Ha Ha, whilst searching out the pictures of the THV470 teardown I found that I had already written a guide to the model series and the faults that I discovered in my THV470 and a friends THV450. It is likely a first draft as I lost interest in posting it but I include it here anyway........... There will be repetition so you can skim that part
Agema Thermovision 450 & 470 Thermal imaging cameras.
Specification:
Wavelength : Medium Wave Infra Red (2um-5um)
Cooled or Uncooled: Thermoelectrically cooled
Detector/sensor : MCT 'hot mode' doped SPRITE Detector
Image forming system : Mechanical scanning raster using Lenses and Mirrors
Resolution: 140 lines x 100 sample points per line
Line rate: 4kHz
NETD: 100mK
The Thermovision 450 & 470 are self contained scanner type thermal imaging camera of the shoulder mount format. The camera uses a mechanical mirror based image creation system combined with a SPRITE cooled detector. The detector is cooled, but not by Liquid Nitrogen, Argon gas cooler or a Stirling Cooler. It uses a Peltier element 'stack' operating on the rear of the detector within a vacuum Dewar. The detector is operated at -70C so a long way away from the -196C used in a Liquid Nitrogen or Stirling Cooler based camera. This is achieved by special doping of the SPRITE detector element and careful selection of the Band-Gap.
The Thermovison 870 and its associated 800 controller form a Peltier cooled thermal imaging system that has a separate scanning head driven by a bench mounted controller. The THV450 and 470 are basically a THV870 and its controller combined into a single casing so this teardown may interest owners of the THV870 as well
.
A THV470 appeared on eBay at the right price. This particular example had clearly been stored in conditions conducive to corrosion
I was hopeful that the camera might respond to some TLC but it was very much a calculated risk.
Initial assessment of the THV470 confirmed that it was not booting and had been stored in damp conditions. The AR coating was separating from the Germanium lenses which is symptomatic of damp induced degradation. The unit appeared to be complete and cared for so had likely fallen out of use and was just placed in storage.
There is a lesson to be learned here though, thermal imaging equipment needs to be stored in appropriate conditions if serious deterioration is to be avoided.
Access to the internal parts in these cameras is made simple by the easily removed side panels. After removing the securing screws the two panels come away from the chassis to reveal the scanning head at the front and the image processing and control electronics at the rear of the camera. The Peltier cooler sits on the rear of the detector at the mid point of the camera design and has forced air cooling of its heatsink. The scanner section of the unit is basically the head assembly of a THV870 that has been cleverly grafted into a longer chassis that accommodates the image processing electronics. The electronics package comprises a mixture of fixed and removable PCB's. A card rack and backplane is used for the image processing and control PCB's whilst two large PCB's sit either side of the scanner mechanics in the scanner head. These PCB's deal with the SRITE detector related signals and the Servo drive functions.
All PCB's are easily accessed and removed and it is clear throughout how much thought has gone into the design of this camera and its stable mates. These units a mechanical-electronic hybrids that required expertise in many areas of design and production within the Swedish R&D team. These scanning cameras came from the heydays of AGEMA thermal imaging camera production and predate the BST and Microbolometer sensor arrays. The wonderful AGEMA Thermovision THV550 came after the THV4xx series and provided a very compact cooled thermal imaging camera that made older AGEMA scanner type cameras look very bulky indeed. The cooled THV550 was a palm camcorder format and used a Stirling Cooler and 320 x 240 pixel staring array operating at -196C. The THV550 effectively marked the beginning of the end for the bulky and heavy scanner type thermal imaging cameras like the THV4xx, 8xx, 9xx series. Industry bought the new compact THV550 and either traded-in the old scanner models or left them languishing in storage. Such was likely the fate of my THV470.
So what did I find inside the poorly THV470 ?
First impressions of the interior of the camera were initially good. It looked very clean. I noted some corrosion around the AA sized Lithium Thionyl Chloride memory battery. Sadly removing the other side panel immediately revealed significant leakage of the Lithium cells electrolyte onto the panel. This was an age related failure. the electrolyte has damaged the power PCB that accommodates the cell and the top area of one of the cameras image processing PCB's that sits beneath the cell in the chassis.
Not great, but by no means fatal.
All easily removable PCB's were extracted and inspected for damage. No serious damage was found. These cameras are literally stuffed with integrated circuits. Very different to modern 'black blob' SOC technology ! The firmware resides on UV erasable EPROM's. Owners of these cameras should make back-up copies of their firmware in case of future age related data corruption.
The scanner section of the camera head uses a pair of moving mirrors and several fixed precision mirrors. These are all highly polished aluminium. As those familiar with aluminium will know, it can be highly polished and works well as a mirror. Sadly it is also vulnerable to corrosion when exposed to a damp atmosphere. I was expecting the worst when I removed the dust covers that protect the mirrors. I was not 'disappointed' by the patient.... all of the mirrors were corroded beyond use
The first moving mirror creates the vertical scan movement using a tilt action and is driven by a precision galvanometer. The second moving mirror is a facetted mirror wheel that rotates at high speed to create the horizontal scan movement. This wheel is attached to a precision engineered motor. The mechanics of these cameras must have been very expensive to produce.
Once the poor condition of the mirrors was confirmed, the THV470 effectively became a spares donor as to return the camera to operation would require specialist restoration of the mirrors that placed the unit well beyond economic repair. You will recall that I bought this camera to study its design and I had low expectations regarding getting it serviceable again if faulty. I knew, however, that its parts are of value to others so I could always sell its parts to help others maintain their THV4xx series cameras. I hate to declare a thermal camera beyond help, but on this occasion it was clearly the case. It might go on to save another camera with its parts however.
The story of a friends poorly THV450
The THV470 languished in its case in my garage for a few months before a fellow forum member asked for some help with a THV450 that was booting but not producing a thermal image. After my last dealings with the series I had sworn myself off of them as they do take a lot of time to disassemble and reassemble. They are also a precision mechanical device so there is no simple way to work on the optical scanning section if that has a fault. Remember, no service information is available for these cameras and mechanical alignment of scanning optics can be a real nightmare without the correct documentation and jigs. This was a request from a friend though so I parked my negativity and agreed to look at the THV450, but only at a superficial level for anything obvious. Well that plan soon went out of the window
I got stuck in to the THV450 and quickly diagnosed a non functional scanner section but the video processor board and detector were working so there was hope for the unit. The dead scanner section could be caused by a number of faults but a hot favourite would be a failed servo drive board. My THV470 has a servo driver board that is the same as that used in THV450 so an obvious path for testing was to fit that board in the poorly 450
The replacement servo board was fitted and then the camera was retested. I was rewarded with the sweet sound of the mirror wheel motor starting so life had returned to the scanner section. The motor did not sound right though and its speed varied all over the place. The tilt mirror was not running and that was a sign that the mirror wheel PLL was not achieving rpm lock. I switched off the camera and noted the almost immediate stopping of the mirror wheel motor. That is not normal in these cameras. The pinning mirror wheel acts like a flywheel and the motor spins down slowly after power is removed. I did wonder whether the THV4xx series employed a motor brake so would need to investigate that as well. The uneven motor speed during the test was the priority. There was the possibility of faults in the motor, motor drive, servo PLL or mirror position sensing system. As an initial check I tested the mirror wheel for free movement....... it offered a weird spongy resistance not typical of a motor or bearing fault. This investigation was getting more and more interesting so I could not resist delving deeper into the scary precision mechanical scanning system. I knew it could turn into a rabbit hole but I hate to be beaten ! I tested the camera again but this time the mirror wheel did not spin-up and only a quiet hum was audible. I had either killed the good servo board or something had changed since the last test was started. I went back to the spongy feel of the mirror wheel when rotated. Maybe something was now jamming the drive to the wheel ? That was an assembly that must be precisely aligned to the other optics in the system and its looked a pig to work on. Oh well, down the rabbit hole we go !
Before attempting surgery on the patient I practiced on my THV470 scanner block to learn about how best to disassemble parts of it in a way that allowed correct alignment upon reassembly. The THV470 was a parts donor and it had already proved invaluable when working on the THV450. Investigation of the mirror drive system in the THV470 revealed age related degradation in the adhesive that was used in a shaft coupling for the mirror wheel and its associated motors anti-vibration mount assembly. Could it be that the mirror wheel coupling had failed in the THV450 causing uneven mirror speed due to slippage ? That would definitely be high on the checks list and the coupling would need testing anyway, even if still complete, as the adhesive may also be degraded. The adhesive had turned to a white powder in the THV470 but thankfully such a failure is repairable as the foam material used in the coupling and motor mount was in excellent condition with no signs of degradation.
The tilt mirror assembly is self contained and easily removed from the chassis. If there was a fault in that sub system I could easily use the motor out of my THV470 after attaching the good mirror plate fro the THV450. I established how best to approach the disassembly of the mirror wheel, its mount, the motor mount and the motor without causing damage to any mirrored surfaces. I also marked the various components positions with a scriber to realign them when reassembling the unit. I was ready to start the work on the THV450.
My first investigation of the scanner head mechanics would be the mirror wheel and its associated drive. Due to my experience with the THV470 I knew that the small hole in the motor mount allowed hex key access to the coupler securing grub screws on the motors shaft. This needs to be released before the motor is separated from the mirror drive shaft. Upon turning the mirror wheel it became clear that the motor shaft was not turning and I could not access the grub screw. It appeared that motor shaft coupling had completely failed. Upon releasing the motor from its two clamps I was able to pull it away from the mirror wheel and lift it free of the camera. As suspected, the motors shaft had only an aluminium ring attached to it. This was a third of the coupling assembly ! An inspection of the tube in which the coupling rotates revealed the other aluminium ring attached to the mirror wheel shaft and some odd looking lump trapped between the coupling ring and the housing tube wall. I was able to remove the trapped 'lump' and it turned out to be a rather mangled shaft coupling foam 'donut' ! It would appear that during the first test of the camera with the working servo board, the mirror wheel motor tired to spin the mirror but failed adhesive in the shaft coupling meant that there was slippage that prevented PLL lock. When power was removed it would appear that the foam donut in the shaft coupling flew out of its correct location and became jammed in the coupling shaft ring and hence the mirror wheel stopped almost immediately. The foam donut 'died' in the event
Fortunately I know someone with a spare coupling assembly that just needs new adhesive..... me
I inspected the motor mount foam vibration isolator and found it partially separated from its metal faces. It would also need to be rebuilt with fresh adhesive. Inspection of all other parts of the scanning assembly showed its mirrors and mechanics to be in good condition.
The repair of the motor mount and coupling components was not a simple "slap on some glue....job done" type of job. I had to select an appropriate adhesive for the task and create alignment jigs to ensure that all mounting and coupling parts were axially aligned. Failure to align these parts would lead to off axis shaft forces and vibration.
Thankfully my mechanical workshop is well equipped and I had suitable mandrels and a decent lathe to help align the parts correctly. The adhesive used for this repair was Evostick contact adhesive used in the fluid jointing mode to enable correct alignment on the jigs. This adhesive is very similar to that used by Agema when they assembled the parts at the factory. Before reassembly of the parts with adhesive, I scraped away all traces of the original adhesive from both the metal and foam parts and then finished them with 1200 grade wet &dry paper to provide a fine surface finish with keying for the adhesive. With great care, the parts were reassembled with the adhesive and left to cure for 4 days. These are load bearing parts so ensuring an excellent adhesive cure is essential.
The mirror wheel, shaft coupling, motor mount and motor were all carefully refitted into the scanning head chassis. They were very carefully aligned with all location marks that I had created before disassembly. The final part to be refitted was the optical mirror wheel position sensor. This was deliberately installed at a slight angle originally so I had to align this component exactly as I had found it to ensure correct sensing. After some fine tuning of the various components final positioning, all screws were tightened and sealed with red varnish.
The camera was then tested. The mirror wheel motor and mirror could be heard accelerating to their normal rotational speed and after a few seconds of "training" the PLL was in lock and the tilt mirror glavanometer servo started driving the tilt mirror.
The camera produced a decent image of my hand after a few adjustments in the user menu for level and span. She lived once again !
The camera was given some tests and pronounced healthy. She is running with the servo board out of my THV470 and I have yet to look at her original servo board to determine the fault with it. I was just happy to have a running THV450 in front of me to be honest ! Her owner is pleased that this camera was not beyond repair. For anyone considering buying such a camera though, do consider the fact that I have worked on such cameras previously and had the benefit of a parts donor on which to learn and source spare parts. These cameras are not an easy repair without a service manual and spare parts availability.
Fraser
