FLIR MS / PS / LS thermal scope teardown by Fraser

[Fraser]

This restoration is on hold for the minute as there is a case donor in the offing. If I receive it I will likely install my units innards into that case.

The next time you see the scope it may well have changed colour to the PS24/32 camouflage scheme 

I also have a sizable multi thermal camera repair commission arriving next week so that has to take priority.

I will update this thread as soon as there is news.

Fraser

[Fraser]

OK, it is confirmed, a kind benefactor is gifting me the parts I need to refurbish my MS224. I will also be receiving the controller PCB, but not the Tau core.
More from em when I have some news.

Fraser

[Chanc3]

That's great news! Glad to hear they managed to get one.

Interestingly, I've just opened up an A35 camera to discover it is also a Tau core, as suspected, with a Pleora GigE module. It's quite a tidy little package, however I wouldn't trust myself to take each PCB layer apart. I'll take some photos and maybe get another thread going. The reason I mention this is that there is also a readily available A65, which of course, has the 640x512 detector it in. My thinking is that the hardware is the same (not sure how I can verify), so then it may be possible to modify it to the full resolution. There seems to be numerous mentions in the camera's settings that the resolution can go up to 640, but again I'm not sure how I could do this. I think it has something to do with Pleora EBUS SDK with custom FLIR name spaces (called sensorResolution, with up to 4 values, with 0 being 640x512). Some of the config is customisable, however many settings are locked out. I'm investigating further and will let you know if I get anywhere, then hopefully it will help modify your core to its full potential.

[Bill W]

Can you reverse the lens design focal length / pixel pitch / pixel count / field of view ? That would tell you the sensor native pixel count ?
That was the 'smoking gun' on the FLIR E4-E8 on early datasheets.

Bill

www.fire-tics.co.uk

[Chanc3]

Can you reverse the lens design focal length / pixel pitch / pixel count / field of view ? That would tell you the sensor native pixel count ?
That was the 'smoking gun' on the FLIR E4-E8 on early datasheets.

Bill

www.fire-tics.co.uk

How would I go about reversing them?

[Fraser]

The donor FLIR scope arrived today My sincere thanks to the top notch Gentleman who donated it to me.

I attach some pictures of the unit. You will see that it is a camouflage case which raises an interesting possibility. The donor includes everything except the Tau core and the front panel plastics. That means I now have a spare controller PCB, EVF, eye cup, chassis and lithium battery 
Back to the camouflage colour scheme, as far as I know, this was a colour scheme used only on a limited number of PS32R scopes that had a longer range capability thanks to a telephoto lens. If this is the case, the donor unit will contain a PS32 firmware.

So what happens when you attach a MS224 240x180 Tau core to a PS32 320x240 controller PCB ? We shall see. My Tau core is clearly marked up as a Tau320 so maybe the reduced resolution is created by commands from the Scopes controller firmware ? If so I may just have also received an upgrade to the PS32 spec scope if I use the donors PCB 

Further notes on the donor unit. In line wit its 'covert' design, it does not appear to have button illumination. The visible front edge of the controller PCB is without the ambient light sensor and no unused pads are present. Interesting as this is a different version of the PCB to that used in the MS224. Could it be a Scout II I wonder ?

More on this when I have time to work on the unit. I am fully committed to the repair of 21 ARGUS 2 thermal cameras at the moment though. 

Fraser   

[Chanc3]

Is the button led backlighting changeable using that program you linked to earlier? I cannot seem to find the link now.


Sent from my iPad using Tapatalk

[Fraser]

Hi Chanc3,

Yes, the back lighting can be switched off using the FLIR configuration program.

Fraser

[Fraser]

The FLIR configuration tool is here:

https://flir.app.box.com/s/s9hqyuesr945k0bjhqem

Cayman is the code name for the PS/MS and LS scopes. The file is the PS32 firmware.

Fraser

[Bill W]

    Can you reverse the lens design focal length / pixel pitch / pixel count / field of view ? That would tell you the sensor native pixel count.  That was the 'smoking gun' on the FLIR E4-E8 on early datasheets.

How would I go about reversing them?


You need to run the design equations for field of view and as long as there is only one unknown you'll get the missing item.

So say we know the horizontal field of view is 50° from the datasheet, and the lens has 'FL = 9mm' written on it.
For this you can assume that it is a 'simple' lens.
Therefore when focused at infinity the 'simple' lens must be 9mm from the sensor.
Therefore centre to left/right side of the sensor is 9*TAN(25°) = 4.2mm, and the whole sensor active area is 8.4mm horizontally.

If we know that this is a 25um pitch sensor from the datasheet or sensor family then:
pixels = 8.4/0.025 = 336.  So it is most likely a 320 x 240 sensor given the accuracy to which the lens focal length was known.

Alternatively if this was advertised as a 160 resolution camera:
pixel pitch = 8.4/160 = 0.0525mm, 52 um
As no-one still makes sensors at that pitch we would conclude it is really a 320 sensor at 25um pitch being downsized, or possibly a 640 @12um.


    Bill

    www.fire-tics.co.uk

[Fraser]

Thanks for this Bill. Very interesting and helpful.

Fraser

[Chanc3]

    Can you reverse the lens design focal length / pixel pitch / pixel count / field of view ? That would tell you the sensor native pixel count.  That was the 'smoking gun' on the FLIR E4-E8 on early datasheets.

How would I go about reversing them?


You need to run the design equations for field of view and as long as there is only one unknown you'll get the missing item.

So say we know the horizontal field of view is 50° from the datasheet, and the lens has 'FL = 9mm' written on it.
For this you can assume that it is a 'simple' lens.
Therefore when focused at infinity the 'simple' lens must be 9mm from the sensor.
Therefore centre to left/right side of the sensor is 9*TAN(25°) = 4.2mm, and the whole sensor active area is 8.4mm horizontally.

If we know that this is a 25um pitch sensor from the datasheet or sensor family then:
pixels = 8.4/0.025 = 336.  So it is most likely a 320 x 240 sensor given the accuracy to which the lens focal length was known.

Alternatively if this was advertised as a 160 resolution camera:
pixel pitch = 8.4/160 = 0.0525mm, 52 um
As no-one still makes sensors at that pitch we would conclude it is really a 320 sensor at 25um pitch being downsized, or possibly a 640 @12um.


    Bill

    www.fire-tics.co.uk

Very interesting! Plugging that in for the spec of the Tau 324 I have here (9mm focal, 48°FOV), I get either ~800 pixels at 25um or ~1100 for 17 um.

Now to see if it's possible to modify the firmware!

[Bill W]

You forgot to divide the field of view by 2

 2*9* TAN(48°/2) = 8mm = 320x0.025

Bill

www.fire-tics.co.uk

[Chanc3]

You forgot to divide the field of view by 2

 2*9* TAN(48°/2) = 8mm = 320x0.025

Bill

www.fire-tics.co.uk

That makes more sense!

[Fraser]

I have disassembled the donor PM32 scope for inspection.

It is a very different PCB layout. Not sure why yet. Initial cursory observations are that the buttons do have back lighting, or at least it is fitted on the PCB, yet no light sensor on this version. I suspect it will be set as very low level back lighting. Of greater interest is the deletion of the FLASH memory chip adjacent to the ARM microprocessor. Hmmm makes me wonder if the MS224 has on board picture capture that was not used ? Ah, whilst typing this I have had a thought. The PC24 has a freeze frame function. The PS32 has a X2 zoom function. Maybe the PS24 and 32 PCB's are different for this reason ?

I am referring to the donor as a PS32 as that is what it is  I managed to read the latent image of the part number on the worn label that sits on the underside of the unit.

I should state that the PS32 did not switch on or charge its battery when received. A quick visual check identified a lump of green corrosion between the pins of one IC. I cleaned away the corrosion and the unit returned to operation and is currently charging its battery. This raises another matter with these scopes. Both PCB's had small spots of weird grey solder residue ? on  them. It looked like the crud you rub off your soldering iron tip. The MS224 had two 'blobs' on it and this PS32 also had two blobs, but in different positions. One was between the pins of the twin LED charge indicator and another was along the side of SW2 push button switch. I will did out the pictures I took of the MS224 PCB that showed the contamination. The green corrosion between the pins of the PS32 IC was vivid green and looked like a tiny sponge. Most odd. All contamination was lifted off of the PCB's using just IPA. The location of the green corrosion will be investigated further before final assembly.

As a further note, the aluminium chassis is slightly bent in the PS2. I suspect dropping these scopes caused them to bend their chassis. The damage is nowhere near as bad as that which I found on the MS224. For the price I paid for the PS32 (nothing) I am not complaining though 

OK now to the pictures of the PCB's. The MS224 still has gold test points adjacent to the torch LED. The PS32 PCB appears to have had test leads connected to its test points at some time as they have been soldered. I suspect that the PS32 was used as part of a project with the TAU and front panel separated from it. That may explain moisture getting onto the PCB ?

Fraser

[Fraser]

More pictures

[Fraser]

More pictures

[Fraser]

The USB common mode choke is damaged on both PCB's. It is vulnerable to impact from the TAU's metal case. I will be replacing these chokes with bypass links as they actually degrade USB performance. For those unaware, these chokes are only needed for EMI compliance and the USB chipset manufacturers actually advise against their use unless essential for reasons of emissions compliance.

Fraser

[Fraser]

Pictures of the weirs deposits found on the MS224 PCB. They are not corrosion but rather some form of contamination during production.

The deposits mat be clearly seen over the SW3 and U29 PCB legends. They are grey in colour and had a granular dust like appearance.

Fraser

[Fraser]

Things just got interesting 

I reassembled the PS32 PCB, Battery, TAU interface and TAU core.

Result: The scope booted, raised an alarm and shut down 

I had a little suspicion about the model of TAU that had been in the PS32 so fitted my MS224 TAU 1 interface PCB.

Result: The scope booted and displayed the TAU image perfectly..... BUT, the scope does light its error LED and no push button functions work.

Hmmmm, interesting.

I carefully observed the image being produced in the EVF. It looked clearer than I remembered on the MS224 EVF. This was more evidence of what I already suspected.

A quick check on the fitted KOPIN LCD driver IC confirmed my suspicions. It is an A910 instead of the A210 fitted in the MS224/PS24. The difference is simple. The A910 is 640x480 VGA and the A210 fitted in the MS224 is 320 x 240 QVGA 

My PS32 donor is in fact a SCOUT II that is normally fitted with a TAU 2, hence the issues when using a TAU 1 on it, and it upscaled to 640 x 480 resolution for a better EVF image.

Well I am chuffed to have a SCOUT II but now I need to see if it is possible to convince it to work with a TAU 1. Theer are obviously differences between the TAU 1 and TAU 2 interface boards. Maybe the pinouts are different at the TAU connector ? I will have to check. I am hoping that teh TAU is just not receiving its RS232 commands due to a pinout issue.

It would be great to be able to use the SCOUT II PCB due to the better EVF image and the added functionality of a composite video output capability via the USB port. This video out feature became available on the PS32R and SCOUT II PS32 so I suspect that I actually have a PS32R and that, though not stated by FLIR, it is a SCOUT II masquerading as a late model SCOUT I. All interesting stuff 

To clarify:

Standard SCOUT I PS32

320 x 240 pixel QVGA EVF
TAU 1 320 x 240 pixels
Video output via contacts on bottom of case

SCOUT I PS32R - Camouflage colour scheme and long range lens fitted

640 x 480 VGA EVF
TAU 2 320 x 240 pixels
Video output via USB socket

SCOUT II PS32

640 x 480 VGA EVF
TAU 2 320 x 240 pixels
Video output via USB socket

There are also the SCOUT I LS32 and LS64 scopes but those are always found in black cases.

My hypothesis would also explain the completely different PCB used in PS32 donor.

Sadly i must get on with other chores so more when I have time.

Fraser

[Fraser]

OK, an update.

I did some checking and a Tau 2 can be fitted in a Tau 1 based design. The reverse is also possible provided the additional functions of the Tau 2 are not being used. This suggests that my Tau 1 should have worked in the SCOUT II.

I inspected the Tau 2 interface board and did some tests. I found an open circuit power supply inductor for the core. This explained the lack of Tau video. The inductor was replaced easily.

Another test using the MS224 Tau 1 core in the SCOUT II PS32 proved successful  Now that I am using the Tau 2 interface board, the scope is happy and all functions are working fine. I now have the higher resolution EVF, USB video output and the X2 zoom function that was not available on the MS224. Whether or not I have 240 x 180 pixel resolution, or the full 320 x 240 pixels is not known yet and will take some more testing. If the controller board sets the Tau resolution I may be lucky. The Tau 1 is clearly marked up as a Tau 320 with no indication of resolution reduction in the part number.

Fraser

[Chanc3]

That's great news - I suspect the control board is what dictates the resolution of the camera. Fingers crossed that it is and you're experiencing the 320x240 res!

[Fraser]

Quick update......

1. I have confirmed my unit is a SCOUT II
2. The full TAU320 320x240 pixel resolution is presented to the user
3. The Viewfinder is the later 640 X 480 VGA resolution upscaled from QVGA
4. The composite video output is present at the mini USB socket
5. Enhanced SCOUT II firmware is loaded

Feeling pretty chuffed right now.

Fraser

[Fraser]

A side note for those interested in the TAU camera and its resolution.

FLIR do not resolution cripple TAU cores.

That is to say, if you have particular model of TAU 1 or 2, it will always output its stated resolution i.e.

TAU 160 - 160 X 120
TAU 320 - 320 X 240
TAU 640 - 640 X 480
TAU 1024 - 1024 X 768

FLIR fit the TAU core appropriate to the hosts spec and do not in any way reduce the resolution available at the TAU connector. So any thoughts on a TAU 640 being used in a camera specified as 320 X 240 pixels is sadly wishful thinking. FLIR insert the complete TAU core a bit like a Lego brick. It just plugs in so no need to cripple it in any way.

I thought this might be interesting to fellow TAU owners.

Fraser

[Fraser]

I neglected to mention..... There is no such thing as a TAU240 with 240 x 180 pixels

Fraser