 its hard to get comparable pictures than with the DCI (I have a BCF here..). Futureelectronics ordered about 50 new parts and they promised me that those will be available soon.

Fantastic project by the way got the rest of the parts but having a problem soucing the
MLX90614ESF-DCI is there alternative that would give the same results FutureElectronics are saying there is 22 week leadtime to get this sensor from Melexus ? :frowning:

SPI-Investigator:
Fantastic project by the way got the rest of the parts but having a problem soucing the
MLX90614ESF-DCI is there alternative that would give the same results FutureElectronics are saying there is 22 week leadtime to get this sensor from Melexus ? :frowning:

There is no alternative (in this price range). You can buy the sensor for 48$ plus 5$ international transportation costs by writing an email to info@as-electronic.net. This is the fastest way to get the DCI version and as far as I know the only one, too.

Thank very much for your prompt reply I had feeling there was no alternative ! I shall email the company you mention save having to wait a long time from future electronics :slight_smile: Your help is much appriciated ! Keep up the good work.

Just an update it would appear Melexis are having some quality problems with this IR Temperature Sensors below is copy of the email I got back from them yesterday. As I already have all the parts I have emailed them back and they will let know when they have delivery date. They seem very good company to deal with !

Email from as-electronic.net
Hi Johan,

I have bad news, at the moment. We sold last week all of our stock. Melexis has quality problem with the IR Temperature sensors with the 5°FOV. We allready ordered the MLX90614ESF-DCI but we don´t have a delivery date at the moment.
I will contact you, when I get the delivery date.

Best regards,
Matthias

I was browsing today and found this thermal sensor, do you think it is useful for this kind of project given the limit of 4-100 degrees C only

http://www.robot-electronics.co.uk/acatalog/Thermal_Array_Sensor.html

martin_bg:
I was browsing today and found this thermal sensor, do you think it is useful for this kind of project given the limit of 4-100 degrees C only

http://www.robot-electronics.co.uk/acatalog/Thermal_Array_Sensor.html

The Field of View is much too high, have a look at this project which uses the TPA81: Thermoscanner

Email from As-Electronics

Hi Johan,

Thank you for your mail. Today I received a delivery from Melexis the MLX90614ESF-BCI. The MLX90614ESF-DCI is specialy made for medical reasons. On page 31 in the datasheet you can find the differents.
For the MLX90614ESF-DCI, I don´t have a delivery date yet.

Best regards,
Matthias Schenzinger

Vertriebsingenieur
Sales Engineer

I just wondered if the MLX90614ESF-BCI if could used in its place as still no delivery date on the MLX90614ESF-DCI . Spec's seem very similar according to the datasheets the only difference seems to be the supply voltage . Other wise will have to wait a little longer for the exact ir sensor.

SPI-Investigator:
Email from As-Electronics

Hi Johan,

Thank you for your mail. Today I received a delivery from Melexis the MLX90614ESF-BCI. The MLX90614ESF-DCI is specialy made for medical reasons. On page 31 in the datasheet you can find the differents.
For the MLX90614ESF-DCI, I don´t have a delivery date yet.

Best regards,
Matthias Schenzinger

Vertriebsingenieur
Sales Engineer

I just wondered if the MLX90614ESF-BCI if could used in its place as still no delivery date on the MLX90614ESF-DCI . Spec's seem very similar according to the datasheets the only difference seems to be the supply voltage . Other wise will have to wait a little longer for the exact ir sensor.

I think you can use the BCI version without any problems. The supply voltage is also the same (3V). I don't know exactly if you can use the EEPROM Sketch with this version, but as it also work with other, more different models i hope there won't be any issue.

Max
Congratulations on winning the Jugend forscht award. It's a great idea. I have built one but am having problems with it.
I can download the software to the eprom and if I open the serial port monitor and send a "1" it will go through the scanning process ok. When I fire up the "JAR" file it opens up, grabs camera frames ok and when I press "START" the "remaining time: 70 seconds" screen comes up but nothing happens. The log file says "Error the arduino is not correctly connected". Now this is where I'm getting confused. Does all the output data go to the computer via the USB serial port or is an RS232 port (Null modem??) also needed? I guess I'm asking what port the "JAR" file expects to talk to the arduino board or do you have a connection diagram? Oh, while I'm here there may be a bug on your home page. The schematic sensor pinouts are back wards I.E. 1234 should be 4321 as pin 4 is GND and 3 is 3v3.
Cheers
Stephen

inmyshed:
Max
Congratulations on winning the Jugend forscht award. It's a great idea. I have built one but am having problems with it.
I can download the software to the eprom and if I open the serial port monitor and send a "1" it will go through the scanning process ok. When I fire up the "JAR" file it opens up, grabs camera frames ok and when I press "START" the "remaining time: 70 seconds" screen comes up but nothing happens. The log file says "Error the arduino is not correctly connected". Now this is where I'm getting confused. Does all the output data go to the computer via the USB serial port or is an RS232 port (Null modem??) also needed? I guess I'm asking what port the "JAR" file expects to talk to the arduino board or do you have a connection diagram? Oh, while I'm here there may be a bug on your home page. The schematic sensor pinouts are back wards I.E. 1234 should be 4321 as pin 4 is GND and 3 is 3v3.
Cheers
Stephen

Hi Stephan,
its great you built my project :slight_smile: The Java Software talks to the Arduino via the "virtual" USB Serial port (no extra port is required :smiley: ). For this, the rxtx java library from rxtx.org is required. Have you already installed this library properly ? Follow the install instructions on their site and you will be fine i hope..
Concerning the schematic: I copied the sensor bottom view graphic from the official datasheet and therefore didn't adjust my own schematic to this one. I will fix the bottom view to avoid any confusion.
If you have further questions, feel free to write me an email (contact data on the official website).

Update:
There is a new software version (2.0) which includes several new features and bugfixes.
I created a Printed Board so that no soldering will be required to build the device ( -> http://cheap-thermocam.lima-city.de/images/Board.png) and an enclosure to make it stable and nice-looking ( -> http://cheap-thermocam.lima-city.de/images/enclosure.png).
Both components will be available as soon as I have tested them for the material price of 20$ each.