Flir E4 Thermal imaging camera teardown

[Maxpax]

 

That's on me

[tomas123]

Ok, I managed to extract all info from putty by copy to clipboard in sections and puzzle it together. Phew!

bad news, your sensor is only 80x80

Code: [Select]

i5

.image.flow.framebuffer.geom:
r---r--------- 0 root   root   <i> bitsPerPixel                  16
r---r--------- 0 root   root   <i> height                        80
r---r--------- 0 root   root   <i> width                         80


.calib.detector.geom:
rw--rw-------p*0 root   root   <i> imgPixX                       80
rw--rw-------p*0 root   root   <i> imgPixY                       80
compare the values with an unpatched Flir E40
image 160x120 <-> sensor 320x240

Code: [Select]

E40

.image.flow.framebuffer.geom: (3)
r---r--------- 0 root   root   <i> bitsPerPixel                  16
r---r--------- 0 root   root   <i> height                       120
r---r--------- 0 root   root   <i> width                        160


.calib.detector.geom: (8)
rw--rw-------p*0 root   root   <i> imgPixX                      320
rw--rw-------p*0 root   root   <i> imgPixY                      240

we need 17 posts to check it 

[Maxpax]

Okay. If you are 100% sure, then - I'll live with that.


Thank you so much for your help and time spent on this.

[Maxpax]

But hey, what's up with this section then:

Code: [Select]

.caps.config.image.settings:
r---r--------- 0 root   root   <b> enabled                     true
r---r--------- 0 root   root   <i> resIR                        120

[encryptededdy]

Ok, I managed to extract all info from putty by copy to clipboard in sections and puzzle it together. Phew!

bad news, your sensor is only 80x80

Code: [Select]

i5

.image.flow.framebuffer.geom:
r---r--------- 0 root   root   <i> bitsPerPixel                  16
r---r--------- 0 root   root   <i> height                        80
r---r--------- 0 root   root   <i> width                         80


.calib.detector.geom:
rw--rw-------p*0 root   root   <i> imgPixX                       80
rw--rw-------p*0 root   root   <i> imgPixY                       80
compare the values with an unpatched Flir E40
image 160x120 <-> sensor 320x240

Code: [Select]

E40

.image.flow.framebuffer.geom: (3)
r---r--------- 0 root   root   <i> bitsPerPixel                  16
r---r--------- 0 root   root   <i> height                       120
r---r--------- 0 root   root   <i> width                        160


.calib.detector.geom: (8)
rw--rw-------p*0 root   root   <i> imgPixX                      320
rw--rw-------p*0 root   root   <i> imgPixY                      240

we need 17 posts to check it 

That's so odd...

I'm pretty sure they don't make a 80x80 sensor. The lowest res non-lepton sensor is a 80x64 50 micron sensor that you find in the FLIR A5.

Unless they're using the ULIS Pico80P, but that's very unlikely.

[tomas123]

ok, you don't give up.

The next proof is your calibration file. The calibration with bad pixel mapping is taken by full resolution.

your calibration file name is ds250C_we_ap_fi_le.gan

Code: [Select]

.registry.calib.detector.deadPixMap.ds250C_we_ap_fi_le:
rw-dr--------- 0 root   root   <a> fileName        "ds250C_we_ap_fi_le.gan"
rw-dr--------- 0 root   root   <i> type                         100

you find it here

Code: [Select]

+---FlashFS
|   +---system
|   |   +---maps
|   |   |       ds250C_we_ap_fi_le.gan
you can read the gain map with exiftool

here a sample from a E40

Code: [Select]

$ exiftool ds250C_we_ap_fi_le_static.gan
File Type                       : FLIR
File Type Extension             : fff
Gain Dead Map Image Width       : 320
Gain Dead Map Image Height      : 240
Gain Dead Map Image Type        : TIFF
Gain Dead Map Image             : (Binary data 153804 bytes, use -b option to extract)

 

[Maxpax]

Ok.

exiftool output for ds250C_we_ap_fi_le.gan:

Code: [Select]

ExifTool Version Number         : 10.02
File Name                       : ds250C_we_ap_fi_le.gan
Directory                       : .
File Size                       : 13 kB
File Modification Date/Time     : 2000:01:01 00:34:00+01:00
File Access Date/Time           : 2015:10:14 12:44:10+02:00
File Creation Date/Time         : 2015:10:14 12:44:10+02:00
File Permissions                : rw-rw-rw-
File Type                       : FLIR
File Type Extension             : fff
MIME Type                       : application/unknown
Creator Software                :
but there's another file in the maps directory called "ds250C_we.crs" mentioned, among other places, here in the allsettings-file:

Code: [Select]

.registry.calib.detector.data:
rw-dr--------- 0 root   root   <x> ds250C_we
with the exiftool output:

Code: [Select]

ExifTool Version Number         : 10.02
File Name                       : ds250C_we.crs
Directory                       : .
File Size                       : 13 kB
File Modification Date/Time     : 2000:01:01 00:29:00+01:00
File Access Date/Time           : 2015:10:14 12:44:09+02:00
File Creation Date/Time         : 2015:10:14 12:44:09+02:00
File Permissions                : rw-rw-rw-
File Type                       : FLIR
File Type Extension             : fff
MIME Type                       : application/unknown
Creator Software                :
Coarse Map Image Width          : 80
Coarse Map Image Height         : 80
Coarse Map Image Type           : TIFF
Coarse Map Image                : (Binary data 13004 bytes, use -b option to extract)
That kind of convice me that you are right, tomas123. But I just can't help to wonder why the resolution 120x120 is suggested here and there in filenames and the allsettings-file. Could it be that the sensor has different "modes"? ...But I'm not the one to speculate on this, all you other people here have way more knowledge about how it's normally done...

[tomas123]

 the fw is the same for all ix cameras (all resolutions)

[stefbeer]

@ Maxpax:
Back then I successfully upgraded my i3 (bought about 3 years ago) with the instructions from this post. When I look at your original config file, then all the lines essential for the hack are missing. Which means that you have a later hardware revision and / or firmware version. And please note the last line of the linked post. My TIC was able to do the full 144 but it seems yours really only can do 80.

And the 120 you mentioned here are from the config file you copied from the pdf with the instructions. Because the "new" config file is successfully loaded (which means you did all the # ID and # CRC01 stuff correctly ) this parameter is set but sadly because of the hardware / firmware this parameter doesn't do anything. And even if it did, you really seem to have only a 80 sensor.

And the timestamp from January 1st 2000 is just a default time if the actual time isn't set, this doesn't mean the TIC is 15 years old.

I have no idea why the i-Series has three different types of sensors but it seems that you just have the wrong one.

[joe-c]

Hello,
try to go to: FlashFS\system\maps\  there should be a *.gan file
ds150C_we_ap_fi_le_static.gan (from FlirC2)
ds250C_we_ap_fi_le.gan (from Flir i3... later enhanced to 120x120)
ds250C_we_ap_fi_le_static.gan (from Flir E4... later enhanced to 320x240)

I guess... this Calibration file was ever in the full resolution of the Camera.
Try to decode it with my Dev Tools file to image decoder:
http://joe-c.de/pages/posts/programm_dev_tools_133.php#bilder

from the i3 and E4 i get a 320x240 resolution, the C2 brings only 80x60... it has a Lepton core.

[tomas123]

@ Maxpax:
You wrote that exiftool can't read your
ds250C_we_ap_fi_le.gan
Please zip and upload your gan file. I will check it.

[Maxpax]

vielen dank!

[tomas123]

just for fun some exiftoool and imagemagick magics.

your ds250C_we_ap_fi_le.gan has the Flir Tag FFF_TAGID_DeadMap = 4 (see pdf)
with the config.txt file from attachment we can extract the RAW 80x80 image

Code: [Select]

>exiftool -config config.txt ds250C_we_ap_fi_le.gan
ExifTool Version Number         : 9.91
File Name                       : ds250C_we_ap_fi_le.gan
File Size                       : 13 kB
...
File Type                       : FLIR
MIME Type                       : application/unknown
Creator Software                :
Embedded Image Width            : 80
Embedded Image Height           : 80
Dead Map                        : (Binary data 12800 bytes, use -b option to extract)

>exiftool -config config.txt -b -DeadMap ds250C_we_ap_fi_le.gan > 1.hex

now convert with IM

Code: [Select]

>convert -depth 16 -endian lsb -size 80x80 gray:1.hex  t2.png
resize for better bad pixel peeping

Code: [Select]

>convert t2.png -filter point -resize 320x resize.png

[encryptededdy]

Still no idea where FLIR is getting these 80x80 sensors.

Perhaps in the later versions some image processing (specifically downscaling) got moved on chip?

But that would make no sense because the E4 doesn't do that...

hmm...

[Maxpax]

My i5 is blue and gray, not black. I think it is one of the earlier versions of i5, and therefore one of the older designs. I suppose it is quite natural to change designs i the beginning of new models as a company gains in understanding from testing, customer feedback, ideas on how to achieve cheaper/simpler manufacturing and not least, providers who approach the company with deals on technology parts, such as sensors.

Anybody want to buy it? 

[tomas123]

sorry for double posting

I published in the seek thread an excel sheet to show, why the RAW values of an E4 microbolometer (response of "A/D Counts") are linear to radiance [W / m^2] (per nm wave length) and not linear to temperature [Kelvin]. The difference is not great.

https://www.eevblog.com/forum/testgear/yet-another-cheap-thermal-imager-incoming/msg779709/#msg779709

[brobbuilder]

Would be cool to get a firmware on the i-series that dumps full rez (320x230) to SD, since the screen can't handle more than 140x140.

We never did figure out the pinouts on those, did we?  If we did, the hazard cams from BMWs/Audis might finally be reusable, since they use the same sensor.

[jumbo]

Hello,

I changed E30bx to E60bx and its really clear view now, thanks eevblog 
but is it possible to change settings in video mode to turn of staff on right?


If you don't understand me I attached picture below.
On first its all staff (temperatures, point etc.) on second its what I would like to have on video mode.

I would like to clear screen only on video mode - not on picture mode.

[Chanc3]

Hello,

I changed E30bx to E60bx and its really clear view now, thanks eevblog 
but is it possible to change settings in video mode to turn of staff on right?


If you don't understand me I attached picture below.
On first its all staff (temperatures, point etc.) on second its what I would like to have on video mode.

I would like to clear screen only on video mode - not on picture mode.

You can turn it all off in one of the settings panels - I'll have to dig out my E60 and see if I can find it.

[jumbo]

Hello,

I changed E30bx to E60bx and its really clear view now, thanks eevblog 
but is it possible to change settings in video mode to turn of staff on right?


If you don't understand me I attached picture below.
On first its all staff (temperatures, point etc.) on second its what I would like to have on video mode.

I would like to clear screen only on video mode - not on picture mode.

You can turn it all off in one of the settings panels - I'll have to dig out my E60 and see if I can find it.

From menu I can only switch off left site

[tomas123]

With telnet you can use commands like

Code: [Select]

rset ui.userSettings.overlayLogo false
It's a sample from my Flir Exx old version without MSX
For more informations google this string on eevblog

[Chanc3]

Or you can delete the logo.bmp file! (Make sure you have a backup).

Good news for those with an i50 - I have successfully upgraded the GUI to include better measurement functionality and I have also updated the IR resolution to 240x240! (up from 140x140). Funnily enough, the i60 only goes up to 180x180 I think.

Does anyone have an InfraCAM SD? I picked one up dead cheap the other day but it doesn't get past the boot screen (doesn't even show it loading up). I've tried connecting to it and pulling off the files so I can inspect them, but FTP only seems to work very briefly - any ideas?

[simclardy1]

  celebrate! not scared.

I managed to get the resolution resolved. I managed to get the menus updated. Problem. the picture in picture and thermal fusion are on the menu but I can't operate them? the other thing is Bluetooth and wifi don't seem to do anything? I don't care too much about this but the fusion or pip would be a plus. The succeeded to activate: camera record, resolution, dots, delta, isotherm, zoom.....wow it's really cool.
Thanks to mike, and everyone!

[simclardy1]

I got it! I had to edit the appcor config file a bit more. I made a mistake on the framegrab entries.
cheers! if I can help a newbie like me let me know. 

[lefedor]

Hi, I'm new in this topic, need scope owner help with my thermal sensor (have sensor (flir photon camera) with dead electronics):



Is here exists any fella with oscilloscope (with logic analyzer may be, but i guess it's not critical) and also brave, enthusiastic and open to make few start up/initialization measurements on his sensor pins?
Looking for both owners of 'I' series and 'E' series.