FLIR - Interesting dual FOV thermal camera on eBay. Save me from myself !

[Ultrapurple]

Re-reading the brochure that Fraser so kindly located and uploaded, it seems that maybe all I need to to is frighten the thing with 12V and I'll get analogue video out. How about that?

Looking at the main electronics assemblies there appear to be four 16C550 UARTs and four LT1381 RS-232 driver/receivers, so identifying the serial connection(s) shouldn't be too hard. Likewise, the larger PCB that's visible is clearly a power supply board, so it should be easy enough to work out where to apply volts. As for the video output - well, analogue video probably won't be too much of a challenge to spot.

Rather than dive into all this though, my first move has to be to ask if anyone has got any technical data, pinouts and/or command set info for these beasties?

[Fraser]

Ultrapurple,

That is not what I was expecting to see inside this camera ! This looks more like the PCB’s in the AGEMA/FLIR PM series cameras. Even the PCB naming convention looks the same (RIPO2, RISL 2). There is nothing there that looks like it came from an Indigo Photon core. This may be a previous version of the camera module to that for which I found the datasheet.

The unit exudes high quality as found in the Industrial PM series cameras and the lens assembly will likely be very high quality as well.

Sadly I do not recognise any of the PCB’s in your pictures so cannot give you a steer on the pin outs or command protocol used. When we look at the chipset I may be able to ascertain whether this camera is based on one of the FLIR Industrial models that I have worked on. If so, it’s command set may be more AGEMA-FLIR than Indigo-FLIR

Fraser

[Ultrapurple]

Thanks Fraser - very interesting indeed. The plot thickens, as they say!

One gratifying thing worthy of note is that all the screws I undid had what appeared to be their original seals in place, so there's no suggestion that anyone had been inside before me - well, not since FLIR fastened the case shut on the production line.

I still hope I'm on course for at least basic functionality with just 12V in and video out. But we'll see. It'll be a little while before I can do any more investigation.

[Fraser]

The date codes on the IC’s suggest that this camera was built in 2005. That is around the time that FLIR bought Indigo but would explain why this appears to be a pre FLIR-Indigo merger design. 2005 would place this in the era of the FLIR Pxx series and the FLIR A40 static camera.

http://alacron.com/clientuploads/directory/Cameras/FLIR/A40M%20Researcher%20Datasheet.pdf

Fraser

[Fraser]

I will see what pictures I have of the FLIR A40 PCB’s as that may provide clues to your cameras design origins.

For those unaware, the A40 was a static QVGA thermal camera with excellent capabilities. It had a local keyboard for control via menus plus IEEE1394a and RS232 connectivity to a PC. It would make a good basis for the camera that Ultrapurple has and both were from around 2005.

Fraser

[Fraser]

Pictures of the A40 PCB's. Sadly, apart from the PCB identity notation being typically AGEMA Thermovision, there is no match that I can see. I have pictures of the SeaFLIR PCB's so will look for those as well.

Fraser

[Fraser]

A40 pictures continued....

Sadly, no match to the Dual FOV camera.

[Fraser]

Just checked the SeaFLIR PCB's for similarity. No match but some example pictures attached for interest.

A very different PCB style so the dual FOV PCB's are not from the 'Military/Law Enforcement division'.

Fraser

[Fraser]

Looking at the pictures of my AGEMA/FLIR Thermovision PM570 PCB's you will see why I thought the PM series might be the basis of the dual FOV camera electronics package. Note the design style and PCB ID format. The PM series ended with the PM695 released around 2000.

Pictures of my 1997 PM570 attached for interest.

Fraser

[Fraser]

After some more digging in my archives for FLIR/Indigo technology PCB pictures I have realised I am mistaken about the Photon core not being present in the camera...... I now believe that the Photon is indeed fitted  I looked at the likely depth of the microbolometer module at the rear of the cameras casing and think there is enough room for a Photon but I am not certain so Ultrapurple will need to investigate to confirm. The Photon is normally self contained so I assume that the other PCB's are associated with power and the lens control/auto focus.

Take a look at the attached pictures and look at the FFC flag pivot, the three front cover mounting screws and the front cover locating dowels. I think you will agree that we are looking in the face of a Photon.

The Photon teardown from which I borrowed the image is to be found here:

http://www.rtftechnologies.org/general/thermal-imager-flir-photon-320.html

I should of checked this before setting the Hares coursing by saying that the unit was not Photon based  Time to give the somewhat fatigued brain a rest me thinks.

Fraser

[Ultrapurple]

Thanks Fraser

I managed to squeeze in a little more time to disassemble the camera further and take snapshots of the main boards. They appear to be
- power board RIPO2
- serial interface RSIL 2
- logic / framestore RICO3
- what I'm calling (for now) 18Mbit store (the board with 2 chips that is definitely part of the core PCB set)

Anyone recognise anything?

[Fraser]

It uses the MC68340 ! ..... my old friend from the Thermovision PM series 

That processor board is called the “LiCo” in the PM series cameras and I immediately recognised the AGEMA hand in the “RiCo” board design    In the PM series that board is the heart of the camera that is connected to another large PCB that has its own processor board and deals with the output of the microbolometer. Your camera design is somewhat different however. I will need to study the pictures some more and try to make sense of them. Parts of the RiCo PCB look familiar, others not. I think I can recognise the video encoder stage as it looks similar to the PM695 LiCo that uses a FPGA in place of the dedicated BT series video encoder IC that went obsolete.

Can you check to see if any of the Photon ‘stack’ PCB’s are fitted behind the Microbolometer ? I am hoping that you will find all but the power PCB mounted on the photon and that FLIR are feeding the cameras digital output to the PCB that you have shown on the end of the FPC.

The camera will clearly comprise of some core elements that should not be hard to identify. I have made a list that draws on my knowledge of the AGEMA Thermovision PM series.

1. Power supply PCB with main power rails created for board set. Some power rails may be further processed locally on other PCB’s

2. Main processor PCB that controls the camera and possibly provides the auto focus function.Any video output often comes from this PCB.

3. Interface PCB between the microbolometer and LiCo PCB. This PCB can be a dedicated processing stage with its processed output being fed to the LiCo for use.

4. Communications interfaces. There is usually a Serial interface on the LiCo but other interface options may have their own dedicated PCB’s that feed into the LiCo.

5. Servo PCB. Where a camera has additional functionality such as auto focus and Zoom functions, there can be a dedicated servo driver board that is either ‘dumb’ or contains processing for the focus element position as part of an auto focus feedback loop. Heavy current drivers can be included on this board where larger servos are being used.

6. Display PCB.Where required a local display screen or EVF can be integrated into ten design. This is usually connected to the LiCO PCB that contains ten video encoder. Available video types include composite and S-Video but composite is often used for a EVF.

7. Battery power. where there is a need for battery power here will often be a dedicated Battery PCB that deals with the needs of the battery pack and connects to the main Power supply PCB.

Fraser

[Fraser]

Just comparing the PM series LiCo and the Dual FOV RiCo, the provenance is clear to me 

BT858 video encoder is located in the bottom left corner and has a FPGA to its right (covered by thermal pad in my picture) In later PM series cameras the BT858 was replaced by an FPGA. That area of the RiCo looks like what I saw in the PM695.

Looking at the left side of the PCB there is a lot that appears familiar to me from the PM series, and more specifically, the PM695. Look at the connector positions on the left of the PCB and compare the LiCo with the RiCo. The small additional FPC connector tor is used for the visible light camera input from a video encoder IC in the PM695 visible light camera ‘pod’. The PM series included a PCMCIA card port so IC’s relating to that function will not be needed in the dual FOV camera. From memory there was a PCI bus IC that serviced the PCMCIA port.

Looking at this RiCo PCB I have hopes that this camera will behave like a PM series camera and will present the user with an on screen control menu that is navigated by simple serial commands (Up, Down, Left, Right, Enter etc). More advanced commands would directly access functions and set parameters. If the same team designed the PM series and this dual FOV camera, there us a chance that the firmware is similar to that with which I am familiar. I have a copy of the Thermovision complete command set so we may be in luck 

I own a FLIR static camera that was developed from the PM570 but that uses PM570 PCB’s and just deleted the EVF and user controls. The Dual FOV camera appears to be a development of the PM695 using a new microbolometer, namely the Indigo Photon core. The PM series used a large and expensive microbolometer and by 2005 it would have been quite old technology to use in a new design.

This is really interesting to me 

Fraser

[Fraser]

The other side of the LiCo and RiCo PCB’s.....

Note that the large white connector at the centre of the LiCo PCB is for the PCMCIA card port. There are IC’s on this board associated with the PCMCIA memory storage system that are not needed on the dual FOV camera.

Fraser

[Fraser]

I have dug out some documentation that may be helpful/of interest 

I attach a picture of the FLIR Thermovision 320 that I own but that unit is basically a PM570 and not like the dual FOV camera we are looking at here. The Thermovision command set is common across PM series models so there is a good chance that the dual FOV camera uses the same command set. It is not a dedicated military FLIR design and was a commercial product so using the standard Thermovision command set seems logical. The cameras ID plate calls it a "THV" and this means it is from the Thermovision family

Fraser

[Fraser]

This old tread may be of interest.

https://www.eevblog.com/forum/thermal-imaging/flir-pm695595-teardown-pc-connectivity-firmware-modifications/50/

Ratatosk included pictures of his PM675 LiCo PCB that I am including here as a comparison to the earlier PM570 PCB.

Fraser

[Fraser]

I have not got a lot of time today to study the pictures of the dual FOV camera PCB’s but I have marked up the 12 pin I/O connector on the RiCo PCB that, in a PM570, provides the following connectivity via the I/O Lemo connectors

Composite Video out
S-Video Out C
S-Video Out Y
RS232 TX
RS 232 RX
Earphone audio out
Microphone audio in (with phantom power)
Signal ground

The +12V supply goes from the 10 pin Lemo connector directly to the Power PCB.

Fraser

[Fraser]

I just remembered, if serial communications can be established with the RiCo board, there is a member of this forum who has been successful in extracting the firmware from the PM series over the serial data link. He was then able to extract the command set from the firmware download  If the dual FOV camera is not using the standard command set for its control, it may be possible to download its firmware and obtain the commands that way 

Fraser

[Fraser]

I will add the pin out of the I/O connector later as I will need to dig out my PM570 notes.

There is no guarantee that the pin out is the same on the dual FOV camera of course but it may still help.

Fraser

[Ultrapurple]

Thanks Fraser

I am led to believe that document PM501708 is the technical specification for the THV Integrator 5/20. This is the source, on manualslib. The document I found it in is a manual for a cased camera that probably uses the same or very similar core + optics package, but which has a different set of connections to the outside world - and though I didn't find any info on the internal wiring harnesses, there is a lot of info on the UI that I haven't had a chance to look at yet.

[Fraser]

An excellent reference document 

Looks like you are going to need the standard Thermovision hand controller for local control. I have these but sadly they are very expensive    I was recently approached by someone wanting such a controller but, as he has no responded to my emails, I consider it back on the market. Sadly it costs almost as much as your camera though so we may need to consider other remote controller options.

Fraser

[Ultrapurple]

Ouch! I think a laptop and RS-232 are going to be the likely candidates - unless of course there's a simple way to reverse-engineer the FLIR box?

[Miek]

I just remembered, if serial communications can be established with the RiCo board, there is a member of this forum who has been successful in extracting the firmware from the PM series over the serial data link. He was then able to extract the command set from the firmware download  If the dual FOV camera is not using the standard command set for its control, it may be possible to download its firmware and obtain the commands that way 

Fraser

Here's the info on dumping the firmware over serial by the way: https://www.eevblog.com/forum/thermal-imaging/getting-digital-data-out-of-my-pm695/msg3186726/#new

I assume the official FLIR controller just has a microcontroller inside sending RS232 commands. It wouldn't be too hard to make one from scratch.
On my PM695 at least, the serial command to do button presses is:

Code: [Select]

BUTTON:arg1[,arg2]
where arg1 is one of U, R, D, L, ENT, ESC, F1, F2, LEV, SPN, EMI, DST
and arg2 is optional and is either 'P' or 'R' (I'm assuming press/release)

[Fraser]

I have just read the Sentinel manual and I am pleased to say that this camera looks simple to recommission 

The manual provides most of the information needed and my Command manual contains the serial commands you will need  This dual FOV camera is a pretty standard AGEMA/FLIR PM series static camera design.

It would be possible to build your own RS232 remote control that sends the required data strings to the camera. It is plain old fashioned RS232 and nothing unusual.

The camera may also be controllable from FLIR software of the era, such as Researcher 2001 or 2.8. There was also a Thermovision control program but I cannot remember if I have used it so do not know if it works with the PM series.

The Sentinel uses RS232 or RS485 for its control. I suggest sticking with the standard RS232 for simplicity. RS485 is common in CCTV systems for remote controlling PTZ’s but it appears that the common PELCO D command decoder is not included as standard so your unit may not understand it. It is likely that there was a board in the camera housing that manages the RS485 commands to the visible light and thermal cameras. I see no advantage in using RS485 for you.

The pin-out of the Sentinel camera connector gives you an excellent insight into what is likely present on the dual FOV modules D-Type connector. The ‘RS232 Service Port” is the one you will be using.

Fraser

[Fraser]

Miek,

Yes building a remote controller is pretty simple and an Arduino could easily meet that need  I had plans for such a project but life just got in the way 

Thanks for detailing the firmware download method 

Fraser