EEVblog Electronics Community Forum
Products => Thermal Imaging => Topic started by: peppy88 on July 05, 2023, 09:39:09 pm
-
Hi,
Was wondering if anyone knows the cost to service a MWIR camera's cooler. I know overtime they lose helium pressure.
What do they do to get it back to peak efficiency?
In this article it shows the MC-3 Cooler found in some of FLIRs current lineup:
https://www.flir.com/discover/instruments/gas-detection/see-smaller-voc-leaksfrom-a-safer-distance/ (https://www.flir.com/discover/instruments/gas-detection/see-smaller-voc-leaksfrom-a-safer-distance/)
-
When a Stirling cooler reaches the end of its service life there are two options open to the owner in terms of ‘repair’.
1. Have the cooler recharged with UHP Helium, but no other work carried out (not normally offered by service centres)
2. Have the cooler rebuilt/replaced and filled with Helium. A full camera calibration is usually carried out after this procedure and such is essential if the original cooler is changed and not rebuilt, due to differences in the attached sensor array.
The last time I had a Stirling cooler rebuilt (not replaced) it was £5K and that was back in the mid 1990’s. I saw a quote from FLIR to rebuild (not replace) the cooler in a PM280 in the early 2000’s and they quoted £7K for the job. If the cooler required replacement rather than a rebuild, it was going going to cost £15K.
It is often uneconomic to have a Stirling cooler rebuilt or replaced on obsolete thermal cameras and FLIR may only offer such a repair on some models of camera for which there is still Stirling cooler support from the supplier.
Note that I say repair or replace when it comes to coolers. The coolers use extremely fine tolerances to achieve their low ‘cold finger’ temperature and wear in the moving parts results in inefficiency. Some coolers may be rebuilt with certain wear prone parts replaced but others have to be replaced with a new cooler and this will have a new sensor array already installed. New coolers are very expensive as a result. The Helium used in these coolers is UHP type, meaning Ultra High Purity, as contaminants such as air or moisture will cause the coolers pistons to lock up in use due to freezing. Particulate contamination just destroys the moving surfaces.
I wrote a post on the perils of buying cooled cameras for this forum a while ago. I will find it and provide the link.
Hope this helps
Fraser
-
Here you go, the post I created on the perils of buying cooled thermal cameras. I own a few and the ones with failed coolers will never be repaired.
https://www.eevblog.com/forum/thermal-imaging/thermal-camera-purchasing-the-perils-of-buying-a-cooled-camera/ (https://www.eevblog.com/forum/thermal-imaging/thermal-camera-purchasing-the-perils-of-buying-a-cooled-camera/)
-
Gotcha. Do you know what they do during the rebuild? Also what constitutes a rebuild vs replacement?
Im guessing they do a rebuild for a cooler that still works but just has lost efficiency over the years? And a replacement for a completely non functional one?
Just want to know what I am getting into if I ever get my hands on an A-Series or X-Series Camera.
-
The parts that wear on a Stirling cooler depends upon the type of cooler.
1. The Linear Cooler (induction driven oscillating piston) has few parts that move, is very long lived, and reliable. The main areas of wear are the cylinder liners which are normally coated with a special material that reduces Helium gas penetration into the cylinder walls. A rebuild will assess the wear on the cylinder walls and piston to determine whether they require replacement/refurbishment. Any seals in the system will also be replaced to avoid Helium loss through age and vibration related seal failure. The seals are usually made from a very soft metal called Indium. The springs in a linear cooler normally have a long life but as they are part of a mechanically resonant system comprising the spring, piston and AC excitation magnetic field, they will be tested for performance and new springs fitted if required.
2. The Rotary Cooler looks like a model aeroplane Nitro engine but it’s internal parts are very different. The ‘L’ shaped housing contains two pistons, a crankshaft and a flexure (similar purpose to a connecting rod in an internal combustion engine). The crank is rotated by an electric motor and the pistons are driven in their cylinders. There are bearings involved in the Rotary cooler design and it has more points of failure than a linear cooler. Common points of failure beyond wear to the cylinder walls are piston wear due to lateral forces, bearing faults and flexure failure. When these coolers fail they can become very noisy, just like when the connecting rod bearing shells fail on a car engine. Material ejected from damaged bearings contaminates the cooler and it may require extensive rework in order to return it to a serviceable state. Dead rotary coolers can be expensive to repair. As with the linear cooler, the cylinders are often sealed against Helium leakage with a special impermeable layer. This will be subject to wear and may require rework. The life of a rotary cooler used to be only around 2000 hours but design improvements have extended expected life to 10,000 hours in some cases. The Rotary cooler will have special gas seals just as in the Linear Cooler. These will be replaced as part of the service.
Rebuilding a Stirling cooler requires clean room conditions, skilled staff, specialist reworking capabilities and it is a time consuming task. It is a relatively small market, specialist, service and so attracts a significant premium.
You have to go into cooled thermal camera ownership with your eyes wide open and, in some cases, in the knowledge that your cooled camera is running on borrowed time, could fail at any time and is likely uneconomic to repair. Hence my warning post referenced previously. If you are willing to lose the money spent on a cooled thermal camera, then life is good and you can enjoy ownership of such, as I do. If you are concerned about longevity and the availability/cost of repair, then a Stirling cooled thermal camera may not be a good choice for a stress free life :-DD
Fraser
-
Good to know haha. I knew these cameras were expensive and expensive to maintain. I just wanted to know "how" expensive.
Here is an x6900 that is for sale at the moment:
https://thermalvisionresearch.co.uk/demo-stock-sale/
I wonder what type of cooler that uses? I know FLIR Sells Neutrino cores that use the linear cooler you are talking about. They are prob the most optimized coolers to date.
-
Yep, they are not exactly cheap, even ex Demo !
My SC4000 high speed SWIR+MWIR cooled camera cost its original owner $150,000 :scared:
From memory the X6900 uses a linear cooler, just as in my SC4000. Some LWIR versions of the SC4xxx/SC6xxx series used rotary coolers. Linear coolers are the cooler of choice these days.
Fraser
-
Oooops, I am incorrect. The X6900sc uses a Rotary Stirling Cooler !
http://www.flirmedia.com/MMC/THG/Brochures/RND_065/RND_065_US.pdf (http://www.flirmedia.com/MMC/THG/Brochures/RND_065/RND_065_US.pdf)
Fraser
-
Yeah I am assuming that even though these are Rotary Coolers, they probably aren't the rotary coolers of the past. I am guessing they can run for a few thousand hours before need for servicing.
Also they never really list the prices of these cameras. So I was always wondering how much. Funny enough I stumbled upon this webinar in German:
https://youtu.be/6vQ-AsY59OM?t=1222
They show the aprox pricing. Good webinar btw just had to watch in subtitles lol
-
Hobby owners of these systems may not need days and weeks of cooling operation and simply purchasing LN2 and piping it in may be a satisfactory alternative. You can buy a Dewar and a LOT of LN2 for 7000 pounds. It will take some cleverness to set it up, but it shouldn't be expensive.
A cryostat is another possible alternative but can be much more difficult to sustain.
-
Some previous stories from me involving buying used cooled thermal cameras……..
https://www.eevblog.com/forum/thermal-imaging/high-speed-thermal-imaging-cameras-when-60fps-is-just-not-enough-!/ (https://www.eevblog.com/forum/thermal-imaging/high-speed-thermal-imaging-cameras-when-60fps-is-just-not-enough-!)
https://www.eevblog.com/forum/thermal-imaging/the-story-of-a-radiance-1-camera-and-frasers-quest-to-find-information-on-it/ (https://www.eevblog.com/forum/thermal-imaging/the-story-of-a-radiance-1-camera-and-frasers-quest-to-find-information-on-it/)
https://www.eevblog.com/forum/thermal-imaging/a-stirling-cooled-qvga-thermal-camera-for-99-pence-!-yep-no-kidding/ (https://www.eevblog.com/forum/thermal-imaging/a-stirling-cooled-qvga-thermal-camera-for-99-pence-!-yep-no-kidding)
There were bargains to be had but with the increase in publicity surrounding thermal imaging, such bargains are rare now.
-
I contacted a buddy that has a GF series camera
Here is a quote he got recently:
"I have confirmed with a technician that services this model camera.
If the cooler needs to be replaced the cost will be $8,499 + $3,199 for the calibration.
If the cooler replacement fails it would require an IDCA replacement. This would estimate around $32,000."
Good to know lol...
Also what would constitute a IDCA replacement?
-
That quote makes no sense to me.
An IDCA is an Integrated Dewar Cooler Assembly …. That is the Stirling Cooler in my language, yet the quote states that the cooler replacement is $8499 + calibration. More information needed to better understand the quote. Maybe FLIR consider the IDCA to be the Stirling Cooler plus other PCB’s. A bit like buying a short block engine vs a whole engine ?
FLIR spare parts for professional thermal cameras are very expensive. As someone once told me…… if you buy a Mercedes, you pay Mercedes prices for parts and service.
-
Yeah not entirely sure. Maybe they mean the detector array as part of the assembly?
Also the evaluation fee for looking at the camera is $999, but is waived if you perform the maintenance/calibration.
Yeah that is true haha. But I guess it is nice to know what it actually cost to maintain in the even you happen to stumble upon a used model
-
If buying a used model, please do be aware that many of these high performance cooled cameras have a lot of controls placed on their sale and use. This includes requirements for export licences and paperwork. If a person finds one of these cameras and lives outside the USA, it can be challenging to get FLIR to support the camera or even talk about it. My SC4000 is a direct product of a US DoD contract with FLIR to create a high speed thermal camera for missile range use. As a result, the documentation and support for my model is heavily controlled. If you are not the original purchaser of the camera, FLIR sometimes refuse to provide any help. I note that some of the later versions of these cameras have less controls on their documentation at least. This may be because of the direct connection of my model with the DoD contract and very detailed manuals that come with it.
Expect to find an SC4000 or SC6000 for between $5K and $15K on the used market. Later models will likely be more expensive. Sadly you cannot see the “hours run” on the camera itself. This information is accessed via the camera configuration utility loaded on a PC. My camera is circa 2008 and the cooler is in great condition thanks to very low hours on it. Such may not be the case with other cameras used in industry.
If buying one of these science cameras, make sure that you can obtain the configuration utility and image analysis software. Nothing about these cameras is cheap !
Fraser
-
On the IDCA: I've come across a lot of companies sales reps using that when they actually mean IDDCA (integrated detector Dewar cooler assembly). The IDDCA is the camera 'core' consisting of the cooler, cooler controller electronics, Dewar, detector (FPA w/ROIC), and close proximity electronics for the detector. Additional electronics cards are not include: 1) daughter cards to convert into CameraLink, SDI, RS-170, ethernet, or similar video; 2) power distribution; 3) any shutter for NUCing; 4) filter wheels;
Good way to tell is by the price. To me the 32K is a dead give away.
Not sure why there is a 'calibration' needed if the cooler is replaced. Normally the compressor (linear or rotary) and cold finger would be replaced. The system would need to have the Dewar pumped back down after welding to the new cold finger. The FPA and ROIC are not replaced. Unless they mean rebuilding factory NUC tables and/or if the sensor is factory radiometrically calibrated. I've found factory calibrations good SWAG (scientific wild a$$ guesses) but not really useful for detailed radiometric or thermographic measurements. On site external high quality blackbodies are a must.
-
IR_Geek,
Thank you for the excellent insight into Cooler replacement. When I had cooled cameras repaired it was often unclear whether the cooler was actually rebuilt or a new one fitted. The requirement for a calibration seems to be a FLIR standard operating procedure, along with installing the latest firmware version, as they include such in many repair operations. This may be done in order to provide the customer with an assurance that the camera has been returned to factory fresh operational condition. I thought that FLIR included a new detector array with all its cooler replacements rather than reusing the original detector but this was never made clear in the repair notes I received.
Fraser
-
Any time Fraser. I'm not on here much, but feel free to reach out anytime. Quite enjoy 'geeking' out out on all things IR and especially the history.
Unless the FPA/ROIC is bad or it fails a spec threshold for number of bad pixels, it is rarely replaced. Too expensive. It may not seem like it but prices for cooled sensors have dropped dramatically. What makes it seem like prices are going up is the vast increase in array size, reduction of pixel sizes, increase in digital bit depth, increase in frame rate, and reduction in sensor noise. I'd say the price per pixel has been very competitive for what a person (really a company !) can afford.
-
Any time Fraser. I'm not on here much, but feel free to reach out anytime. Quite enjoy 'geeking' out out on all things IR and especially the history.
Unless the FPA/ROIC is bad or it fails a spec threshold for number of bad pixels, it is rarely replaced. Too expensive. It may not seem like it but prices for cooled sensors have dropped dramatically. What makes it seem like prices are going up is the vast increase in array size, reduction of pixel sizes, increase in digital bit depth, increase in frame rate, and reduction in sensor noise. I'd say the price per pixel has been very competitive for what a person (really a company !) can afford.
Thanks for the insight. That's good to hear that it is rare the FPA needs to be replaced. The price to replace the cooler($8k) is a little more reasonable than $32k for the entire IDDCA.
I'm guessing if you are still able to take good images with the camera and it just takes longer cool it just requires a cooler rebuild in most cases?
Here is a FLIR MC-3 cooler.
[attach=1]
They still use this in the newer GFx320 cameras today. And if I am not mistaken in some of their A series cameras that use the "rotary" style cooler.
Here is an article of some testing they done with it:
https://cryocooler.org/resources/Documents/C16/077.pdf
-
Similar to what Fraser said in the 4th post. Lots of moving mechanical parts with either rotary or linear. Also you have to imagine that some of those moving parts are at cryogenic temperatures. Taking longer to cool down then its design/spec is an indicator of a problem. Determining the root problem really takes an expert and somebody familiar with that specific cooler/camera. Could be a poor vacuum on the Dewar, gas leaking, contaminates in the gas, seals wearing, or heck even 'bad' power causing poor motor driving. Lots of consideration into how many hours it ran, what kind of vibrations it was exposed to, and the big one of running during really hot (or cold) conditions.
I'd love to get my hands on a couple of older cooled cameras. Unless I could score them really cheap, the cost of fixing them (if even possible) is not worth the hassle.
-
I've got all the equipment necessary to refill a cryocooler(ultrapure gas handling, ultrahigh vacuum), but given that all that equipment is currently in Belarus it's not particularly useful to most people on this forum.
It's not very difficult to assemble this kind of kit - it's just regular high vacuum and UHP gas stuff. Conflat everything, swagelok for tubing and valves, etc. The most expensive single part will likely be(assuming you buy everything else used) the 6N or 6.5N pure helium tank, which is pretty much always a custom order item anywhere in the world, and it typically involves buying the gas bottle outright with the gas inside, because they're shipped completely sealed(both in the "closed" and in the "tamper" sense) from the specialty gas plant. A 10L cylinder of 6-6.5 helium costs around $600, and while that might seem expensive, there's enough gas inside to completely service hundreds to thousands of cameras.
You definitely don't need a cleanroom, but a laminar flow HEPA downdraft hood is a must - thankfully bio labs have the same level of contamination cleanliness requirement for growing sensitive cultures, so they're quite affordable, and worst case not very hard to build yourself out of a few large sheets of stainless and a heavy multi-filter stack.
If, however, the cooler itself has failed(i.e. detached debris ended marring up a cylinder), then best case scenario you need a precision(Ra/Rz<0.1um) grinding/polishing capable machine shop, and worst case the entire mechanical unit is scrap. You could buy another donor cooler and replace everything but the cold finger(which is part of the IDDCA, where the sensor is bonded to), but I don't know if anyone would want to take that risk, since that requires voiding the cooler's warranty and venting it prior to frankenstein reassembly.
-
I have been considering disassembling the cooler in one of my failed PM280 cameras. That particular cooler is an interesting case as, if the camera is placed in the fridge, it will achieve operating temperature at the cold finger but cannot maintain it for long. The cooler has a mechanical fault that means it is both VERY noisy and creates masses of vibration in the camera. To run the camera I have to wrap it in a towel to muffle the horrendous noise that it makes. When the thermal image appears, there are hundreds of dead pixels present and grouped together in such a way that the detector FPA would have failed QC at time of build. I suspect that vibration from the cooler is responsible for the masses of failed pixels, rather than loss of the dead pixel map data. There is no hope for that cooler, it is effectively End of Life and requires either a very expensive rebuild or a new cooler and detector FPA.
I have not dismantled the cooler to date as I saw little benefit in doing so. I know what is inside a rotary Stirling Cooler and the disassembly would be purely to take pictures fir the forum and discover which bearing had failed to cause the noise and vibration. I may revisit that camera and extract the cooler fir a teardown but I do not think it will make for a particularly interesting or useful photo set as these are not normally a DIY rebuild device. Remember that the cylinder liners are treated/coated to help Helium ingress into the aluminium and no spare parts are offered fir DIY rework.
I have a very nice, fully working, low hours, Rotary Stirling cooler from an AGEMA THV900 but it would be wrong to destroy that cooler despite me not having a practical use for it. Maybe I will mount it on a nice stand as a desk ornament :)
The attached pictures are of the Inframetrics PM280 Rotary Stirling Cooler with detector array and lens block attached.
Fraser
-
Wow that looks really similar to the FLIR MC-3 (see pic in previous post). Same cooler used in many of their cameras today
-
I hope that they have improved it then as the Inframetrics PM2xx/3xx series cameras are infamous for suffering cooler failure on a pretty regular basis ! Not to say they were premature failures, just that they were not the longest lasting of rotary Stirling coolers. There are not many of those cameras still running on their original cooler. FLIR bought Inframetrics and renamed the Inframetrics PM2xx/3xx series the SC1000 series. Inframetrics owned technology was absorbed into the FLIR product development work flow.
Fraser
-
I think they have improved it. Here is the manual to a G300a which is the fixed mount version of the GF320 and GFx320.
https://www.wingold-mt.de/wp-content/uploads/2016/07/FLIR-G300a-Benutzerhandbuch-ENG.pdf (https://www.wingold-mt.de/wp-content/uploads/2016/07/FLIR-G300a-Benutzerhandbuch-ENG.pdf)
Towards the end the MTBF: is 2 years or 15000 hours running 24/7 !
I'd say that is way better than the stats for the older rotary coolers haha.
Most of the time we don't run the cameras 24/7 so it's good to know that there is decent reliability to the MC-3 microcooler
-
Found this gem of a video (old but good): https://www.youtube.com/watch?v=-H3TwSux4Ic (https://www.youtube.com/watch?v=-H3TwSux4Ic)
Looks like the production of their MWIR cameras. I see a few SC5000 and SC7000 cameras lol. Look at all the IDDCA's too lol
-
Peppy88
Thank you for sharing this excellent video. I loved watching it :-+
It was the team in France who helped me with my SC4000 reds. The FLIR USA support team pearled less familiar with cooled camera technology and I was told Paris was their specialist centre for all matters relating to cooled camera technology. The Chief science camera engineer could not have been more helpful to me and is a really friendly chap :-+
The cooled cameras are built with love and professionalism :)
Fraser
-
Yes, formerly CEDIP (see 1:35 and 2:04 - 2:08 frame), not fully rebranded to FLIR.
They were specialists in building cameras for ULIS / Sofradir sensors. Which explains why FLIR USA would not know much about them.
-
Nice video on rotary and linear coolers. These videos are hard to come by:
https://www.youtube.com/watch?v=P2epxhmQuoY (https://www.youtube.com/watch?v=P2epxhmQuoY)
-
What an interesting article. These guys are trying to make an "affordable" cryocooler that cools down to 150K. Which is what is considered Hot MWIR today.
https://cryocooler.org/resources/Documents/C21/191.pdf