Flir E4 Thermal imaging camera teardown

[Fraser]

Close up lenses are little different to a pair of reading glasses in terms of how they work. This technology has been in use for many years on video presenters to enable the standard integrated lens video camera to close focus. A No2 auxilliary close up lens is placed in front of the cameras lens.  That is where I  got the idea of using ZnSe lenses with thermal cameras. A few experiments with cheaply available CO2 laser focus lenses proved that they would work OK.

Please bear in mind that the ZnSe CU lenses are auxiliary and not primary lens structures. They present only loss in the optical path and will always have a transmission of less than 100%.

The ZnSe lenses have a very good transmission figure of approximately 97% between 8um and 12um

Take a look at the graphs on this Thorlabs page

http://www.thorlabs.de/newgrouppage9.cfm?objectgroup_id=1781

If a calibrated temperature reading is needed, an offset can be applied using the rough transmission figure or by testing the camera against a Black Body source whilst fitted with the ZnSe lens. Some cameras permit a lens transmission figure to be entered and these automatically compensate in the measurements.

Close-Up lenses are just that, close-up only. The depth of field is very small as well.

To produce an auxilliary telescope for the E4 you would need at least two lens elements as a minimum.
This has been covered previously by myself and others.

Take a look through my 'FLIR E4 Useful information' thread

https://www.eevblog.com/forum/reviews/flir-e4-the-useful-information-thread/

https://www.eevblog.com/forum/reviews/flir-e4-the-useful-information-thread/msg339922/#msg339922

I bought all of my thermal telescopes as used parts as new ones cost a fortune and they are difficult to make at reasonable cost due to lens element prices.

Aurora

[mikeselectricstuff]

Just did a simple test for framerate.
Set up a motor with a foil segment on the shaft, set speed to around 9 revs/sec.
If it was reading at 60fps and avaraging, it should show a blurred image
If it was 9fps and duplicating it should show a sharp image of the segment once the speed was adjusted to be in sync.
Is showed a sharp image, so looks like the effective data rate from the sensor really is 9fps

...of course there could be some magic command you can send to it....
 

[SeanB]

Might be you have to set the bit for PAL/NTSC to all zero, or there is a fuse bit. Wonder how much fuzzing you can do to the input config data stream before you brick it.

[Solare]

"Magic command" that unlocks everything! I like the sounding of this...  O0

[Fraser]

Sadly it is not a 'magic command' as such as FLIR never intended end users to change the settings.

You are dealing with changing a bit, or group of bits, by direct access methods. No friendly 'commands' here as this is not an OS based change. You are changing the ROIC settings.

It is interesting to find out that no averaging is in play and that the E4 uses just a single frame to produce its images from what is a very small and relatively inefficient lens. The Ex series microbolometer is a sensitive little beast 

Aurora

[Fraser]

@Solare,

Just so you are aware, the matter of the E4 frame rate and associated legal issues has been discussed in this thread previously. It was agreed by those involved in 'upgrading' the E4 that it was not a good idea to pursue upgrading the frame rate. Such action would bring significant pressure to bear on FLIR to not only countermeasure such a frame rate 'hack',  but to also properly lock down the camera against other 'upgrades'. This would not benefit the greater community.

Aurora

[Solare]

@Aurora and 'those who may concern'!

I am happy to comply (c)

Really, who am I to break the Rules accepted by Community. But, in reality, I never said I was going to!

Moreover, in previous posts, I stated clearly that I am not interested in getting 60 fps in realtime, but of course, would welcome the possibility to record some 2-3 seconds of video in full 60 fps temporal resolution. Since using 2-3 seconds video to guide in realtime a manned missile in total darkness from 30 meters after downloading it to PC thru USB, sounds NONSENS even for ITAR Guys ))) I consider matter closed.

I never insisted, or implied, that 60 fps unlock should be done. But, I wondered... what IF it could be recorded?

Since you've explained in detail that this is not technically possible, I say okay not possible means not possible, what can I do.

But what Rule on Earth can forbid me dreaming ? 

[Fraser]

@Solare,

Please do not take my comment as 'warning you off' , that was not my intention. It was just to let you know why the frame rate has not been greatly investigated in this thread and why you may not get answers to certain questions that you may innocently ask 

By the way... the "ITAR Guys" see things very much in black & white so even a short video grab at 60fps upsets them 

Best Wishes

Aurora

[Solare]

@Aurora

Okay

@All

Still, this is weird... replicating frames mean all the computing power is engaged in data stream processing matching 60 fps... just think of for how many hours the E4 could operate, if it really skipped the frames...

[mikeselectricstuff]

@Aurora

Okay

@All

Still, this is weird... replicating frames mean all the computing power is engaged in data stream processing matching 60 fps... just think of for how many hours the E4 could operate, if it really skipped the frames...

Remember the sensor was developed for automotive applications, where it needs a high framerate.
For other, lower volume uses, they obviously thought it would be simpler to keep the interface the same ( and the same as earlier sensors).  Even if the sensor outputted a lower data rate they'd need an FPGA to handle it, so it doesn't really make a lot of difference..

[max-bit]

I have a problem
E4 updated firmware 1.19 to version 2.3.0
I acted in accordance with hack http://fubar.gr/hacking-the-flir-e4/
Unfortunately it does not work HACK MENU
First, the file extension "conf" is a .cfg ( my camera) and according to the description on this page conf.cfc? (new camera)
Secondly, when trying to use a Python script, there is a mistake:

applying 5178 bytes of delta at 000004d8
traceback (most recent call last:)
file "menu_conf_e4.py" line 12, in <module>
assert len(d) ==  len(delta)
assertionerror …

E8 menu exists but it is incomplete.
Does not work "MENU free enhancend menu"
etc ...

Does anyone know the solution?

PS
Camera had a previous hack to version 1.19

[max-bit]

Hack resolution from what I've seen has been maintained from the previous hack.
I also wrote to the people of the website "fubar" because that's where the problem was also prompted tem.
Does anyone know what is the difference between conf.cfg files and conf.cfc (conf.cfc are in newer releases of cameras)

[Fraser]

If anyone is interested in the hardware aspects of thermal cameras, I have been posting a teardown and analysis of a FLIR E2 here:

https://www.eevblog.com/forum/testgear/flir-thermal-camera-for-teardown-my-latest-patient-the-e2/

Its a work in progress and really just for fun.

Aurora

[Solare]

Hi all,

anybody knows what is current status and progress of development menu's, sofware and reverse engineering the E4 device?

Who is doing what? Open problems? New solutions?

thanks

[max-bit]

I think that for the time being, it would work out, the problem of E4 and cameras firmware update to version 2.3.0 and HACK menu (and HACK resolution)

[staze]

so now I'm curious... what do you get by updating to 2.3.0?

[Solare]

Sorry but isn't Firmware 2.3.0 already hacked as well as the MENU?

[max-bit]

from what I see, in the software 2.3.0 is a little more powerful.
Although part of me just a function does not work, wrong hack through the menu (File issue conf.cfg / conf.cfc)
As a topic is being developed and is updated then  need to solve this problem (and also for those who carry out such update)

[Fraser]

Be very clear on what firmware 2.3 is and what it represents......

Firmware 2.3 came about as a COUNTERMEASURE to users upgrading their cameras.

Firmware 1.19.8 was the last non-countermeasured firmware, As such it is inadvisable to move to a later firmware. That said, Firmware 2.3 has had its defences circumvented but I have seen no real benefit to installing this later firmware. You would effectively be voluntarily installing a firmware that was designed to be non upgrade friendly.

In the end the choice is yours though

Aurora

[staze]

Be very clear on what firmware 2.3 is and what it represents......

Firmware 2.3 came about as a COUNTERMEASURE to users upgrading their cameras.

Firmware 1.19.8 was the last non-countermeasured firmware, As such it is inadvisable to move to a later firmware. That said, Firmware 2.3 has had its defences circumvented but I have seen no real benefit to installing this later firmware. You are would effectively be voluntarily installing a firmware that was designed to be non upgrade friendly.

In the end the choice is yours though

Aurora

That's kind of what I figured... I'm happy with 1.19.8.

As a note, anyone else see the wifi symbol show up in the menus after the menu hack?

[max-bit]

OK update to version 2.3.0 is the user's wishes
But there are people who have done upgrades and they have a problem with full support (HACK menu)
And for this reason you need to solve this problem
Nor has on many pages, clear information that the upgrade to version 2.3.0 is a problem with the menu HACK

And there is no possibility of withdrawing Software 1.19.8 ?!

[max-bit]

So maybe let's discuss about the issue or the configuration file conf.cfg (older versions) and conf.cfc (newer versions).
What these files are different and why the PYTHON does not modify them?

[Bud]

You would effectively be voluntarily installing a firmware that was designed to be non upgrade friendly.

Do you expect any more updates for 1.19 ?

[max-bit]

Is a software update to 2.3.0
However, it is a problem.
However, Hack, higher resolution is maintained.
But there are problems with getting the full menu (menu hack).
The case is based, on various files conf.cfg / cfc.
You have to recognize what these files differently, and if there are still any problems?
 At least in this case there is no answer!

Can anyone share a file structure: conf.cfc? (I do not have such file) It is known that at the same time config.cfg file (a file with the old version of the camera) does not work PHYTON script.

[max-bit]

Can anyone share a file conf.cfc?
Is anyone working on the theme file conf.cfc / conf.cfg?