Thermal images


riumplus

Still Fresh
Joined
May 6, 2009
Messages
32
Location
Brisbane, Australia
Website
www.riumplus.com
Hi everyone,


So I have access to a thermal camera at work. I don't follow these boards too closely but I don't think anyone's ever taken some thermal images of a Pandora... So I thought I should share some snapshots!


First, my testing methodology: my pandora was overclocked to 800MHz OPP4, wifi was turned on & connected to my access point but I wasn't transferring data. The screen was put on full brightness and the speakers at full volume. I had Chromium open on a website with a lot of animations to consume 100% CPU while I played a video with Panplayer. My Pandora was left to sit on a pillow without being touched for 1 hour while I watched some shows so the heat from my hands didn't disturb it. After one hour my Pandora was moved with insulated mittens to a wooden table for photographing.


Some additional thermally-relevant information: the thermal camera has an 80x80 sensor, accuracy to 0.1'C and all temperatures below are in Celsius (sorry Farenheit users but science calls for sensible scales :p ). The ambient temperature indoors was 24'C with 50% humidity. All images have been pre-calibrated to these environmental settings and I've assumed a surface emissivity of 0.95 which should be close to the matte black plastic of the Pandora (but not necessarily the buttons, or the screen, or even the wooden table). I'm auto-scaling the temperature scale between images to show up the best visible contrast so a certain shade of orange isn't the same amongst all images.


And now, on to the images! All images have been pre-scaled to 480px high for your Pandora-wallpaper-setting pleasure ;)


-The front view with the screen open. It's peaking in the center at 35.4'C.


pandora_thermal1.jpg



-Closed from the front. You can see the heat spot on the back on the screen peaks at 33.2'C.


pandora_thermal2.jpg



-The underside from the rear. I didn't line up the Pandora's position on the table exactly so there's a heat ghost left on the table. The heat is mostly concentrated at the rear of the units, with the metal on the USB ports peaking at 35.6'C


pandora_thermal3.jpg



-the underside from the front. If you're comparing this to the image above, don't forget I'm auto-scaling the temperature/colour scale. You can see a small localised hot-spot around the screws holes. I had two SD cards installed and one is notably hotter - guess which one contained the video I was watching!


pandora_thermal4.jpg



...More to come in part 2, I can only post a few photos at once. :)
 
And now for part 2. I closed all applications, disconnected & shut off the wifi, underclocked to 400MHz opp1, closed the screen (but did not suspend it) and plugged it in to charge with the official charger. As I expected the heat to concentrate at the bottom of the device, this time I suspended the Pandora 30cm in mid-air, touching only the ends of two books.


-Closed, from the front. The top reached only 31.2'C, the SD cards up to 32.8'C


pandora_thermal5.jpg



-Open. The heat signature is in a slightly different place to the previous open image, guess which chip is where! ;p This time the temperature peaked at 39.0'C


pandora_thermal6.jpg



-upside-down from the back. Surprisingly the back of the batteries reached only 31.1'C. The charge ended at 80% so maybe the battery would heat up hotter if I ended the charge at 100%. Once again the USB port is the hottest at 35.5'C


pandora_thermal7.jpg



-the bottom and the front. 32.6'C for the battery back, though the hottest part in the image is the SD card at 33.7'C. Suspending it in the air cools the battery well.


pandora_thermal8.jpg



Aaand that's all I could think of! Let me know if there's any other photos someone would like to see. :D
 
Looks like the Pandora is pure Lava. :D But since the Max Temp is only 40°C or so, I guess it is OK.


Most heat the Battery produces during charging, I noticed with my own Pandora too. But nothing wild.
 
You can calibrate emissivity. You can do that with a thermometer. Given the temperature of Pandora measured with the thermometer you just have to chose the emissivity that gives that same reading in the camera. Another way is using the pieces of materials that comes with the camera (cameras for thermography come with this). Those usually are tiny discs with a declared emissivity. So you can leave one of this discs on the spot you want to know emissivity. You can read temperature on this disc with the camera (you have the emissivity). If you can assure some good thermal equilibrium, remove the disc and adjust emissivity of the spot with the temperature that you previously read on the disc.


This are very interesting measurements you have made.
 
Ha, I knew (I knew!) my battery felt warmer after an extended period of play; I thought I was just crazy, though.


I would think a thermal camera would lead to all sorts of fun and shenanigans.
 
Quick and Dirty GIMP job that needed to be done, hopefully most people get it.


pandora_thermal_pred.jpg



hmm very quick..
 
Last edited by a moderator:
^ explain?


Also amazing pictures! I wonder if this would help with case cracking? instead of measuring heat maybe measure cold? (absence of heat)
 
^ explain?


Also amazing pictures! I wonder if this would help with case cracking? instead of measuring heat maybe measure cold? (absence of heat)

Here yer go! cold in white, warm in black. Ctrl-I in photoshop!


pandora_thermal6_inverted.jpg
 
Cool pics! ... I mean warm pics... er... "awesome". :D


TrashyMG: also awesome lol.
 
Trashy: "If it bleeds, we can kill it" ;)


The thermal camera is very sensitive, 0.1'C shows up a lot more than you'd think. The small heat marks on the table in that image are from when I momentarily brushed my fingers against the wood. My fingers never stopped, just a fraction of a second of moving contact, but that's enough to leave some heat marks!


The emissivity should be mostly correct for the case's matte black plastic. Anything even remotely glossy (buttons, screen & even oily fingerprint marks on the black plastic) will act like a mirror & reflect some of the room's ambient temperature. In fact even the black plastic reflects a little so I've taken that into account. Given my living room has a very varied thermal profile, it's difficult but possible to calibrate for the screen & almost impossible to calibrate for the curved buttons.


Oh, and if EvilDragon or CraigIX read this thread: If you want some more thermal photos of a Pandora for publicity or troubleshooting purposes, assembled or in pieces, the same angles as above or from other angles, in this colour palette or any other, send me a PM :)
 
Honestly, if I had a thermal camera, I'd be taking pictures of everything I could get my hands on.


Computers. Car engines. Turkey sandwiches. The sun.
 
Slight diversion, but if I point my camera at the sun (or other bright object) (probably not a good idea!) it shows a blue-white vertical line through the object.


What aspect of photosensor design causes this?
 
In light of disturbing findings, here is:


04012012510.jpg



04012012511.jpg



04012012512.jpg



04012012513.jpg



The Pandora Cooling Dock.
 
Oh wow, the most useless thing I have seen in a long while ;)


But well, if it does what it should...
 
Last edited by a moderator:
Back
Top