Fraser has two new patients inbound - E2V ARGUS 4 & Seek Reveal Pro FF

[Fraser]

 I blame it all on the lockdown that the UK is in at the moment ! It made me do it, honest 

What am I talking about ? Well as regular readers of my posts will know, I have stopped my intensive searching of eBay for thermal imaging camera bargains since January 2018. My resolve weakened this week however and I bid on some suitably interesting and well priced "spares or repairs" cameras that caught my wandering thermal eye. I first saw the relatively rare sight on eBay of an E2V Argus 4 fire fighting camera that was being sold 'as-is'. It looked to be in pretty good condition and was shown working in the auction (always a good sign) The camera was not only an Argus 4, but the higher resolution model '320' that has a frame rate of 60fps. The model was stated as the higher specification HR320 but the camera straps show the standard spec '320' blue logos. We will have to wait until the camera arrives to establish its exact model and specification. Either way, both are QVGA and of very high quality 

I already have two ARGUS 4 series cameras and own several of the very nice optional accessory modules that fit into the battery bay. E2V Argus offered battery modules that provide Wi-Fi/LRT IP streaming, Wired Ethernet IP streaming, Analogue composite Video, Digital Video Recorder and External Power for static installations. It is a very versatile fire fighting and surveillance camera The camera has its own picture storage and can store imaged for later viewing or download via USB to a PC. More about the camera once it arrives.

The Seek Reveal Pro camera will be well known to many readers of this forum. I already own the standard <9fps and 15fps Fast Frame versions. The eBay auction was offering a faulty Reveal Pro Fast Frame so it was far more interesting to me than a standard Reveal model that the seller was also offering in another auction. For those unaware, the difference in image quality between a Standard Reveal and the Reveal PRO is significant and far more than just the pixel count. The reveal Pro contains a decent image processor IC and more features than the standard Reveal model. For this reason I was tempted to bid on it. The fault looks potentially nasty though. The auction image is pretty awful so little can be determined from it, but the seller stated that the display showed a thermal image with varying temperature measurement readout but with vertical lines present. It could be just a faulty LCD panel that I can source for around $30 from China, or it could be a fault in the microbolometer..... two ends of the fault spectrum in terms of repair ! Thankfully the Reveal cameras use a relatively common LCD panel. I am wondering whether the fault lies elsewhere however, such as the power supply buck and boost converters. The vertical lines appear equidistant from eachother, suggesting to me that an fixed frequency signal is involved. Hopefully it is not a failed microbolometer as that would render the unit a spares donor only, I do own a suitable spare Seek Thermal Pro microbolometer that could be fitted, but calibration would be an issue. We shall have to see what the symptoms suggest once I have the unit in my hands.

Needless to say, both of these cameras fall into the 'bargain' price category for eBay sales, especially at the moment with increased interest in thermal cameras and so, higher prices. I was fortunate but must harden my resolve to not search for any more 'bargains' 

Watch this space for an update on the two units once they arrive, hopefully later this week

Fraser

[Collat3ral]

I think I found the same bidder. Trying my luck on some items

[railrun]

Fraser, are you the buyer? Never seen this camera before.
https://www.ebay.com/itm/DYNAMIC-ATHERMALIZATION-FLIR-INFRARED-UNIT/402306300169?hash=item5dab52f909:g:Z6gAAOSwAwVe86fs#vi__app-cvip-panel

[Fraser]

No, not me 🙂

I do not recognise that camera or the manufacturer.

It will be a specialist product and the challenge with such is controlling them without the command set or driver software. As such I tend to avoid them unless a real bargain.

Fraser

[Fraser]

Collat3al,

Good luck.

The one with the dead battery looks promising. Be careful spending much on any that are not displaying a thermal scene as a dead microbolometer module renders them good for parts only. I bought the Reveal Pro because the seller stated that it’s measurement Function was responding to thermal stimulus so I am hoping the microbolometer module is OK. For information, the Reveal does not use a Mosaic core, or even the standard complete Seek Thermal Core module. It interfaces the microbolometer PCB direct to the video processor on the main board without the usual J3 core processor PCB that outputs USB. Buying these is very much a calculated risk as faults in the firmware memory or the Image Processing chipset will render them likely beyond repair.

Fraser

[Fraser]

The Argus 4 arrived today. The Seek Reveal Pro has yet to be dispatched 

The Argus 4 is an Ex Fire Brigade training unit that has an intermittent fault with the FFC flag. At first power on the camera sometimes suffers a stuck FFC shutter in the closed position. Further investigation will be required but it is hopefully a mechanical issue that can be easily rectified. Once ‘unstuck’ the FFC shutter works normally and an excellent image is produced. More about this camera when I get a chance to spend some time on it. It’s good condition and lack of the usual smoke smell is likely because it is a training unit so never saw a true fire in its life.

Fraser

[Fraser]

For information. I was going to bid on the Seek Reveal with a dead battery or charge circuit but if there is interest from others on the forum, I will step back on this camera. I believe that camera to be eminently repairable, but the other three to be somewhat of an unknown quantity ! I would not recommend buying a faulty Seek camera unless you can afford to write off the money spent on it. They are highly integrated units so faults can be a challenge to track down and repair. Caveat Emptor and all that. Please do not blame me if you buy a dud from this seller as they are clearly customer returns with unknown history. Like mobile phones they can suffer abuse, such as being dropped or submerged in water   I bought a camera that looked to be mainly functional so it was a calculated risk and worst case scenario, spares for my other two units.

Fraser

[Bill W]

The Argus 4 arrived today. The Seek Reveal Pro has yet to be dispatched 

The Argus 4 is an Ex Fire Brigade training unit that has an intermittent fault with the FFC flag. At first power on the camera sometimes suffers a stuck FFC shutter in the closed position. Further investigation will be required but it is hopefully a mechanical issue that can be easily rectified. Once ‘unstuck’ the FFC shutter works normally and an excellent image is produced. More about this camera when I get a chance to spend some time on it. It’s good condition and lack of the usual smoke smell is likely because it is a training unit so never saw a true fire in its life.

Fraser

Toss a coin between:
Part cold weld of the flag to the end stop where it has worn through the anidising.  Gentle polishing will fix, maybe some mylar tape.
Clearances issue, later cameras had a tiny ptfe washer added.

Either way a total pain to get to on these cameras - so a very comprehensive teardown is likely to result.

Bill

[Fraser]

Dear Bill,

Thanks for this.

So I have a voyage of exploration to undertake with the 320. The cold weld issue is a new one on me. I thought there might be a rubber end stop pad that had become slightly sticky.

Wish me luck with the journey into the deepest, darkest parts of the Argus 4 camera 

I used to dismantle and service Gov’t grade Combination locks and Siemens T100 electro mechanical teleprinters, so mechanical disassembly of equipment holds no fear for me

Fraser

[Fraser]

Bill,

Out of interest, why is the Argus 4 320 so much more complicated on the disassembly front than the 160 model ? Were other changes done beyond a different (physically larger) microbolometer that required a redesign of the front section component parts ?

Fraser

[Bill W]

Out of interest, why is the Argus 4 320 so much more complicated on the disassembly front than the 160 model ? Were other changes done beyond a different (physically larger) microbolometer that required a redesign of the front section component parts ?

The case was designed for the 160x120 35µm sensor, and so a 6.3mm f/1 lens.  No allowance was permitted (we did ask  ) for any larger sensor or lens.  "Firemen will never pay for a 320 sensor".
So on the original Argus4 the sensor PCB comes out as a complete assembly with minimal clearances.

So when we did (inevitably) put in a 320 sensor with an 8.6mm f/1 lens, and that larger diameter still needs to see out of the case, which is quite thick there.
So the lens ends up partly through the case wall and the sensor PCB assembly is captive until you undo the lens.
Naturally the lens needs a tool and has threadlock and a grub screw just to make it harder.

[Fraser]

Thanks Bill,

I now see the issue 

Thankfully you recommended a suitable screwdriver for the lens lock a while back and I bought one. I also have a range of adjustable lens spanner’s with various tips so hopefully have the covered as well. Acetone will deal with the thread lock varnish, as it did with another lens that you kindly gave to me 

Thank you for the picture. That really helps 

Don’t you just love it when the casing design team refuses to consider the expert input from the electronics payload team 

Fraser

[Bill W]

Don’t you just love it when the casing design team refuses to consider the expert input from the electronics payload team 

Fraser

To be fair to them, the mechanical folks were OK either way.  The problem was the product specification coming from sales.  When challenged with 'Can you guarantee that we will never ever ever need to fit a 320 sensor in this box ?' they replied 'Yes'.

This was despite still selling Argus 3 with 320 sensors in them at the time.  It was more that they were getting hit by Bullard T3's - smaller cheaper - so wanted something smaller and cheaper.

At least they were talked out of putting the batteries in the handle.  I was probably never forgiven for saying "OK, so how would you use the camera without the handle if that is where the batteries are ?'

Bill

[Fraser]

Bill,

One of the features of the Argus 4 Design that I love is the “option port” on the top of the casing that is home to the battery pack or several other option modules that incorporate a battery for the camera as well. It is a really excellent idea 
Now that I have Sourced most of the optional modules, my Argus 4 cameras are very versatile indeed

As to placing the battery in the handle....... nope ! As you say, the handle should be optional. A handle has its place but should not be a fixture unless it can fold up out of the way on a knuckle joint.

Fraser

[Fraser]

Just in case it helps anyone, I will provide my thoughts on the three faulty Reveal units that are on eBay from the same seller as my unit.

To be clear, these cameras are in no way connected with me and I do not know the seller. They are customer Raw returns that the seller has plugged in and switched on. We are reliant upon the sellers description and pictures provided but I can assess some aspects of the units from that information. I will not be held responsible for any errors I make however !

Seek Reveal XR30 - Camouflage

Physical condition? : Looks Good
Completes Boot ? : Yes
Displays expected OSD and thermal scene ? : Yes
Fault : The camera apparently will not run from its internal battery but will run from its USB socket. This suggests to me a failed battery (locked out?) or a failure in the charging/power management IC.
Repairable by competent person ?  : Yes, eminently repairable as the battery sub system is not complex.

Seek Reveal XR30 Black

Physical condition? : Looks Good
Completes Boot ? : Yes with a caveat - no thermal scene is produced on the display.
Displays expected OSD and thermal scene ? : No.
Fault : The camera apparently boots to the point of presenting the OSD bars but fails to display a thermal scene or temperature measurement. The lack of a temperature measurement may just be down to the measurement mode not being selected. The blank display could be a stuck/damaged FFC flag mechanism. This is quite fragile but I would not expect a 'perfect' blank screen in this particular FFC flag actuator design. The unit may well have booted to the point of displaying the OSD but halted due to an issue in the final stages of the boot sequence.
Repairable ? : We lack the information to determine the actual state of the camera after/during boot. It is an unknown and could be a basket case if the problem is in the VLSI side of the unit. Firmware corruption is a possibility but I cannot be sure of that. A fault on the Microbolometer cannot be excluded. This unit would be a risky purchase.

Seek Reveal Pro Black (carbon fibre look)

Physical condition? : Looks Good
Completes Boot ? : No - no thermal scene or OSD is produced on the display.
Displays expected OSD and thermal scene ? : No.
Fault : The camera apparently boots to the point of presenting the SEEK logo and Measurement overlay but fails to display the OSD bars or a  thermal scene.  The blank display could be a stuck/damaged FFC flag mechanism. This is quite fragile but I would not expect a 'perfect' blank screen in this particular FFC flag actuator design. Then there is the displayed temperature measurement.... it is reading 32F and that is 0C ! With a stuck FFC flag I would not expect to see a measurement of 32F(0C) That particular aspect of the displayed symptoms is worrying.
Repairable ? : We lack the information to determine the actual state of the camera after/during boot. From what can be seen though, the camera had not correctly completed the full boot sequence, or if it has, it has crashed soon after. The displayed temperature measurement is pure fiction and indicative of something very strange going on on the measurement side of the system. There is an ambient temperature monitoring sensor in the camera but failure of that would not produce a 0C measurement reading. To me that reading suggests a default value used in the cameras measurement system against which the readings from the microbolometer are offset. What is causing this I cannot say, but a failure on the microbolometer PCB should be considered a possibility. 


OK there are my thoughts on these units. I hope they help anyone interested in them. I already have my very own Seek Reveal project and I hope to repair it when it arrives. If you buy one of these three units, I wish you well with the repair but please do not ask me to help. I just do not have the time and remote repair is a total nightmare ! Please do consider how much to spend on any faulty SEEK product as working used units are selling at reasonable prices. A Reveal Pro in its box sold for just over 200GBP yesterday. These units are not that easy to work on and would not be what I would call a beginners project. Only spend money on these if you are willing to accept it may be beyond repair. There is a good reason why customer returns like these basically get dumped. They are not a commercial repair proposition.

Now if these were FLIR PM series industrial cameras..... I would be all over them like a rash as I love working on those little beauties 

Fraser

[Fraser]

The seller has added another Seek Reveal listing for a dead unit that does not start or operate the LED. The fault could be varied as no further information is provided except that the buttons feel “solid”. That does sound weird.

The seller has opened the unit to see if it was a dummy (display model). It apparently contains a main board and battery.

The only concern I have is that the seller even bothered to open the unit. Have the others also been opened I wonder. I personally like to work on virgin returns that have not been messed with in any way. Most returns resellers will not wield a screwdriver as they just want to shift stock. It makes me a little nervous that this seller is not afraid to open these units  We can only hope no parts swapping has been going on. Caveat Emptor and all that.

I am thinking the seller might be working their way through a bulk lot of Seek Reveal units from a company like Amazon. They appear to test a few, list them and then test a few more and so on. For anyone thinking of buying a Seek Reveal repair project it may pay to wait until something is listed that does not look like a total basket case or high risk. I am certainly pleased that I bought a Reveal Pro that looks function albeit with a display issue.

Fraser

[Fraser]

Anyone thinking of taking on a Seek Reveal repair project would be wise to look at Mikes teardown of a Reveal Pro and someone else’s teardown of a Reveal. I will add the links in a minute.

The Reveal and Reveal Pro use different main boards and Microbolometer PCB’s. The microbolometer calibration data is held on the main board chipset. Unlike designs with multiple PCB’s, the Reveal design does not lend itself to the old “make one good unit out of two dead ones” approach. You have to swap chips if you want to go down that path and then you have to consider any data held in the chips that is camera specific. I advise caution if anyone thinks the Reveal or any other Seek Thermal camera will be a simple fix if it has more than a simple power management or battery fault. A broken USB connector is always a possibility on a unit that will not charge. This is especially so on the Standard Reveal as it does not have USB socket strengthening.

Fraser

[Fraser]

Seek Reveal Pro teardown by Mike....

https://youtu.be/mgaBAm_sRnk

Fraser

[Fraser]

Seek Reveal Teardown by Alex Udanis.......

https://www.allaboutcircuits.com/news/teardown-tuesday-thermal-camera/

Fraser

[Fraser]

As can be seen in the above teardowns, the Reveal is a highly integrated electronics package so there is not much that can actually be ‘repaired’. If a fault is in the main processor, the camera is likely toast. If it is a BGA solder ball fault, rework may be an option but it is not a task for the faint hearted !

Firmware corruption can be a possibility, especially if a camera loses power during the update process. It is a sure way to brick a camera and is potentially unrecoverable by most users. Firmware updates that have gone wrong or attempts to hack firmware can result in a very dead camera that may not be worth the hours needed to unscramble the mess and repair it, even if possible. Flash memory corruption through no fault of the user is also a possibility and the seek Reveal is known to have an issue with Flash memory self corruption.

Fraser

[Fraser]

The Seek Reveal series are a good example of why I effectively ceased to be a component level repair technician in my working life. The level of integration can make a piece of equipment challenging to diagnose without diagnostic software and ‘bed of nails’ test jigs. When a PCB contains just a very few lumps of clever silicon running code that is not public and the power supplies and other supporting hardware is present and correct..... you are faced with board replacement as the fastest means of repair and it is sometimes the most economic path to take. This is why Radio Officers, like me, on Ships became obsolete. ‘Techs’ were then employed who only knew how to follow board change or even whole equipment change routines. The OEM repaired the kit back at the factory.

Modern electronics is becoming more and more integrated and is often more of a hardware-software symbiotic hybrid that can be very challenging to work on without inside knowledge and diagnostic utilities. The thermal cameras I tend to repair are from a slightly earlier period where multiple PCB’s were used and, whilst they remain complex, the PCB’s contain separate processors, RAM, ROM, I/O glue logic and FPGA’s. With those units I have a chance of locating the fault and correcting it. In the case of the FLIR PM series you are basically working on a MC68K series based computer much like the Amiga 520 !

[Fraser]

The teardown by Alex revealed the identity of the two security screws that secure the front metal section to the cameras casing. They are Bryce Pentagon centre pin screws that require a hollow pentagon driver bit. They are not pentalobe.

https://www.brycefastener.com/penta-plus-vandal-proof-screws.html

Fraser

[JimM]

The Reveal and Reveal Pro use different main boards and Microbolometer PCB’s.

That's really surprising to me. I would have thought Seek would have used the same main board for the Reveal and Reveal Pro for commonality of parts, firmware/software, etc. You could easily account for the difference in resolution between the Reveal and Reveal Pro in firmware code. Instead the board has different processor and even a different USB connector mounting.

You certainly can't liberate a Reveal to a Reveal Pro like you can a FLIR E4 to a FLIR E8 through firmware/software changes. (On second thought, maybe that's why Seek took the approach they did!)

[Fraser]

A tiny update on the a Argus 4 320 camera.

I accessed the cameras “About” page in its setup menu and it only has 111 hours on it 

Accessing the stored images revealed that it was used during fire fighter training from 2014 until late 2016. The cameras sticking FFC shutter problem initially occurred in early 2016. It was used in training for a little longer with the intermittent problem present, but looks to have been removed from service in late 2016. Why this camera was not repaired, I do not know. It was certainly not a cheap purchase and repair would normally be sought by the Brigade.

I just spent a few minutes listening to the camera as it booted with my ear pressed to the case. What you should hear after starting the camera is the FFC shutter being activated at very short intervals as the camera takes some FFC readings during the microbolometer warm up period. The period between FFC events then increases greatly once the microbolometer temperature has stabilised.

With this model of camera there is normally a very clear ‘clunk’ when the FFC shutter operates. On the faulty camera I carried out a series of 10 bench mounted starts with a run time of a few minutes each to see what happened. The blank screen remained throughout. When listening to the camera I could hear the faint sound of the FFC solenoid being energised but no associated ‘clunk’. This is because the FFC shutter was already in the activated position (closed). Having proven my suspicion about the stuck FFC shutter, I gave the cameras front rubber buffer a light tap and was rewarded with The thermal image appearing on the screen and the next FFC event occurred normally. The camera behaved itself perfectly after this little test exercise.

So, as Bill said, I will be delving deep into the microbolometer block assembly to carefully inspect clean and ‘Tweak’ the FFC flag mechanism to remove its penchant for sticking in the enclosed position if left inoperative for very long.

Bill can you provide any more detail on where the thin PTFE washer was placed to create a better clearance please ?

I like faults like this. With many mechanical issues you see or hear what is happening, or not happening and your ears combined with an experienced diagnostic brain can image the problem before you even see it. It is like listening to a car engine, you get to know what it should sound like and any deviation in that sound profile is often detected by the experienced ear of the owner or mechanic.

Fraser

[Fraser]

The faulty Seek Reveal Pro arrived a few minutes ago 

The unit arrived in its original “Export Controlled” box along with its lanyard and USB cable. Battery was at 80% charge.

I switched the camera on and was pleased to see a perfect boot screen, followed by a noisy thermal scene. The LCD panel is perfect, no fault there. One golden rule before reaching for the screwdrivers is to full assess the patient and note any symptoms that might help localise the fault location to a sub system within the camera. Once this is done, always carry out a full system restore if such is offered. This can flush out ‘bad’ data from the configuration that could be causing issues. A full cold boot is also recommended but that may require removal of the battery in some equipment.

OK, to my Reveal Pro patient...... I am very pleased with what I am seeing. The camera is correctly imaging the thermal scene and trying to measure the spot temperature. The problem is that there are vertical lines all over the scene so the measurement is jumping all over the place. I checked stored images and they are the same except with static vertical lines of course. All menu options operate as expected with no other fault symptoms. A system restore did not change the symptoms.

From this simple observation I am able to make some early deductions......

1. Battery appears OK
2, firmware is OK and not corrupted
3. System is running and not crashing so no fault is being detected or encountered
4. System is oblivious to the vertical lines present being a fault and is treating them as thermal data from microbolomter.
5. The vertical lines are dynamic and fast moving in the lateral direction when viewed. Some might describe them as random  rather than predictable or static.


The Seek Reveal operates with the thermal core rotated 90 degrees to provide a portrait image format. The human brain interprets what it sees and tries to rationalise it using memories. The vertical noise looks like nothing particularly familiar but when you rotate the camera 90 degree to create a landscape image the lines immediately take on a more familiar pattern (for me anyway)

With the camera display in landscape orientation the noise lines are clearly just that..... noise. They are what I have come to expect in an analogue system that has noise on a power supply rail feeding a sub system that directly deals with image creation from an analogue input. The power rail noise becomes superimposed on the analogue signal and becomes part of that signal, just like a mixer..... though an unwelcome addition is added to the wanted signal.

So from my very early days diagnosis I suspect that there is a Buck or Boost converter that is producing a noisy output and that power rail is used to power the microbolometer PCB, or an analogue sub system that follows, before the high speed ADC translates the analogue image into the digital domain used by the image processor. It could even be noise on the ADC supply rail.

Well that will do for now. I am very pleased with my purchase and hope to return it to full health when I get some spare time. Such is in short supply at the moment. I have not had to open the casing yet ..... I like this kind of closed case diagnostics 

As a side note...... From witness marks on the screw heads, I have a suspicion that someone went in search of easy to fix ‘loose wires or connectors’. Not an unreasonable action but not what I like to see as it can indicate ‘fiddle fingers’ have been inside a unit. If anyone is thinking of buying one of these faulty units in the hope of a loose connector type fault, I suspect you will be disappointed if the seller has been inside looking for the same.

We do not know the history of these units but the sellers company name is known to me and they are basically a professional WEEE waste processor who take electronic waste and recycle it. They clearly also do a side line in selling some kit on eBay ! There is likely a tidy profit to be had as such waste disposal companies normally charge the originating company for disposal of WEEE waste based on weight.

I suspect these Seek Cameras came from a large retailer as Seek Thermal would likely have wanted them crushed if a U.K. service centre was the source.

Fraser