Flir E4 Thermal imaging camera teardown

[mago]

Since so many people seem to struggle with generating their own conf.cfc , I took the time and wrote a quick'n'dirty (well, not really quick but still a bit dirty) GUI for the tools in AutoIt. See the attached .zip .
I did some testing and everything should work. Also most errors should be handled. But ("sadly" ) I only have an E4 with firmware 1.19.8 , but maybe someone could do some testing by herself / himself and give me some feedback.

@ flir-ID10T : Maybe this will help you.

Hi stefbeer, link to .zip file is broken! Can you upload again?

EDIT: I tried with guitool, now it works!! 

[Jaimslaw]

Has anyone received recently the free gift for the E4 purchase?
I`ve sent my request early in March, they requested me the invoice around the 15th of March,and since then I have no feedback.

It took them about 3 weeks to ship mine.  You can find out the status by calling FLIR Sales at 800-464-6372.

Thanks for that phone number.  I'm going on five weeks since mailing in all the required info and no free stuff yet.  I'll give them a ring and report back.

[tomas123]

Apologies in advance if this is old news. My searching for a Linux based way to connect to the E4 only turned up some frustrated sounding posts with no success stories.

It seems the E4 does indeed use a modified DHCP/RNDIS that doesn't play well without the FLIR tools installed, and manually configuring your interface to the expected 192.168.0.0 or 192.168.1.0 range fails as expected.

great
this way is simpler as descriped here
https://www.eevblog.com/forum/testgear/flir-e4-thermal-imaging-camera-teardown/msg340074/#msg340074
or
https://www.eevblog.com/forum/testgear/flir-e4-thermal-imaging-camera-teardown/msg339941/#msg339941

I checked it with an old  clean ubuntu machine. An RNDIS Device (like E4) has never plugged in since now.
(no extra rules for ethx edited - see /etc/network/interfaces)

$ uname -a
Linux xxxx 3.0.0-12-server #20-Ubuntu SMP Fri Oct 7 16:36:30 UTC 2011 x86_64 x86_64 x86_64 GNU/Linux

$ cat /etc/network/interfaces
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).

# The loopback network interface
auto lo
iface lo inet loopback

# The primary network interface
auto eth0
iface eth0 inet dhcp

$ dmesg
[922392.410990] usb 2-1.3: new high speed USB device number 3 using ehci_hcd
[922392.548860] usbcore: registered new interface driver cdc_ether
[922392.568661] rndis_host 2-1.3:1.0: RNDIS_MSG_QUERY(0x00010202) failed, -47
[922392.575979] rndis_host 2-1.3:1.0: eth1: register 'rndis_host' at usb-0000:00:1d.0-1.3, RNDIS device, 00:40:xx:xx:xx:xx
[922392.576000] usbcore: registered new interface driver rndis_host
[922392.581075] cfg80211: Calling CRDA to update world regulatory domain
[922392.582848] usbcore: registered new interface driver rndis_wlan
[922392.594211] cfg80211: World regulatory domain updated:
[922392.594213] cfg80211:     (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp)
[922392.594215] cfg80211:     (2402000 KHz - 2472000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[922392.594217] cfg80211:     (2457000 KHz - 2482000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
[922392.594218] cfg80211:     (2474000 KHz - 2494000 KHz @ 20000 KHz), (300 mBi, 2000 mBm)
[922392.594219] cfg80211:     (5170000 KHz - 5250000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)
[922392.594221] cfg80211:     (5735000 KHz - 5835000 KHz @ 40000 KHz), (300 mBi, 2000 mBm)


$ sudo ifconfig eth1 192.168.250.1 netmask 255.255.255.0

$  ftp 192.168.250.2
Connected to 192.168.250.2.
220 Service ready for new user.
Name (192.168.250.2:vbox): flir
331 User name okay, need password.
Password: 3vlig
230 User logged in, proceed.
Remote system type is Windows_CE.
ftp> ls
200 Command okay.
150 File status okay; about to open data connection.
01-01-98  12:00       <DIR>          FlashFS
01-01-98  12:00       <DIR>          FlashIFS
01-01-98  12:00       <DIR>          FlashBFS
04-02-14  16:44                 4096 Printer.swap
04-02-14  08:44       <DIR>          profiles
04-02-14  08:44       <DIR>          Documents and Settings
04-02-14  08:44       <DIR>          My Documents
04-02-14  08:44       <DIR>          Program Files
04-02-14  08:44       <DIR>          Temp
04-02-14  08:44       <DIR>          Windows
226 Closing data connection.

after plug in the E4 I get

Code: [Select]

$ cat /etc/udev/rules.d/70-persistent-net.rules
# This file was automatically generated by the /lib/udev/write_net_rules
# program, run by the persistent-net-generator.rules rules file.
#
# You can modify it, as long as you keep each rule on a single
# line, and change only the value of the NAME= key.

# PCI device 0x8086:0x10ef (e1000e)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="d8:d3:85:7f:88:e6", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth0"

# USB device 0x09cb:0x1002 (usb)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:40:xx:xx:xx:xx", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth1"


[ixfd64]

Interesting read about thermal cameras in general: http://www.prnewswire.com/news-releases/the-world-market-for-commercial-and-dual-use-infrared-imaging-and-infrared-thermometry-equipment-vol-irw-c-2014-edition-253528351.html

[pingo]

Has anyone explored whether it would be possible to increase the range of measured temperatures? Would it be possible to have the TIC measure temperatures higher than 280°C?

[Bolt]

Has anyone explored whether it would be possible to increase the range of measured temperatures? Would it be possible to have the TIC measure temperatures higher than 280°C?

Possible, but IMHO temperature limit is about 320°C, in RAW data its equal 65535. Low limit theoretically is 0°K, I have seen -150°C.

[tomas123]

Possible, but IMHO temperature limit is about 320°C, in RAW data its equal 65535. Low limit theoretically is 0°K, I have seen -150°C.

It's not so simple.
You can play around with this excel sheet to calculate Temperature=function(RAW)
 
source:
http://u88.n24.queensu.ca/exiftool/forum/index.php/topic,4898.msg27546.html#msg27546


Hint: the embedded RAW value is not the direct result of the sensor ADC measurement of a single pixel

[mikeselectricstuff]

Has anyone explored whether it would be possible to increase the range of measured temperatures? Would it be possible to have the TIC measure temperatures higher than 280°C?

I wonder if maybe there might be some interesting rset parameters, e.g. to tweak ADC gain/range

[max-bit]

Hi
Is the last HACK 100% working with FW ver 1.22.0 and HW 1.1L?
There is information that is not of the bid (or users are making a mistake)
Thanks all info

[humdinger13]

To install Taucher's menu hack:

1. Install the RNDIS drivers on your PC. (http://cdn.cloud.flir.se/swdownload/assets/other/flir_device_drivers.exe)
2. Turn on RNDIS on the camera via the hidden menu (Settings>Device Settings>Camera Information - Press <right> for 10 seconds, select "USB Mode" then select "RNDIS")
3. Connect camera to PC via USB.
4. FTP to camera (eg. with FileZilla): Normal logon type, Host "192.168.0.2", user name "flir", password "3vlig"
5. Make a back-up of the files on your camera (copy files from camera to your computer into a back-up directory of your choice)
6. Copy the ADDMENU-Beta files over the files in the camera (eg. unzip the files to your computer then copy the folder "FlashBFS" from your computer over the folder "FlashBFS" in the camera - overwrite files)
7. Power-down the camera, remove battery and USB cable.
8. Re-insert battery and turn on.
9. Enjoy! 

Re: menu hack: 
Issue 1:  I made it to step 5 (above).  I selected ALL the folders/files on the camera (to backup everything) and during the transfer there are about 139 failed transfers.  I tried this on 2 different computers with similar result. 

Issue 2:  The FlashBFS\system folder in the 'ADDMENU-BETA3' mod has 1 folder and 5 files, however the same folder in the camera has 9 folders and a lot of files.  I'm guessing this means overwrite the 1 folder and 5 files, but KEEP all other files in the cameras "FlashBFS" folder as is.

I installed flir tools 4.0 from the two zip files posted somewhere, and the camera (apart from a few quirks like an occasional failure to find my images until I reboot and it reindexes them) seems to work OK. 

I am trying to upgrade but run into problems installing the device drivers.  It fails with FLIR Device Drivers -- Error 1935 An error occurred during the installation of assembly 'policy.8.0.Microsoft.VC80.OpenMP,version="8.0.50727.4053",type="win32-policy",publicKeyToken="1fc8b3b9a1e18e3b",processorArchitecture="x86"'. Please refer to Help and Support for more information. HRESULT: 0x80070005. assembly interface: IAssemblyCache, function: CreateAssemblyCacheItem, component: {04B9F3B6-9645-7658-A01F-C8B3B9A1E18E}.

The thing is is that I was expecting to be installing a 2.1 x 64 version and not 1.9, but have been unable to locate one, if that matters, since my fw version is 2.1.  Can anyone help me on this please?  I've spent a lot of time looking around to see if I could find a different driver file.  Also, the installer says something about not recognizing the windows version, which I find strange.  If I manage this part the rest should be pretty straightforward.  Thanks.  Please be kind, I'm getting old and my mind isn't quite as sharp as it used to be.

Edit:  I should add that I am using Windows 7 x 64

[Nemonic]

Not strictly a question for this topic, might want to contact FLIR support, but a quick google gives some things to try.

http://blogs.msdn.com/b/astebner/archive/2005/08/16/452523.aspx

[ixfd64]

Has anyone explored whether it would be possible to increase the range of measured temperatures? Would it be possible to have the TIC measure temperatures higher than 280°C?

I've wondered the same thing but didn't get any definite answers. Some cameras from FLIR and other companies (such as Fluke and General Tools) are able to handle temperatures as high as 1,000°C or more. Do those cameras use different hardware, or are temperature ranges also firmware-limited?

[emptech]

I don't have an answer, but if thermal cameras are like video cameras, at least like mine (canon gl-2), once the sensor
becomes saturated, it can no longer capture information.  In visible light photography (digital) you are better off having a
photo that is under exposed than over exposed, again, saturation.  My camera (gl-2) has a mechanical neutral density
filter that goes into place once a picture tends to be over exposed.

It would seem reasonable that the micro-bolometer has limits, I would assume the cells, pixels, could become saturated,
like pointing at the sun.  Of course in that case, you could have permanent damage.  We know the IR cameras have a
mechanical shutter, but that seems to be for some kind of internal calibration.

Would some kind of neutral density filter be used to extend the upper limit of a thermal camera?

Jim

[tomas123]

I wonder if maybe there might be some interesting rset parameters, e.g. to tweak ADC gain/range

The Flir Exx with two temperature ranges don't use a filter like the neutral density filter in photography.
I think, that the TIC only changes the integration time.

see here for samples with variation of the integration time:
http://www.flir.com/cs/emea/en/view/?id=41923

[humdinger13]

Hi
Is the last HACK 100% working with FW ver 1.22.0 and HW 1.1L?
There is information that is not of the bid (or users are making a mistake)
Thanks all info

The last hack (1.21/1.22 firmware hack info or page 1 of this thread) is working very well for current fw 1.22.0 and hw 1.1L.
Follow the instructions. Do not replace the original file "conf.cfc" (on camera)! Better way: Rename the new, protected config file "conf_new.cfc" to "e8.cfc" and copy it to "FlashFS\system\appcore.d\config.d\" on your camera, cold-start and you are done.

Thanks  for the help, but I can't seem to install the drivers at all.  Based on what you're saying, the drivers I have should work.  I don't know what to do at this point. 

[humdinger13]

Not strictly a question for this topic, might want to contact FLIR support, but a quick google gives some things to try.

http://blogs.msdn.com/b/astebner/archive/2005/08/16/452523.aspx

Not very keen on contacting flir about this.  haha.  I did have a look at that link, tyvm, and tried some suggestions like disabling certain s/w, but still no go.  I'm tearing what little is left of my hair out!

[humdinger13]

Hi
Is the last HACK 100% working with FW ver 1.22.0 and HW 1.1L?
There is information that is not of the bid (or users are making a mistake)
Thanks all info

The last hack (1.21/1.22 firmware hack info or page 1 of this thread) is working very well for current fw 1.22.0 and hw 1.1L.
Follow the instructions. Do not replace the original file "conf.cfc" (on camera)! Better way: Rename the new, protected config file "conf_new.cfc" to "e8.cfc" and copy it to "FlashFS\system\appcore.d\config.d\" on your camera, cold-start and you are done.

Thanks  for the help, but I can't seem to install the drivers at all.  Based on what you're saying, the drivers I have should work.  I don't know what to do at this point.

Try to let you help from google.
I think you need the Microsoft Visual C++ 2005 Redistributable Package (http://www.microsoft.com/en-us/download/details.aspx?id=5638)
But I can also be wrong.

The first time I installed flir tools, it also installed a few .NET programs and I think that 2005 C++ one as well as 1 or two others.  .  Subsequent installations were never the same, and the last one didn't install it.  The last couple of times I happened to check the options box for network install which I had missed before.  I will see about manually putting in that distributable, but what concerns me more is that error with VSS.  I am not sure which voume is causing it, and  it still failed after disabling a bunch of software and turning VS off, as well as turning security essentials off.  If anyone has any other ideas I am all ears.

Thanks again.

[inline]

I ordered and received my 20mm diameter and 100mm FL ZnSe lens off eBay, shipped from China the other day and just sent my lens holder off for production to a local outfit (I found them through www.makexyz.com if this helps anyone, because I could never get ahold of the LuneValley3D place, I tried like 7 times...)  The lens holder was only about $10 with shipping, so much better than thru Shapeway which wanted like $30.  The guy who printed it even sent me a picture and it ships tomorrow, so hopefully I'll have it on Saturday.  Finally I'll get up close ability for PCB work!

[Nemonic]

Here's mine + scope attachment. I added finger holes to help remove it.

The scope attachment works, however the lenses either need to be a larger diameter or a smaller focal length, also images are inverted.... and the whole thing looks ridiculous.

[Teledog]

Hi,
Newbie here.
Not sure if this is the correct thread to post on..my apologies if not.
Was about to order an E4, but with their latest firmware, is it still possible to re-flash it? (Now April 3, 2014)
I am familiar with Linux/firmware flashing (unlocked my HTC, made/used a GoodFET, etc.)
TIA! 
Teledog

[john19]

I wonder if maybe there might be some interesting rset parameters, e.g. to tweak ADC gain/range

The Flir Exx with two temperature ranges don't use a filter like the neutral density filter in photography.
I think, that the TIC only changes the integration time.

see here for samples with variation of the integration time:
http://www.flir.com/cs/emea/en/view/?id=41923

How does that camera control the integration time? 0.01 ms seems too fast for a mechanical shutter.

[Fraser]

@nemonic,

Interesting work you are doung on auxilliary lenses.

Some comments from me .....

With regard to the close-up lens attachment please be aware that tle lenses from China come in three forms depending upon focal length. Bi-Convex, Convex Meniscus and Plano Convex.  My Inframetrics Industrial Close-Up lens is a Convex Meniscus type that has the concave side facing the target. This makes sense and I used the smaller Chinese meniscus lenses in the same manner. Other members of the forum discovered that the lenses actually worked better with the convex side facing the target. In the pictures it looks like you have a concave side of the lens facing the target ? You may wish to experiment with the convex side facing the target as you may see a significant image improvement (especially edge focus) I also note that you appear to have the close-up lens positioned further forwards in your mount than that of Georges80 design....any reason fo this ? Do you get vignetting ?

With regard to your telescope....... I am in exactly the same position as you. I have two lovely Inframatrics X3 Telescopes, but they invert the image. I have used a simulator to investigate why they invert and it is definitely a product of the design and not an issue with the positioning of my E4 in relation to the lens rear. My X3 telescopes use 4 lens elements. I X-Rayed the telescopes to determine the lens shapes. I found two Convex Meniscus lenses in the front (moveable) section and two Concave Meniscus lenses in the rear section.

The image inversion occurs between the 1st and 2nd lenses at the front of the telescope and the 3rd and 4th lenses just shape the rays to suit the very large aperture lens output.

As you may know, refractor telescopes come in two forms, inverting and non inverting. For a non inverting telescope you will need a primary lens that has a long focal length to prevent the image inversion that would otherwise occur in the ray path between the two lenses. I failed to find such a ZnSe lens as most on e*ay are designed for less than 100mm FL to suit CO2 lasers.

To correct the inversion, you could add another biconvex lens between the 2nd lens and the camera ? That is something that I will be pursuing with my Inframetrics Telescopes in the hope of making them more user friendly.

If anyone is interested, I will post the X-Rays of my Industrial X3 Telephoto lenses and their dimensions etc.

I have managed to buy a FLIR X2 compound telescope for my FLIR PM695 cameras and that does not invert the image. It should work well with the E4 provided the E4's wide FOV does not cause vignetting. The FLIR lens is very short in length and uses complex optics that would not be available to home users, so sadly I doubt its internal design will offer much to your lens experiments. Sadly the FLIR X2 telescope is also a very expensive product, costing many thousands of Dollars new.

[Nemonic]

@Aurora

The lenses I have are a 25.4mm and 101.6mm FL Plano-Convex, however positioning the elements more around 130mm centre to centre yields a sharper image, so i am not convinced the reported FLs are accurate.

I currently have both Plano sides facing each other within the lens tube as i believe this is the normal arrangement with optical telescopes. Apparently this is to reduce aberrations but i am not certain it will matter for my case. (Edit: got this the wrong way around i think)

The reason the lens is closer to the E4 body is that its a 25.4mm FL and was under the impression the same FL rules apply to the ocular lens such that the ocular lens is positioned FL + (E4 lens 6.6mm) distance. The interesting thing from having the 25mm lens at this distance is the magnification at close range is incredible, i've managed to capture the individual pours on my fingers "sweat" cool and dissipate heat.

With regards to inversion, i know of two simple ways to correct the image. One is the addition of a 3rd lens called an erector lens like you mention and the other is with replacing the ocular lens with a convex. The benefit of a convex lens would shorten the design and similar element arrangements are used in theatre binoculars.

I was planning on using a 3rd lens as i wasn't sure i would be able to get a convex with suitable FL properties. The problem i have is that the ocular lens has to be positioned at a distance that doesn't cover the FOV of the E4, so you get the strong vignette effect in the image i posted. The solution would be a (i haven't calculated this) 38mm dia+ ocular lens and an even larger objective, but then it starts getting pricey as you're after non-common lens sizes and the FL with an erector makes the whole thing crazy long.

Ideally i would have a 40mm FL 38mm dia+ objective with a 10mm ish FL ocular lens + 10mm FL erector (i think), but i cant see me getting these lenses from anywhere but a lens lab for crazy money. The benefit of the larger objective would also be it would collect more thermal light resulting in better detection of low temperature at range.

These are all optical solutions to the inversion problem and frankly just flipping a config flag on e4 to invert seems the much more preferred option. There is no reason why the E4 cant invert and im sure there will be a setting for it.

I shall have a look at the Convex Meniscus type you mention. I am not sure what optical properties they offer over Bi-Convex/Plano-Convex types.


- Disclaimer, i am not by any means an optometrist or have any expertise in optics. I happen to have a complete opticians lens set which i use to experiment different lens combinations with tho.

[Fraser]

@Nemonic,

Ditto, I am very not knowledgeable on the intracasies of optics, and all that I have learnt was from Googling and practical experimentation    I learnt another lesson whilst using a lens simulator.... Index of Refraction is very different to glass when simulating a Germanium lens ! Glass is approx 1.55 and Germanium is 4.

I have some Germanium 1/2" diameter 15mm FL Thorlabs Plano Convex lenses that were very kindly gifted to me by a member of this forum for experimentation. He is also selling them on e*ay. I was thinking of some experiments with them to see if they could help with my X3 Telescope inversion issue. They may be too small for that task though.

The greatest challenge for me has been in obtaining reasonably priced lenses. As you state, the larger they are, and more uncommon FL's, quickly make them unaffordable for experimentation.

I will do some experiments with my ZnSe close-up lenses as well. I may have them too close to the E4's lens. I discovered the use of No2 close-up optical lenses several years ago when working on video document presenters. The standard Sony Zoom camera module could not focus at the short distances from the lens to the platen so the manufacturer placed a No.2 lens directly in front of it to act as a 'monocle' and provide closer focussing. Once I discovered the availability of ZnSe lenses, with useful FL's, it was an easy step to inproving the close range focussing of my Ex Fire Brigade ARGUS thermal cameras. The same technique works with the E4 as we know. I have also experimented with GaAs lenses that are AR coated and designed for CO2 lasers. They also seem to work well but have lower transmission figures. They have the advantage of not scratching as easily as the soft ZnSe material. I have yet to experiment with Si lenses but from memory they are more suited to Shortwave thermal imaging applications.

[Corn]

Just check if you copied the conf.cfc file as conf.cfg to the camera. Don't ask me what makes me think so. 

Just tried (over and over) the above with my just arrived E4 1.22 with no success 
i can only see some extra menu after installing the ADDMENU-BETA3 package, although not fully functional, some of them just freezes camera, like multi temp measure, but i think it's due to the main hack not working.

BTW apart a supposely better looking picture on screen, where am i supposed to see the full resolution? Camera info maybe? or some new menu which lets me to choose resolution? Sorry i'm completely new to this cameras..