-
The story of another thermal camera purchase by Fraser
If you are sitting comfortably, I shall begin
As many readers will already know, I collect and repair thermal cameras as my hobby.
On many occasions I have decided to stop buying these wonderful pieces of technology, but my will is weak in this area. Inevitably an unusual thermal camera will stray into my field of view, at a price that just begs my attention.....and my resolve will buckle under the strain !
Such an occurrence recently resulted in my becoming the proud owner of another unusual thermal camera. Before I go on, I have decided to not identify the camera or its manufacturer at this time. Such details are not essential to the story yet could cause some issues as you may see as the story progresses.
So what is unusual about this thermal camera ? Well its specs are nothing earth shattering at 384 x 288 pixels producing a 60fps output to the user. The NETD is specified as <70mK but in high sensitivity mode this drops to <35mk. None too shabby really. This improvement is made possible by carefully set bias voltages and the use of a thermo-electrically temperature stabilised microbolometer array. The camera outputs are many. It has Firewire, RS232, Composite Video (NTSC or PAL), S-Video (Y & C) and VGA. Plenty to be meet the varied needs of a user.
This camera has no control panel, no LCD display and no viewfinder...... making it not very user friendly eh ? Well it was never intended to operate in a stand-alone mode. It is a specialist industrial thermal camera module that is intended to be incorporated into a larger machine. It is not just a 'Core' as it is a complete thermal camera with a full gambit of on board capabilities. It just lacks the normal human interfaces. The camera is controlled from a computer that sets up the modes of operation and the image is either streamed to the computer or displayed on a suitable video monitor. Control may be via Firewire or RS232 and the user interface is a bespoke PC program. It is possible to command the camera over RS232 using a simple terminal program but this has its complications as I shall shortly detail.
The camera actually appears very boring from the outside. It is just a 'box camera' and, to those unfamiliar with thermal camera optics, it looks very much like the usual heavy duty industrial visible light cameras that are so common in industry. It build shouts 'ruggedized' and such cameras are often very expensive. The seller knew this so its price was set quite high. In truth the seller likely knew nothing of the camera, beyond, "it looks shiny and expensive" !
In communications with the seller I established that the camera was purchased at general auction without any other accessories or paperwork. I received the serial number and sent that to the manufacturer to check that it was legitimate. It was and it was originally sold in 2014 ! It was virtually new compared to most used industrial thermal cameras sold. That may be explained later though, so hang in there
The seller could not test the camera or state that it was working. This is one of those occasions when you have to decide whether the risk is worth taking. My thought process was as follows:
Pro's:
1. Very nice cosmetic condition industrial thermal camera
2. Decent resolution and frame rate
3. Lemo connectors shouted quality
4. It is fitted with a large 50mm close up lens !
5. It is an auto and manual focus camera
6. The camera has an SD card recording option
7. The camera is made by a reputable company and I found data on it.
8. The price might be negotiable
9. The camera is only around 2 years old
10. It is unusual, so I like it
Con's
1. It is an industrial modular camera, so not totally self contained.
2. The camera is in unknown, untested condition and may be faulty or degraded.
3. No leads or software come with the camera. Both are expensive from the OEM
4. The camera needs bespoke software to control it as the command set is not public
5. The OEM is unlikely to provide much assistance
6. The camera is not cheap if it turns out to be a paperweight !
After some thought and communications with the seller of the camera, I decided to risk a sum of money with which I was comfortable. This is essential when considering the purchase of an unknown condition thermal camera. They can end up as expensive paperweights. I knew the lens would be useful even if the camera was scrap. I offered the seller £200 and he happily accepted. To him, this was a total unknown and he was happy to get some decent money for it.
The camera arrived in due course and I quickly inspected it for signs of abuse. The Lemo sockets were all in good order, the Aluminium case had some bumps on the corners and one mounting lug on the lens was badly bent. Hmmm it looked like the camera had suffered impacts to both its case and the lens. There were no dents in the lens barrel but there was a witness mark from a blunt metallic impact on the side with the bent lug. These are Industrial cameras so such damage is not that worrying as they are tough and designed to take a beating.
I had already procured the required LEMO connectors ..... total cost £45 for the two from Germany, and that was the cheapest source ! that is a consideration when buying a camera without cables.... if LEMO or Fischer connectors are used, they can be very expensive. It is sometimes the case that it is better to change the connector on the camera to LEMO from Fischer as the Fischer types are both more expensive and harder to source in the UK.
Now to the challenging part. I had been in communications with the OEM and had provided the serial number so that they could check their records. I was up front about buying the camera used and not being the original purchaser. They seemed happy to talk to me, which was a bonus, but it was not all good news. I had asked them if they were willing to supply me with the pin-out of the cameras connectors, any documentation on the command set and a user manual for my model. The response was polite but disappointing. The pin-out was part of an extended documentation set that was chargeable. The documentation set also includes the command set for the camera. Neither could be supplied to me for free. The software was also a chargeable item. It looked like I had hit the proverbial brick wall with the OEM as everything was chargeable, and at high prices. We are talking thousands of Pounds here and not pennies
This is the risk you take when buying specialist thermal cameras or equipment that requires bespoke software, or a command set that is not in the public domain. It was looking as if I had myself a £200 paperweight.
I am not one to give up so easily though. I asked the OEM if they would be willing to send me just the user manual. This they were happy to do and I received a 20MB PDF by return. The user manual made interesting reading but sadly no remote control information is provided, except to refer the reader to the 'Interface and control' document that may be purchased from the agent or OEM.
I knew I could reverse engineer the cameras interfaces so was not concerned about not having the connector pin-outs. What I needed was either the command set or the software. Both seemed out of reach but if there is one thing I know about the internet, it has all manner of information and software squirreled away in dark corners out of the normal public gaze. I am experienced in searching out such assets on legitimate servers and not hacking sites. A long and comprehensive search brought fruit in the form of an official Agent who was happily providing a download of the software that I was searching for. It is the original software from when the camera was released. This has been replaced with a brand new software package so maybe the old version was made available for download as a legacy support decision ? The OEM was offering me the latest software and made no mention for the earlier software so maybe it is considered obsolete and no longer offered by them ? I downloaded the software from the dealers support page and tried to install it on my Windows 7 laptop It installed without the need for any keys or passwords. It runs perfectly and has no signs of being a demo, or limited use version. Happy days
So I now had the essential software to configure and use the camera. I could intercept the RS232 communications between the PC and camera now and so reverse engineer the command set for the camera
So where are we at this point then.....
I have a camera that remains untested
I have the required LEMO connectors
I have the required OEM software
I have the user manual for the camera and software
I do not have proof that the camera is in an operational state or 'safe' to apply power
I do not know if damage or degradation exists within the camera or microbolometer
I do not know the pin-out of the two LEMO connectors
Next obvious step ..... open her up
And that is what I did
Opening the cameras case was very straight forwards with just two hidden screws. One under a rubber foot, and one under the ID label. Pretty common stuff really.
With one side of the metal case removed, I could inspect the internal parts for evidence of damage. One thing you learn when doing forensic investigations is to use your eyes first and not your hands. I visually inspected the camera and notes some interesting facts about its interior.
1. We already know that there are impact witness marks on the case and lens barrel
2. We also know that one lens retaining lug is bent as the lens was forced sideways
3. There were small flakes of matt black paint floating around inside the cameras case
4. A microprocessor daughter board was only partially retained in its edge connector (similar to a SIMM connector). One end was completely out of the socket. If I had applied power to the camera, all manner of issues could have been caused if the edge connector was shorted by miss-aligned connections.
5. There was no sign of where the mat black paint was being used in the cameras chassis or optical path.
6. The camera looked in good physical condition with no evidence of previous entry or distortion of the mechanical elements.
I know the camera had suffered some form of impact to its lens and case but thankfully it seems to have survived.
I made another discovery whist inspecting the camera. There is a small micro switch positioned behind one of the SD Card cover retaining screws. Both the screws are too short to operate the micro switch. The SD Card recording capability is a chargeable option on this camera. There are two possibilities for the micro switch that come to mind.
1. The SD Card option is enabled by fitting longer screws to the SD card cover ! It would not be the first time that equipment 'upgrades' and options were a very simple physical change like a jumper link or push button sequence etc.
2. The micro switch could actually be a 'hard reset' button concealed behind the sealed SD Card Cover. This protects the camera from accidental reset and dirt ingress if a hole was provide for reset button access. Saying that, such a hole could easily be sealed with a blind grommet. Maybe the designer though this a more elegant solution?
At this point in time I do not know the function of the micro switch. It is connected to an I/O board expander that talks I2C. when I get the camera running I can quickly work out its function though
I was able to quickly and easily dismantle the cameras internal parts down to module and PCB level. Each was then visually inspected for damage or interesting elements.
It was immediately obvious that the cameras main board comes from one of the OEM's camcorder style industrial thermography cameras. In fact I think I know which model. There are several unused ribbon connectors and headers present. The same is true on the microbolometer PCB. The camera does have a hardware option that is not present on my camera but this does not explain all the unused connectors. The camcorder style camera would have the keyboard and viewfinder connected to this PCB.
Upon examination of the main PCB, the architecture of the camera became pretty obvious. No surprises here at all. Thankfully the OEM used COTS IC's and the datasheets are freely available to download. The date codes on the IC's suggest that the camera was built in 2009 which is earlier than I expected for a camera sold in 2014. Maybe they had a lot of stock in the warehouse ? I will likely add pictures of the PCB later but it is nothing unusual or very exciting really. The use of a daughter board for a microprocessor proved a vulnerability as the sprung retainers were unable to keep a grip on the PCB during the impact event. Judicial use of RTV silicone might have addressed this vulnerability, as it does on rugged fire fighting camera internal header connectors.
Another daughter board provides the required major supply rails for the camera. It is a very neat design and easily reverse engineered. It uses a pair of LM5642 IC's to take the input supply voltage and generate 5V0, 5v2, 2V5, and 3V3. No negative rails are present at this boards output. The voltages are annotated on the PCB which is kind of the designer. How often I have wished other designers would do the same. It makes it so much easier to check for correct supply rails.
The 'optical block' with associated microbolometer PCB is interesting in that the whole microbolometer PCB, and its associated chassis, is moved forwards and backwards on linear ball bearing runners that fasten the microbolometer chassis to the optical block chassis. Two stepper actuators are present. One drives a circular plate that contains three different apertures and the other drives the microbolometer chassis back and forth. There is a challenge to the designer when the microbolometer moves. The microbolometer in this camera needs a heat-sink on its rear against which the Peltier temperature stabiliser may operate. How do you attach the microbolometer to the case such that it acts as the heat-sink ? Well this cameras designer has used an impressive form of flexible 'heat pipe'. The microbolometer has a special 'retainer' attached to its rear. This retainer holds four very large copper braided cables. These braided cables are then attached to the case via a bracket. The flexed braided cables permit the microbolometer chassis/sled to move whilst maintaining the required thermal path to the metal case. A neat solution to a challenging problem. FLIR take a different approach in the PM series cameras. They fix the microbolometer to a heat-sink that is in turn attached solidly to the metal case. The focus is accomplished by moving the middle lens in its optical block. The PM series do not have a removable lens however so they do not need to move the microbolometer. Professional thermal cameras are often a combination of high performance electronics and clever mechanical engineering. This camera is no different.
Whilst investigating the optical block I found more flakes of matt black paint. Even the microbolometer had matt black paint specs on its window. It was time to track down the source of the paint flakes.
Where is matt black paint commonly found in a thermal camera ?
Matt black paint is often an excellent black body surface and has excellent emissivity properties. It is anti reflective in a thermal camera optical path. The inside of lens barrels, areas around the microbolometer and the FFC flag are common areas where you will find such paint.
I had inspected the lens and could find no source of the paint flakes. Two areas remained tat I needed to access, the area around the microbolometer and the rotating aperture plate. This was not going to be easy however as I established that I would need to completely dismantle the optical block chassis. Not a job I take lightly as all screws are locked, and for good reason. Dismantling the chassis presents unknown risks such as alignment loss as I have no service manual to help me. You have to be very careful and methodical when undertaking such a task. I would not recommend such to the beginner and in all cases I take lots of pictures so that I can retrace the disassembly steps and see 'what goes where' !
I disassembled the optical block and inspected the parts. I could see where the matt black paint flakes had originated before spalling away and into the cameras interior.
There is a small insignificant metal plate mounted adjacent to the microbolometer. It is likely present to prevent some imaging artefact due to internal reflections or off axis energy from the lens. It may be small but the fact that it was shedding paint flakes in to the cameras optical block was very significant and very bad for the cameras imaging health.
Why is the paint flaking off the small metal plate ?
I cannot be certain as I do not have enough evidence, but I suspect that the impact event that the camera suffered put a shock wave through the optical block via the lens mount, and through the rotating aperture plate shaft and rear bearing. The plate rests on the rear bearing and this flexed the thin metal plate. The paint on it is very brittle so spalled easily. There is no evidence of a bent chassis. the camera received impact energy from two directions. To the side of the lens barrel, and the rear of the case. There may have been quite a lot of energy passing through the camera and its internal parts.
To repair the paint damage to the small plate, I removed all loose paint, roughened the surface with very fine 'wet & dry' paper before cleaning all dust off and re-spraying with an excellent optical black paint that was recommended to me by an industry insider.
The plate is currently drying in the conservatory and I shall await the hardening of the paint before re-fitting it. Sadly I cannot re-assemble the optical block until that small plate is in place. It is buried deep within the optical block All work has stopped as a result.
The camera case and all component parts will be cleaned with compressed air to remove all traces of the paint that was floating around inside the camera.
Well that is the story so far. I shall re-assemble the camera and then see what happens when power is applied. An RS232 link will be used initially to talk to the camera as it has less chance of issues than using Firewire. Once I am sure that the camera is working correctly, I will try accessing it via Firewire and video streaming etc.
I hope this little story has been of some interest. I am not detailing the camera model or OEM as I may need more assistance from them, and they may not appreciate me showing a complete tear down of their product. Maybe later
Watch this space
Fraser
-
Riveting
Tim -
I forgot to share my thoughts on how a 2 year old industrial thermal camera came to be sold in a General auction. I will correct that now......
The camera was purchased along with others in 2014. All were fitted with Close-up lenses so they were going to be up close to something being monitored. The cameras were bought with the most basic support package of 1 year. This is unusual in industry for such an expensive piece of equipment.
No software or cables were purchased with the cameras. They appeared to be either spares or replacements for in-situ cameras.
I have more information that suggests to me that these cameras were almost sacraficial in their tasking. That is to say, the project justified risking destruction of some thermal cameras. Big budgets were involved.
I have no proof, but it is my belief that this camera was damaged during whatever testing was being carried out with it present, and it was written off as a project consumable. It failed due to 'abuse' rather than normal use. This may explain the lack of a support contract. If the camera is likely to be destroyed within the year or so of use, a support contract is not much use. Remember, the owner could afford to write off such a camera as a project consumable.
Scary though this might sound, it is not that uncommon. It is a matter of relativity. If I on my average income destroyed a $6K camera, I would be gutted, however a project with a $100 Million budget would lose no sleep over such a loss provided it could be accounted for and the end justified the means. I have seen new equipment destroyed in testing and the car industry does it all the time during new model safety and performance testing.
My camera failed to work after an impact event that dislodged the microprocessor daughter board. The case was not even opened to investigate. It was written off.
This camera was not the only casualty of the project. There were others similarly disposed of as scrap.
Lucky me though
Fraser -
Yup. A local GM (General Motors) facility was about to open up, then, 20 million later, they just up and decided -- "meh!" and the whole thing was taken apart and auctioned off for pennies on the dollar. Same idea.
My PPoE picked up some steel shelving and cage / securing hardware from that. (This was over 5 years ago.)
Tim -
Some pictures of the disassembled optical block and microbolometer chassis.
The small plate that was losing paint may be seen on the right of the images.
Just look at the complexity of the metal parts used in this chassis. All are milled from solid blocks of Aluminium !
The linear ball bearing slides (like drawer runners) may also be seen. The microbolometer chassis mounts onto these runners.
The aperture disc with its matt black painted rear is also clearly visible.
I have included a picture of the microswitch that is positioned behind one of the SD Card cover screw holes.
Note that the outer case is milled out of solid aluminium as well !
The microbolometer PCB is not shown but I can advise that the camera uses the venerable ULIS UL 03 04 1 which is a 43mK NETD sensor array. It is capable of 27mK NETD when 'tuned' for certain applications.
More pictures to come in due course, once I have removed OEM details
Fraser
-
Some pictures of the troublesome little painted plate that necessitated my disassembly of the whole optical block !
You can see where the paint has shed. The last picture is of the repainted plate drying in the conservatory.
Fraser
-
Great story! Reads like a novel...
But now I am suffering in silence: will the camera boot up and work? Will the results be great? Was it worth it?
Jesus, you have so much patience to FIRST disassemble the whole device to check for faults, painting the plate, etc., before trying it out. I wish I was as methodical. My inner urge is to try out as fast as possible... In this case I would have tried to boot up the camera before painting the plate...
This means that if not anything else, your many descriptions help me to educate myself to get a grip on my urge and check out equipment calmly before hooking it up for a first quick test. Kudos to you for that. I have seen some professionals like you, who do things without hurry. Inspect, measure, clean. Let it rest, think calmly, get a good sleep, wait until everything is quiet so you have time to do stuff without pressure. This is something one has to learn. Otherwise, when under stress, the slightest mistake and you break something.
Keep your promise and continue this wonderful strory!
Regards,
Vitor -
Vitor,
It is 'normal' to want to apply power and see what happens.
I was trained to show restraint though
If you receive a piece of equipment and have any suspicions that it may have had a rough ride or has suffered some damage, it is important to inspect it for issues within the case. Some issues can cause more damage if power is applied as parts can be shorted or jammed. Applying power in an uncontrolled manner is a bit like Russian Roulette. If you are lucky, all will be well, if not, you can cause serious damage to electronics or mechacon systems.
In the case of the camera in this thread, there was obvious evidence of an impact to the camera casing. Once opened I could immediately see that the microprocessor daughter board was dislodged from its socket. Had I applied power without opening the camera, it would not have booted but power supply rails may have started and been shorted by the misaligned edge contacts in the socket. not a good situation. There was also paint flakes in the case and this could have been caused by distortion of a metal painted part. Such could be serious enough to jam a mechanism and overload the motor or driver circuit. Again, not a great idea as more damage can occur. There is also the issue of the paint contamination getting on the microbolometer window and being scuffed across it by an FFC flag. Not a great idea for an optical surface!
So in short, though I was keen to apply power to the camera, doing so might have caused more issues. If I had just re-seated the microprocessor board I could have started the camera with less risk but I was keen to discover the source of the paint flakes as that can be a sign of bent metal parts. This is where my training kicked in. I decided to fully investigate the effects of what is likely to have been a hard impact on the camera before applying power.
Also bear in mind that I did not have the pin-out of the two connectors so I needed to partially disassemble the camera in order to reverse engineer the main PCB. To remove that PCB, I had to remove the optical block !
Once I saw the source of the paint flakes, I could not ignore it and it was a case of just pushing onwards until I could get that part out and repair its surface. It is positioned at the heart of the optical block though and required the complete disassembly of the chassis. That was not something I was keen to do, as I said, but I am a perfectionist by nature and training so I could not ignore the issue and it might cause further issues in the future if more paint flaked away.
After this little exploration of the camera, I am now familiar with every part of it. Not a bad idea when you want to learn about a camera design. Patience is required though.
I have another really nice thermal camera that I bought with a known fault. It is a very expensive NEC AVIO camcorder style high performance camera that was displaying "Stabilizer" as a fault warning. The unit came with its service manual and that was next to useless ! It was just a board replacement manual. For a "Stabilizer" fault it just said to fit a new main board and carry out full calibration ! The previous owner was in a position to discuss the fault with the OEM and they were stumped, saying to just replace the main board and see what happens. The main board is several thousand pounds as a spare part and comes without any calibration entries of course. The owner of the camera sold it to me as a 'lost cause' for £700.
Now £700 is a lot of money to pay for a faulty camera but I had a feeling that i could repair it. These particular cameras sell for around £7K on the secondary market. I decided to take the risk. I was working then so £700 was less scary than it is now that i am retired !
The camera arrived and I carried out my usual visual inspection of the casing looking for any evidence of abuse. I found plenty !
The camera had been a rental unit that came back broken one day. I could see that it had been dropped on its lens ring
The camera is slightly front heavy so this was no surprise. the lens ring had already been removed as it just unscrews. I could see the witness marks on the front casing from the impact though. The camera was otherwise in great condition.
Once I saw the impact damage I knew I needed to inspect the internal parts. To cut a long story short, the camera had suffered minor distortion of the lens mounting plate (aluminium and easily corrected) plus the distortion of the quite thin aluminium main chassis. The main chassis holds the battery bay and the battery comprises two 18650 cells so it has some weight to it. The impact with the ground resulted in the battery slamming against the end of its battery tray and bending the aluminium 'cage in which the battery bay resides. The aluminium cage sits directly behind the microbolometer PCB and the distortion of the cage caused it to touch the PCB ! This resulted in the shorting of a critical power rails to 0V as the chassis sits on the 0V rail. The shorting of the power rail caused the "stabilizer" fault as the Thermo Electric temperature stabilizer was powered from this supply rail. The gap between the chassis and the PCB was normally around 2mm, which is not much ! On inspection I saw that the chassis was slightly bent but could not see the shorting point as it was obscured by other chassis parts. I proved the fault cause when I separated the chassis and microbolometer PCB for testing and the camera booted without the "stabilizer" fault It was then just a case of visually working out what the chassis was doing to the microbolometer PCB when joined together.
That was a fun repair and well worth the effort as she is a lovely camera. I was lucky though. Both the previous owner and I had applied power in order to ascertain the fault symptoms and neither of us had been able to visually detect the distortion of the chassis that was causing a short to 0V of a supply rail. Testing of the TEC circuit would have revealed the supply rail issue but I did not get to that stage before discovering the cause of the problem during disassembly. Also, having power applied to a shorted supply rail for a prolonged period during tests could have caused more issues. I was lucky.
Visual inspection is a crucial element of good trade craft when repairing equipment. Always look for signs of abuse or other potential causes of failure. This is why a methodical approach to disassembly is important. you need to keep looking for clues as sometimes the cause is right in front of you staring you in the face. Electronics are quite reliable in general, so external influences can lead to faults through no fault of the components. If you dive in pulling stuff apart, cleaning surfaces and blowing the dirt out as you go, you sometimes miss or obscure useful clues Like the small spot of brown 'stuff I found on the underside of an equipment top casing air grille...... directly below I found a spot of the 'brown stuff' on an IC's pins. The 'stuff' was likely coffee and a drip had gone through the air grille and onto the PCB below, causing corrosion over time and malfunctioning of the IC. The power supply pins were involved and one had virtually corroded away causing issues for the IC. Now i could have found that fault by tests but I found it a lot quicker by carefully inspecting the equipment casing, inside and out for signs of abuse.
The MK1 eyeball is a very powerful tool and can be a very effective fault finding tool
Look first, prod later
.... and only apply power when safe to do so ! 
Fraser -
If this is a machine vision camera (despite having a LWIR sensor) and has FireWire then it seems highly likely that:
a) maybe it can be powered via FireWire (dunno about the power consumption of the TEC or the shutter),
b) should conform to the IIDC1394 standard and therefore interfacing to the PC should be easy (i.e. dc1394 + coriander on Linux). This is not the case for GigE/USB cameras which almost always require vendor driver.
This is based on my experience with daylight/NIR machine vision cameras from PointGrey/IDS uEye/Basler/AlliedVisiion - my pockets are not deep enough to collect thermal ones, so I collect these (and RGB-D cameras)
-
Hi Lukier,
Many thanks for this.
Indeed the camera can be powered via a desktop PC Firewire port. The user manual states that laptop Firewire ports will need a powered hub due t their lower current capability.
Thank you for the hints on the Firewire interface standard fro cameras. I will look into that. Firewire is not my favourite interface, due to past bad experiences with drivers etc, but it is very efficient and capable for video streaming. I hd all manner of problems getting Firewire thermal cameras to talk to a Firewire PCI card in a desktop PC and also with IEEE1394 PCMCIA cards in laptops. The drivers would not co-operate with the cameras in spite of them being the correct ones for the job. I found debugging the problem both time consuming and annoying as I had no hints as to what was going wrong. At least with RS232 you can monitor communications easily. One NEC AVIO camera continues to challenge me due to its Firewire issues. The camera is seen by the PC but it ignores all commands sent to it, casuing a hang in the control software. I spent hours working on that problem and concluded that something is wrong at the camera end of the link. The camera works fine in all other respects and responds to its dedicated remote controller (RS232)
Interestingly I have since read a guide to Firewire connectivity that FLIR produced. They advise that most PCI card or PCMCIA card OEM drivers will not work well and that the user should always select the generic Firewire 1394 OHCI drivers provided by Windows. They further advised that most PCMCIA IEEE1394 cards would not work with the cameras.
This is why I try to avoid Firewire interfacing when starting out with a new camera. Interfacing 'issues' can really confuse matters. My FLIR A40 is Firewire as well so I had better get used to debugging such interfaces if the camera is not seen by, or communicated with, by the PC host software.
Many thanks again for the comments. Much appreciated. I am not that comfortable with Firewire as you can likely tell !
Fraser -
At least you can power via Firewire and then poke around RS232. Firewire power usually also means the input voltage can be in a broad range. Firewire specifies up to 30V, but most cameras I worked with run on 12V or sometimes even 5V is fine. For example I have an adapter Y-cable from micro-Firewire (found on laptops, doesn't carry power) to Firewire-A 6 pin connector and other leg of the Y cable to USB plug to steal 5V.
I agree that Firewire can be troublesome, mostly because of multiple standards (e.g. A and B) that also can have variable link speed. Although, I never had problems with the host controllers, let it be something embedded in a laptop or various PCI or PCIe cards. I guess that's because I never used Firewire on Windows, only on Linux which is probably the equivalent of generic Windows OHCI driver. I don't know the equivalent of DC1394 library/GUI for Windows.
Sometimes camera manufacturers try to scam people by convincing them to buy Firewire PCI cards from their store for few times more than on Amazon. -
Again, useful information. Thank you
Fraser -
fascinating
waiting for the next chapter -
For interest, here are the pin-outs that I have identified so far. Some need to be confirmed in terms of exact 'type' when the camera is running as the video encoder IC DAC's are customer configured for output signal content.
6 Pin LEMO 1B 306 Pinout
1 - Firewire TPA+
2 - Firewire TPA-
3 - Firewire TPB+
4 - Firewire TPB-
5 - Power supply-
6 - Power supply+
14 Pin LEMO 1B 314 Pinout
1 - VGA Video Sync
2 - VGA Video Sync
3 - Video Signal
4 - Power-
5 - Power-
6 - Power+
7 - Power+
8 - Video Signal
9 - RS232C Data Out to Host
10 - RS232C Data in from Host
11 - Video Signal
12 - Video/RS232 0V
13 - Video/RS232 0V
14 - Video/RS232 0V
The VGA Sync outputs have been identified but which is H and which is V is not currently known. The three video outputs from the encoder DAC may be configured by the OEM to produce any of the required video signals for Composite Video, S-Video (Y&C) and VGA (RGB) These signals will be identified once the camera is up and running under software control.
The power inputs from the two sockets each have a small filter, 1.5A fuse and series diode in the path to the power PCB. Either may supply the power required by the camera as the diodes prevent back-feeds from one socket to another.
The camera specification permits a supply voltage of 9V to 24V DC.
The RS232 at the connector originate from a MAX3222 level converter IC so the camera runs true RS232C levels and polarity.
All this is may appear useless information unless I reveal the identity of the camera to you dear readers, but that may yet come later in the story
Fraser
-
No movement on the cameras re-assembly yet..... we have been out looking at kittens !
I hope to re-assemble the camera today, followed by a power on test and attempted communications via RS232.
I have purchased an iLink (Firewire) powered hub that I can use to power the camera via its Firewire port when using a laptop. That is due to arrive early this week.
As a side note that may, or may not effect this cameras use with Firewire, Microsoft ceased support for Firewire TCP IP communications with the move to Vista. To use TCP IP via Firewire you have to use a host computer running Windows XP. Good old Microsoft eh !
Fraser -
I have more information that suggests to me that these cameras were almost sacraficial in their tasking. That is to say, the project justified risking destruction of some thermal cameras. Big budgets were involved.
I'm going to guess aerospace, and that these were used to look at rocket exhaust. That type of project would certainly have the budget to treat everything around the launch pad etc. as a consumable.
I have no proof, but it is my belief that this camera was damaged during whatever testing was being carried out with it present, and it was written off as a project consumable. It failed due to 'abuse' rather than normal use. This may explain the lack of a support contract. If the camera is likely to be destroyed within the year or so of use, a support contract is not much use. Remember, the owner could afford to write off such a camera as a project consumable.
Scary though this might sound, it is not that uncommon. It is a matter of relativity. If I on my average income destroyed a $6K camera, I would be gutted, however a project with a $100 Million budget would lose no sleep over such a loss provided it could be accounted for and the end justified the means. I have seen new equipment destroyed in testing and the car industry does it all the time during new model safety and performance testing. -
Amyk,
-
Well the camera is now reassembled
time for the teardown and rebuild pictures me thinks. You can likely identify the OEM from thee pics
Enjoy
Fraser
-
Pics
-
Pics
-
Pics
-
Pics
-
Pics
-
Pics
-
Pics