EEVblog Electronics Community Forum
Products => Thermal Imaging => Topic started by: Fraser on July 14, 2021, 02:35:04 pm
-
A little story by way of an introduction to this thread.......
A couple of weeks ago I was contacted by forum member "NukeET" who was asking for some assistance with a Amber Radiance 1 camera and DFOV lens. With regret I was a little 'cool' in my response as I get a lot of requests for help and it can be a little tiring.
I did try to help NukeET however as he is a nice chap who clearly appreciated others help. We discussed the condition of the Radiance 1 and it transpired that the PCMCIA card that holds the firmware is missing. In addition to this the Lithium memory backup battery had been disconnected during investigations prior to contacting me. The PCMCIA card is a Linear Flash Type 1 that is both hard to find and expensive. I offered to program a card with my cameras firmware if NukeET could find a blank one. There was no guarantee that the copy of my firmware would work in his camera and we still had the issue of the lost battery backed calibration memory contents to deal with. This is where life took an interesting turn for me. NukeET decided that he would rather I have the camera and lens than them be stripped for parts, as repair seemed to require a lot of effort with no guarantee of success. We knew the cooler was still good, but that was all we knew about the camera.
NukeET sent me the Radiance 1 camera and DIOP 75/250mm DFOV motorized lens at no cost to me except postage ! This chap is so kind and generous ... a true Gentleman. The camera and lens arrived safely today and I have taken some 'as received' pictures. As can be seen, both units are in excellent condition :-+
My sincerest thanks to NukeET for giving this camera and lens to me. He would like to see how the lens performs and I will document that on this forum when I have established how to drive it.
The Radiance 1 camera may, or may not be recoverable but even if beyond repair, it is excellent insurance in case my working Radiance 1 camera develops a fault. It will also permit me to study the Radiance 1 camera design without risk to a working unit.
We have already met the Amber Radiance 1 camera in this thread........
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/)
The DIOP DFOV lens is new to me and my collection however :-+
This is a lens that was designed for use with the Amber Raytheon Radiance 1 and Radiance HS series of cooled thermal imaging cameras. It is a powerful dual field of view thermal camera lens and not a continuous zoom type. The dual field of view is achieved by the use of a set of lenses that may be moved into the optical path when required. This is done by a servo sytem. This type of 'Zoom'/'Telephoto' function was/is common in military and long range surveillance thermal camera systems. This lens is the Dual FOV 75/250mm model but a triple FOV (60/180/500mm f4) version was also made. The lens is f2.3 :-+ I dread to think what they cost when new back in 1996
The lens also contains a filter wheel that can hold several filters for either attenuation or band filters for specialist applications. This particular lens is fitted with two, currently unidentified, filters. The filter wheel is motorized and controlled by the on-board menu driven controller. This controller is also used to set the FOV and may control Focus as well (TBC).
An 8 pole Fischer 104 series socket is used to connect the lens to the host camera. This connector provides the 5V supply to the lens plus simple focus control and feedback signals. Sadly a 8 pole Fischer 104 plug is eye wateringly expensive so I may have to change this connector to a 10 pole Lemo connector that I have plenty of.
For anyone wondering what the equivalent lens would be in a 35mm SLR format.... I think it is about 800mm :)
Time for the pictures then.......
-
Lens detail continued.....
-
DIOP DFOV Lens specifications when fitted to a Radiance 1 camera :
Optics : Motorized DFOV 75/250mm
f#: f2.3
FOV: DFOV 5.9/1.8 Degrees
IFOV: DFOV 0.4/0.12 mrad
Band: 3-5um MWIR
Focus: Manual Motorized
Filters: Yes. Motorized rotary Filter wheel
Control: From host camera and via built in LCD Menu and buttons
-
The DIOP lens is also compatible with my FLIR SC4000 as the Amber Radiance 1 is an ancestor of the SC4000 and they share the same lens requirements, including the same lens mount :-+ My SC4000 combined with the fast DIOP DFOV lens would make a powerful MWIR long range thermal imaging system.
-
Hello, I also own one of these lenses, the original cost of this lens was about $35k when new. I would imagine that there were fewer than 100 made. If I remember correctly, the lens needs + and - 15 volts to operate. 5 volts is generated internally by a dc to dc convertor inside the lens. Once supplied with power the lens can operate independently of the camera. One of the functions of the lens controller is to find and remember the focus points for the two fields of view, this lets the image stay in focus when changing fov. The manufacturer DIOP (Diversified Optical) was acquired years ago by Axyss who has since been acquired by another company(I don't remember who). Axyss made a range of long focal length surveillance cameras. DIOP also made a 500mm F2.3 lens! I had a chance to use this lens with a Radiance camera. The closest you could focus with this lens was about 100 feet.
-
Olivir,
Thank you so much for this information. I am working in the dark with the DIOP lens as it is hard to find anything about it on the internet. I see that Indigo sold a similar specification of lens for their Merlin and possibly Phoenix (?) cameras. I was hoping that there might be more detail of the lens under the Indigo branding. From what you have said that will likely not be the case though. It is a lovely piece of optical engineering and I am pleased that it can work independently of the camera for focus and FOV setting :-+ My lens and camera came without the little umbilical cable that connects the two so I may make a replacement using LEMO connectors to replace the Fischer sockets that are fitted. Interestingly, Amber used LEMO sockets on the rear of their camera and a Fischer socket for lens control, on the front :-//
I am grateful for any and all information that you and others provide as it helps me better understand the ‘patient’
Thanks again
Fraser
-
A peek inside the DIOP lens to establish the pinout of the Fischer connector .....
The unit contains a microcontroller board and the components are marked 1994.
As has already been stated by "Olivir", the unit requires a +15V and -15V supply. The +5V supply rail is produced by a 78SR105HC 1.5A switching regulator from the +15V rail. Helpfully the Input/Output PCB mounted on the rear of the Fischer connector is cleary labelled with the voltages etc.
The only other connectivity to the Fischer connector is an RS232 port for communications with the microcontroller. I had read in a reaserch paper that this DIOP lens could be remote controlled over RS232 but I was unsure whether that was actually through the Amber Radiance 1's Serial port and lens interface.
The wiring loom within the lens casing is not as tidy as I had expected of DIOP but this lens may have been worked on previously and any loom management removed. I will tidy up the loom once I have finished work on the lens.
The actual lens barrel assembly is as expected with the rotating lens holder is mounted towards the front on a pivot and driven by a worm and sector gear. The focus assembly is a conventional design and driven by a linking gear wheel set, gearbox and motor combination.
The Microcontroller PCB monitors end-stop micro switches and uses two contact thermistors to monitor the temperature of the lens assembly. It uses an Intel 8097BH processor. I was pleased to see that the Varta 1/2AA 3V Lithium memory backup cell still reads 3.135V and its date code is 0694 !
The filter wheel is of conventional design and driven by a motor+gearbox combination under the control of the microcontroller PCB. The filter wheel is indexed so the controller knows the filter positions.
It is an interesting lens and very similar in mechanical design to FLIR DFOV designs that I have worked on.
-
I will be downloading the firmware from the DIOP lens controller to see whether I can determine the instruction set used for RS232 control. The lens provides 'local control' through the four push buttons but a handheld RS232 remote controller might be useful.
-
For anyone wanting to know how these DFOV type lenses work, take a look at the attached pictures taken from
https://www.prc68.com/I/ThermalIMagerDFOV.html (https://www.prc68.com/I/ThermalIMagerDFOV.html)
The military thermal imager shown on that page uses the switched additional lens set as will be found in my DIOP lens. The additional lenses are on a chassis that swings them out of the optical path when not required. A servo system or solenoid brings the additional lenses into the optical path when requested by the user. This may appear simple, but good engineering is needed to maintain correct lens axial alignment and stability.
In the images the front objective lens has been removed to show the DFOV mechanism action.
Fraser
-
You lucky man!
The two units I can lay my hands on are both software V1.02N, Options 06 00 55 02, Menu Options 089B ... this shows up on quickly on the screen when it boots up. Using HyperTerminal COM's are RS232 with baud rate of 9600, data bits =8, parity = none, stop bit = 1, and flow control = none. under the ASCII setup... send line ends with line feed, echo typed character locally, append line feeds to incoming line ends, wrap lines that exceed terminal width.
Some commands I've dug up:
:FWR ... query filter wheel position
:FWWy ... set filter wheel postion with 'y' being 1, 2, 3, 4, 5
:FV- ... set to NFOV
:FV+ ... set to WFOV
:FF- ... decrease fine focus
:FF+ ... increase fine focus
:FR- ... decrease rough focus
:FR+ ... increase rough focus
:SVR ... read software version
:SNR ... read serial number
:MDR ... read model number
Hope that helps you. Love to see if you can dig up more commands as most of mine were 'brute force', conversations with some former DIOP folks, and picking apart some Amber manuals. I know there has to be a bunch more based on the control screen and abilities called out in the Amber Galileo manual.
-
IR_Geek,
Thank you for this very helpful information. I will share any additional commands that I discover.
I was wondering whether the Galileo/HS manual contained information on the DFOV lens. My Radiance 1 manual contains nothing on this lens and the Pinout of the cameras lens connector shows only a +5V rail so I suspect the manual predates the release of the DIOP 75/250mm DFOV lens option.
Please can you tell me whether your lenses use a separate +/-15V power supply or do they take their supply from the camera system ? If it is an external power supply, do you have the current rating for those supply rails please ? I can hook up my dual rail lab power supply to check the current draw but it would be nice to know if there is an official PSU rating.
Thank you again :-+
Fraser
-
some more history: after AxSys, the name was picked up by General Dynamics. By that time, most of the original folks had gone on to other companies ...
The units I have use a separate power supply but believe it could be powered from Galileo. I know it could be controlled via the Galileo. As for power, the notes I have are +12V on pin 3, -12V on pin 6, and grounds on 4 and 5. I'll have to verify the power draw.
According to the old website, the lens control board was called "Smart Control System" and was in a lot of their lenses.
-
My continuity checks showed the following pinout using the silk screened connector pin numbering on the PCB.....
Pin 1 RS232 Receive
Pin 2 RS232 0V
Pin 3 -15V
Pin 4 0V
Pin 5 0V
Pin 6 +15V
Pin 7 ?
Pin 8 RS232 Transmit
I will recheck my work as I have the supply polarity inverse to your notes. Thankfully both rails have series diodes protecting them though :-+ I was hoping a +/- 12V might work with the lens as they are likely just motor drive voltages with the 5V regulator having more than enough headroom at +12V input :)
The power and I/O pcb will be removed and reverse engineered and I will likely fit a LEMO 3B 310 chassis socket in place of the Fischer 104 series fitment. That is an easily reversed modification if I ever need to revert the lens to 'factory spec' :)
Thank you for the information on the controller board. Very interesting :-+
-
Pin 7 was a no connect in my notes. I'll double check the +/- ... it's likely just 5V min with a 15V max. Thinking back on it, I distinctly remember the short cable between Galileo camera and lens. There was no other cable needed if you controlled it via the camera.
[attach=1]
From the website and image below: Looks like the DFOV when connect to the Radiance1 is connected to the viewfinder connector ??? Or do I have those connectors mixed up? Came across something that said the viewfinder was 15V
https://www.researchgate.net/figure/3-5-micron-band-12-bit-infrared-video-camera-Radiance-1-from-Amber-Engineering-with_fig4_224002649 (https://www.researchgate.net/figure/3-5-micron-band-12-bit-infrared-video-camera-Radiance-1-from-Amber-Engineering-with_fig4_224002649)
[attach=2]
-
Wow Fraser, you're a very lucky chap (but, as I well know, your 'luck' is largely self-made, via a lot of on- and off-line work - and well-deserved).
And a big, big shout-out to NukeET for their generosity and selflessness. Bravo: the world needs more people like you. :D
-
Couldn't verify pins on +/- but here is brand and part number along with listed current for each.
TRUMPower … TMP60-T31 … PMP60-31
+12V … 3.0A /// -12V … 0.7A ...
XP Power … PCM50UD07
+12V … 3.0A /// -12V … 1.0A
-----
Oh, yeah. I did mix up the viewfinder and lens connector.
-
IR_Geek,
Thank you. That is exactly what I need. I was unsure how much current the motors draw but it looks like less than 0.7A which is good to know. I shall start a search for suitable dual output power supplies in the UK.
Thank you for taking the time to detail the power supply specifications. Much appreciated :-+
Fraser
-
Not too long ago you were lamenting some not so courteous treatment by someone. Now this. It seems that the bad must be taken to get the good, and the good is very good. Congratulations and a big shout out to NukeET for his kind gift to one of the really good guys on this forum.
-
DIOP DFOV Lens specifications when fitted to a Radiance 1 camera :
FOV: DFOV 5.9/1.8 Degrees
IFOV: DFOV 0.4/0.12 mrad
Band: 3-5um MWIR
Now a lens giving 0.12 mrad costs $ 30,000
https://www.flir-direct.com/product/flir-t199745-ir-5-6-lens (https://www.flir-direct.com/product/flir-t199745-ir-5-6-lens)
https://apliter.com/wp-content/uploads/2020/05/FLIR-IR-Lente-f142mm-7%C2%BA-Serie-T10xx-Ficha-t%C3%A9cnica.pdf (https://apliter.com/wp-content/uploads/2020/05/FLIR-IR-Lente-f142mm-7%C2%BA-Serie-T10xx-Ficha-t%C3%A9cnica.pdf)
-
Fraser - I expect to see some really good pictures of the Moon from you soon!
-
There is a difference between a 142mm f/1.2 LWIR and a 250mm f/2.3 MWIR lens. In fact, you can calculate the difference in aperture alone and end up surprised - but it's also vastly different material. A modern flir lens will be very well calibrated and corrected. Which is a huge chunk of the money you pay.
I have a 150mm f/1 lens at home that is current in multiple parts as I wanted to turn it from motorized back to manual focus, but it's optically a very simple design - just big and heavy. And it cost me nothing close to 35k and won't even sell for anything close to it either.
-
Ultrapurple,
Today the DIOP 75/250 lens ……. Tomorrow …….. ? ……. ;D
-
Vipitus,
Good point. Cooled MWIR cameras permit ‘slower’ lenses to be used thanks to their high thermal sensitivity. f2.3 is actually fast for a cooled camera lens ;D Standard Cooled camera objectives are often a similar f#
Uncooled cameras and small pixels means a similar IFOV lens needs a much lower f# and this is one of the disadvantages of uncooled cameras…. They often need larger, more expensive, chunks of Germanium to achieve similar results to cooled cameras.
Regarding the DIOP 75/250 DFOV lens…… do not be so quick to judge its performance inferior to a modern thermal imaging lens. These sorts of lenses were often a ‘no expense spared’ product that was an off-shoot of military lens production. The Germanium lens elements will have cost a small fortune to make and are likely of the highest quality and best coatings. Modern thermal imaging lenses are built for maximum profit and will still have excellent performance, but unless military grade optics are employed, production cost reduction will find its way into the optical path. ;)
I take your point about ‘simple’ optics…. Many thermal cameras use very basic optical designs that would not be considered very good in visible light imaging applications, but the apparent simplicity can sometimes hide clever design features that ensure high quality thermal imaging. The DIOP 75/250 DFOV lens uses Germanium and Silicon lens materials to provide high performance. Then there is the dual field of view design. Far simpler in optical terms than a true variable Zoom lens design, but it remains a popular system in long range observation systems. Less Germanium tends to mean a faster lens so complex multi element variable Zoom lenses can be relatively slow, or very large and heavy. In many applications the observer desires a wide FOV for target acquisition and a nice narrow FOV for target detail and identity. Variable Zoom offers better framing capability but that is more of a photography and videography requirement and tends to fall outside professional industrial or Military thermal imaging camera usage. With Radiometric thermal imaging, a DFOV or TFOV lens is far easier to calibrate with the camera.
Thermal imaging lenses have always been expensive and I do not know the profit margins on these very expensive examples. It would be interesting to know the cost of making a modern high performance thermal imaging lens. Chalcogenide IR glass has reduced the cost of budget lenses but I would not want to see such a compromise material in an expensive lens…. It is good, but not that good.
-
Thermal imaging lenses have always been expensive and I do not know the profit margins on these very expensive examples. It would be interesting to know the cost of making a modern high performance thermal imaging lens. Chalcogenide IR glass has reduced the cost of budget lenses but I would not want to see such a compromise material in an expensive lens…. It is good, but not that good.
A few factors to consider:
The design cost is in there somewhere, either up front or spread over the first couple of batches of lenses. So is manufacturing tooling.
Both will be a factor of quantities expected. Equally there is a cost saving if the design is a 'close relative' of one that already exists.
Pure germanium cost is also far from trivial when talking 'big' stuff.
Some surprising factors:
Modern small pitch sensors are much less optically forgiving, so are likely to demand some aspheric elements which are always made one at a time. Spherical elements can be made in sets on much simpler machinery #
Similarly long focal length lenses with a larger back focus are a lot simpler, with less elements and less likely to be anything other than spherical surfaces.
# except one supplier, who ONLY had diamond turning machines so wanted to make all the lens surfaces as either flats or aspheres !
From the Argus point of view, attaching image of 'little and large'.
6.3mm f/1 from Argus4 (for either a 160x120 @ 35µm or 320 x 240 @ 17µm)
100mm f/0/7 for a tube camera (in pixel terms about 200 diameter at 100µm)
The 100mm was about £6k in 1990 and around the same price as the camera. It is time to stop if the lens is more than the camera body.
Bill
-
I want to double down on Bill's comments. Infrared lenses are a very small market and suffer from that at every place in their design, except possibly mounts. Raw materials and polishing materials and methods are two more things that are early on the learning curve industry wide. Even when using materials that are potentially compatible with mass production such as molding, the processes are in their infancy. The high indexes of many materials put them in sparsely explored territory for optical design. Every where you turn the lens maker is exploring new territory.
-
Raw materials and polishing materials and methods are two more things that are early on the learning curve industry wide. Even when using materials that are potentially compatible with mass production such as molding, the processes are in their infancy. The high indexes of many materials put them in sparsely explored territory for optical design. Every where you turn the lens maker is exploring new territory.
I would say it depends who and where you are looking. There are enough well experienced suppliers of pure Ge lenses about who have someone who knows how to run the design software, while spherical polishing goes back to Galilieo !.
Moulded chalcogenides are fairly new but again several companies have been at it long enough to get quite competent and to get the best out of the lens design options.
Equally there are a few 'chancers' about who do not know how to run the software (beyond school level optics anyway), do not understand tolerances or centering and think a 10.6um laser is a test bench for 7-14 optics.
With them the MTF of your '75% MTF' lens becomes closer to a lottery number predictor (for a lottery that only goes to no. 49).
You will understand I am not naming any names !
Bill
-
Ultrapurple,
I just checked the lens FOV for a full moon disk from Earth and it is around 0.5 of a Degree to fill the scene.
My DIOP lens in 250mm mode will provide a Moon disk occupying 0.25 of the thermal image width. With my 320x 240 pixels that means only 80 pixels across the Moons width.
I NEED A BIGGER THERMAL TELESCOPE ;D
-
Ultrapurple,
I NEED A BIGGER THERMAL TELESCOPE ;D
No, just a *really tall* tripod ;)
-
Raw materials and polishing materials and methods are two more things that are early on the learning curve industry wide. Even when using materials that are potentially compatible with mass production such as molding, the processes are in their infancy. The high indexes of many materials put them in sparsely explored territory for optical design. Every where you turn the lens maker is exploring new territory.
I would say it depends who and where you are looking. There are enough well experienced suppliers of pure Ge lenses about who have someone who knows how to run the design software, while spherical polishing goes back to Galilieo !.
Moulded chalcogenides are fairly new but again several companies have been at it long enough to get quite competent and to get the best out of the lens design options.
Equally there are a few 'chancers' about who do not know how to run the software (beyond school level optics anyway), do not understand tolerances or centering and think a 10.6um laser is a test bench for 7-14 optics.
With them the MTF of your '75% MTF' lens becomes closer to a lottery number predictor (for a lottery that only goes to no. 49).
You will understand I am not naming any names !
Bill
Agree there are some quality folks in the business. But we are really saying the same thing. Those experienced designers exist, but their efforts are amortized over hundreds or thousands of lenses instead of of hundreds of thousands or millions. And while software has speeded the process tremendously there is not the centuries of experience with finding designs in visible materials that don't have quite the sensitivity to manufacturing tolerances that others do. The visible legacy only partly translates into the IR. That experience base is much smaller in the IR, and has also been fragmented by the unique requirements of the applications. Most, until the last couple of decades have been driven by military needs, which often pushed strange parts of the design envelope, providing a limited experience base in "mass" produced designs like the current pro-sumer products.
Spherical polishing has been around for centuries, but most of those centuries were optimising grits and other process parameters for two kinds of glass. There are subtle differences and that experience base is much smaller. As you say some have it better than others, and they are a little wary of sharing their secrets.
My experience with optical materials is getting dated, but when last meaningfully involved materials were much less stable than visible materials. Dispersion and other critical properties varied from batch to batch and vendors came and went with regularity. These are issues that the good houses you mention can deal with, but it again requires time and expertise that is amortized over small runs.
Again, I think we fundamentally agree. There are many, many reasons why IR lenses are very expensive relative to seemingly comparable visible lenses. They all contribute some element to the cost, often not too large individually, but it adds up rapidly.
-
Or more smaller pixels...
A 250mm lens coupled to a sensor with 12 or 10um pixels gets a lot better though, comes out at around 200 pixels wide.
Bill
-
Bill,
I just tested the DIOP MWIR lens with a 12um pixel size LWIR core. Sadly, as expected, the lens is mono-band 3-5um AR coated so I cannot use it with any of my LWIR cores. Inframetrics used to produce dual band telescopes but the DIOP unit has been optimised for just one. A LWIR version of the DIOP lens was made for a NASA development version of the Radiance 1 that used a Quantum Well IR FPA but that was obviously a special order job.
I will need to look at the FPA specification of my SC4000 to see if it uses smaller pixels than the earlier Radiance 1 FPA.
Fraser
-
My experience with optical materials is getting dated, but when last meaningfully involved materials were much less stable than visible materials. Dispersion and other critical properties varied from batch to batch and vendors came and went with regularity. These are issues that the good houses you mention can deal with, but it again requires time and expertise that is amortized over small runs.
Again, I think we fundamentally agree. There are many, many reasons why IR lenses are very expensive relative to seemingly comparable visible lenses. They all contribute some element to the cost, often not too large individually, but it adds up rapidly.
Indeed, and while 'pure Ge' is going to be reliable, ZnS / ZnSe may be similar but I do not know them enough. In contrast the various 'flavours' of Chalcogenide come and go. Unless (or even if ?) dealing with a totally integrated supplier (say Umicore) there is an obsolescence risk that could force design tweaks on a long lived design as new and better mixes keep coming along.
Bill
-
The situation with lens prices vs production numbers is well demonstrated by the Nikon UV lens that Ultrapurple sought, and eventually found. That Nikon lens appears pretty ‘normal’ to the uneducated (my) eyes but it it has optical properties at the UV end of the spectrum that make it ‘special’. That UV lens is clearly intended for a niche market and so production numbers were far lower than the conventional lenses. The result is a very high original retail cost and a healthy second hand value.
https://www.nikon.com/products/industrial-lenses/lineup/uv/ (https://www.nikon.com/products/industrial-lenses/lineup/uv/)
You can have a lovely brand new example for just £12K !
https://shop.graysofwestminster.co.uk/product/105mm-f-4-5-uv-nikkor-ais/ (https://shop.graysofwestminster.co.uk/product/105mm-f-4-5-uv-nikkor-ais/)
The UV lens in use…….
https://www.mir.com.my/rb/photography/companies/nikon/nikkoresources/special/105UVmm.htm (https://www.mir.com.my/rb/photography/companies/nikon/nikkoresources/special/105UVmm.htm)
You have to really need such a lens to justify such a high price tag !
On the DIOP DFOV lens front, I have just ordered a suitable power supply to power it :-+ I will be carefully removing the Fischer 8 pin connector from the lens and replacing it with a nice new LEMO 10 pin socket for which I have plenty of plugs :) That is preferable to buying the Fischer connectors at their crazy high prices. The substitution operation appears relatively simple and is low risk to the lens. It may also be easily reversed if I ever find the correct Fischer plugs at a sensible price. I will also download and share the lens controllers firmware for future analysis.
There will not be any invasive work on the lens structure itself as I consider such to be ‘off limits’ unless repair is required. Messing around with a well designed high quality lens assembly carries risk and offers little by way of ‘enlightenment’ as we already know what resides inside the lens barrel.
I will tidy the wiring loom in the lens controller area. In days gone by, my mentors would have had a fit if they had seen cable ties securing the loom ;D I was taught to lace wiring looms as cable ties were considered ugly abominations and the work of the Devil :-DD I have to agree that a well laced wiring loom does look nice if it can be seen, such as in a rack installation. Will I lace the loom in this lens ? Well I am thinking about it :)
As an aside, our older techs were our mentors and many were very strict. They would carry a lace spacing gauge and check your lacing quality. If the spacing was not precisely 10mm and straight they would cut the lacing off with a knife and make you do it all again ! It was very much about self discipline and pride in your work back in the 1980’s and 90’s. These days, companies go for speed and cable ties are the norm. I miss seeing nicely laced looms though.
Fraser
-
A tiny update on the faulty Radiance 1 camera.
The Fault: Stirling Cooler runs and reaches operating temperature but the camera displays no LED indicators on its keyboard and no power to the Electronic Viewfinder. Apart from the cooler, it is basically dead. The cooler is a closed loop control system that likely operates independently of the cameras image processing and control systems. As such it starts working as soon as it receives power and drops into its ‘maintain’ state once the cold finger reaches 77K (-196C). Cooling takes 6 minutes which shows excellent cooler condition. The camera is missing its firmware as the Linear Flash PCMCIA card that holds such is missing. That would stop the camera booting.
I sourced a Series 1 Linear Flash PCMCIA card and cloned my working cameras firmware on to it. Upon checking the clone copy it works in my original camera so the clone process did work. Installing the clone card into the faulty Radiance 1 sadly did not cure the fault. Whilst the cooler starts and behaves normally, the camera is otherwise dead. I suspect a power supply fault within the camera. The Radiance 1 uses an external DC to DC converter so I tried the one from my good camera in the faulty unit. No change. The fault lies within the main camera head.
I was already intending to carry out a careful disassembly of the faulty Radiance 1 camera so no great drama or disappointment about the firmware not curing the ‘no boot’ fault. I will document the disassembly on the forum. I will not be dismantling the cooler or associated FPA vacuum Dewar for obvious reasons. I hope to get this camera working again but even if that is not possible, I can use its good cooled sensor array assembly in my other Radiance 1 camera should the need arise.
I have my good camera as a reference against which to compare voltages etc. so my work is made simpler and less reverse engineering will be required.
For anyone wondering why the firmware PCMCIA card was missing from the camera, there could be many reasons but it is not uncommon for organisations to remove storage devices at the time of disposal in case sensitive data is held on them. The “better to be safe than sorry” policy applies. It could be that the camera developed its ‘no start’ fault whilst in service and that lead to its decommissioning. Another possibility is that the previous owner suspected a firmware fault and removed the PCMCIA card to look at the data on it. Replacement firmware was not, and is not, available so that may have marked the end of the investigation for its new owner.
Fraser
-
I used a Dianyang Technology CA-10 thermal PCBA analyzer to image the main PCBA's of my good Radiance 1 camera and the faulty unit. I was looking to see how much of the faulty camera was active. This was not a detailed examination and was for interest as much as anything. The test did reveal that the faulty Radiance 1 camera has a processor that is dormant and not accessing the SDRAM memory. I also noted one IC on the processor board that is running warmer than that on the known good unit. I will investigate that IC in due course to see whether it is being stressed, or is faulty.
Pictures follow: The first image in each pair is the known good unit and the second image is the faulty unit.
-
Very interesting, Fraser - a fascinating comparison and I wish you well with your repair endeavours.
One thing that caught my eye was where the thermals are presented as a pseudo-3D. I've never seen that presentation before, but I could immediately see how effective it is.
(https://www.eevblog.com/forum/index.php?action=dlattach;topic=285172.0;attach=1237698;image)
(I invite anyone interested in a larger version to click on Fraser's originals, above, rather than this one).
-
I must say I find the virtual 3D display mode very useful as my eyes quickly spot ‘peaks’ and it is great for comparing PCBA’s, as in this case. I asked DYT to enable image capture for this mode but they do not think it is needed as no measurement points are present. They do provide a video recording option in this mode though. I thought the virtual 3D display modes found on some other software products to be a bit of a gimmick but in this usage scenario it has proven to be very useful :)
-
I'm with you on screen capture, even if no actual measurements are shown.
As a general rule, if the software can display something then I think it ought to be possible to save a copy in that form. Even with limitations.
But then, I'm firmly from the 'pretty pictures' end of things...
-
As a side note on the Radiance 1 ‘fault’. I confirmed that the Radiance 1 camera will not show its keyboard LED’s or power the EVF if the processor fails to complete the boot sequence. I did this simple test by removing the firmware PCMCIA card from my good Radiance 1 unit and it failed to boot (as expected) This fact expands the fault possibilities beyond just hardware as I am using a firmware in the ‘faulty’ Radiance 1 from another camera. I cannot be certain that the firmware does not contain camera specific checks and so may not work in another camera. Food fir thought but I will continue my investigation into the hardware fir the time being.
With regard to the thermal analysis of the PCBA, I was looking for components that appeared to be in thermal distress or ones that should be active, and were not. There was also the possibility of components that were not in distress, but were unexpectedly perational or generating slightly higher levels of thermal energy. There are a few points on the faulty Radiance 1 processor PCBA that are warmer than those on the reference PCBA and these could point to a owner or processor state anomaly.
All good fun. I am now in the habit of inspecting ‘dead’ PCBA’s in the thermal domain as well as the visible light domain as much can be learnt from ‘visual inspection’. It can give the repair tech a helpful steer towards the cause of a problem….. including a processor being held in a HALT or RESET state.
-
Totally agree with both you guys (Ultrapurple and Fraser): that 3D image is very useful to rapidly draw your eye to different areas. Good way to quickly tell if you are saturated, have blinkers, or maybe even a focus check.
Thermal inspection is the way to go for troubleshooting. Sometimes it's hard to tell where a circuit card has gone crazy if you were not there when the 'magic smoke' was accidentally released.
-
Thermal inspection is the way to go for troubleshooting. Sometimes it's hard to tell where a circuit card has gone crazy if you were not there when the 'magic smoke' was accidentally released.
In some cases I find that being alerted by a current limit and then being able to check thermally can stop the magic smoke from ever appearing.
Even a second is enough to show thermally that power is going where it should not.
Bill
-
Apologies for the lack of progress reports on the camera and lens. I continued my search for a suitable power supply for the lens as the one I ordered turned out to be unsuitable. I found a nice new XP Power open chassis power supply today that is perfect. 3V3 @6A, +12V @ 3A and -12V @ 0.8A. I will fit a load resistor to the 3V3 output that is unused to ensure stability. At £18 delivered from within the UK it was the right spec at the right price :-+
It is my intention to build a power supply unit that provides the lens power plus the RS232 port for remote control.
The camera will need careful diagnostic work to establish what is causing the failed boot process.
It has been too hot to work on the kit this week so I await cooler temperatures…. Dripping salt laden sweat from your forehead into delicate electronics and optics is not a good idea ;D
Fraser
-
DIOP DFOV lens Firmware :
In case it is of use to anyone in the future, here is the content of the ST M27C512-15 PLCC32 OTP PROM used on the DIOP 75-250 lens controller.
-
I have removed the power input board from the DIOP lens for inspection and modification. I will be removing the 8 pin Fischer socket and fitting a 10 pin LEMO socket in its place. The two centre contacts of the LEMO socket will be left unpopulated. The 8 outer pins match the positions of the Fischer socket so this should be a simple modification and the result will maintain the high quality of connector, whilst using a far more easily obtained make of connector.
Upon removing the power input PCB, I inspected it for any issues. I have previously commented that some specialist thermal imaging equipment can appear to have been hand made on a kitchen table as small production runs were often hand made or used 'cottage industry' production techniques. The DIOP DFOV lens power input PCB is a good example of this ! I personally would not have allowed this PCB design to leave the factory door as it is far from perfect and really needed a redesign to take account of errors made in the original layout. I attach many pictures to show the PCB and its issues.
My thoughts on this PCB design....
1. When I have seen a PCB hanging off of the rear of a connector, it is usually just a breakout PCB for ease of soldering wires or a ribbon cable connector. I do not like to see a larger PCB accomodating a relatively large power converter hanging off of the rear of a connector using only the connector pins for support. The power converter could easily have been mounted in the spare space on the control modules lid. This would have aided cooling and is a more appropriate location.
2. The soldering of the power input PCB is less than great. The soldering of the connectors pins to the PCB has a very dull appearance and poor application. Some of the tantalum capacitors have almost no solder fillet on the end contact, though the joints are nice and shiny.
3. The PCB component layout did not take account of some components physical needs, such a clearance height and avoiding other solder joints. The Varistor adjacent to the Fischer socket does not sit on its PCB pads due to height interferance so it has been soldered at an angle in an unprofessional manner. The Power Trends 5V switching regulator module uses a metal bracket at its top end. On the PCB layout, this bracket would short across the two thermistor bias supplies ! The solution used in production was to roughly distort the bracket and in doing so push the whole module in the direction of the three I/O pins, resulting in them becoming bent and the module casing interfering with the wiring landing pads for RS232 link to the controller PCB. Quite frankly, a bit of a botch and a mess.
4. The PTFE insulated wires connecting to the PCB broke off very easily, suggesting either previous flexing or poor insulation removal technique causing cutting of strands. All wires will be reterminated using a hot bar insulation remover.
5. The PCB had a lot of Rosin Flux redidue on it. It is clear that it was not leaned after soldering of the wires and Fischer connector.
There is an old saying that "you cannot make a silk purse out of a Sows ear" and this applies to the power input PCB. There is only so much that I can do to remedy the poor design without major alterations to the input power system layout. I will rectify the poor soldering, remove and modify the 5V switching Regulator to avoid its bracket being so distorted and to correctly position it on the PCB. The new LEMO socket has a different, smaller diameter, rear design and may provide greater clearance for the Varistor so that it may be placed on its PCB pads instead of hanging off of them on solder bridges.
This was a $35K science grade product ! I think it is pretty clear that DIOP may have been an excellent optics manufacturer, but their PCB design team left something to be desired ! The controller PCB is not without its issues either but I will leave that well alone as "if it is not broken.... do not fix it" applies ;)
Fraser
-
In case anyone needs the PCB track detail, or is just interested, I attach pictures of the power input PCB with its 5V switching regulator and varistor 2 removed. I also include a picture of the 5V switching regulator PCB.
Removing both the Switching regulator and the Fischer connector was made harder by the PCB holes being an interferance fit ! DIOP could not even get the PCB hole sizes right on this PCB design. I resorted to Quickchip low melt solder to remove them and it worked but a lot of effort was still needed to slide the pins out of the holes >:( The PCB holes were then wicked, cleaned, re-tinned and re-wicked to remove the Quickchip residue. For those unaware, you need to get rid of the low melt solder to prevent contamination of the new leaded solder when re-assembling. Note that the PCB mask around the Fischer connector pins was already scratched ... I have no idea how.... maybe they soldered it with a hot spoon or something similar :-DD
I am still considering whether to reinstall the 5V switching regulator on this PCB or whether to mount it in the lens casing. I need to look at its needs in terms of bypass capacitors etc.
Fraser
-
The DIOP lens power input PCB is now reassembled. I decided to leave the design 'as-is' as it has lasted this long ;D
The new LEMO socket uses separate pins and I soldered 0.7mm TCW extensions to them for attaching to the PCB.
The varistor is now in its correct location and all solder joints that looked poor were reflowed.
-
This is really an information post, rather than an update on the lens work.
The DIOP lens is using PTFE insulated cables, as is quite common in professional equipment. These cables can present a challenge to some techs when it comes to stripping them for soldering. One method uses a scalpel to cut the insultation off of the silver plated conductors, but this does risk conductor damage if not done with great care. Such damage can cause failure of outer strands in the cable. I use a hot bar type cutter for removing insulation from PTFE cables. This techique does not damage the conductors and easily strips PTFE insulation. PACE used to make one of these cable strippers but I could not find a used one so elected to buy the Kinetics-Teledyne "Stripall" and import it from the USA. Mine is 117V so I bought a compact 100VA 230V to 110V transformer for it. I also managed to buy several new sets of the stripping bars, but do not expect to need them any time soon. The stripping bars come in two types, flat, or notched. Flat is used for smaller diameter cables and that is what I use most.
You do need to use some common sense when using one of these heated bar type strippers.... you are heating insulation to the point that it melts... that can release gases that are toxic so you should ventilate the work area. The "Stripall" units can strip most flexible insulation types, as the name suggests, but obviously not high temperature fire rated types !
I mention this tool here only because it made my life so much easier on this DIOP lens job. Some readers may not be aware that such a tool exists. If stripping many wires, or specialist insulations, the Stripall is definitely worth consideration. It is far better than a knife or the tip of your soldering iron ! How hot does it get ? Well the bard glow red when operating ;) It was possible to buy a crude temperature controller as an accessory that reduces the operating temperature (this was just a simple lamp dimmer type circuit) but I have not needed such since buying the Stripall some years ago.
Fraser
-
Job done... the DIOP DFOV lens is reassembled and ready for use :-+
For anyone wondering what tool I used to tighen the LEMO socket front castellated nut, I employed my heavy duty flat blade lens wrench. I attach a picture. Lens and watch back wrenches can be very useful tools for all manner of tasks for which they were not intended ;D
I include the old Fischer socket in the pictures as well as the two choices of LEMO connector that I can now use...straight of right angled. As can be seen, the Fischer and LEMO sockets look very similar. The mounting hole for the Fischer is almost 1mm larger than that needed for the LEMO socket though :( I overcame this small issue with a strip of 0.5 mm thick ABS plasticard cut to the required 2mm width to perfectly fit the hole as a liner with, no overhang. The hole liner worked a treat and the LEMO socket was a nice tight fit in the case hole. Nut tightened... job done :-+
Now to sort out the power supply unit and revisit the Radiance 1 camera.
-
The results of current draw tests on the DIOP DFOV lens
Dual rail lab PSU settings
Supply Voltages +12V & -12V
Current Limit set to 2.3A on each power rail
Quiescent Current after boot and self check competed
+12V @ 385mA
-12V @ 85mA
Focus Adjustment motor active
+12V @ 480mA Maximum
-12V @ 85mA Maximum
WFOV to NFOV or NFOV to WFOV change motor active
+12V @ 536mA Maximum
-12V @ 85mA Maximum
Filter Wheel motor active
+12V @ 508mA Maximum
-12V @ 85mA Maximum
As can be seen, the originally supplied 12V@3A / -12V@0.7A SMPSU is a little over rated for driving this lens ! The quiescent state current draw on the +12V rail only just meets the 10% minimum load requirements of many SMPSU's for accurate voltage regulation.
I thought the -12V rail was used as part of the motor drive circuit but this is not the case. It is used to power part of the position sensor circuit. Without the presence of the -12V rail, the lens complains of a Filter home position error and the focus system does not work.
I am now re-thinking my power supply design as I was using 3A rated SMPSU's
Fraser
-
Just to confirm the connector pinout on this lens in case anyone needs it......
Pin 1 RS232 RCV (RX)
Pin 2 RS232 Gnd (Data 0V)
Pin 3 -15V @ <100mA
Pin 4 G- (0V for Minus Rail)
Pin 5 G+ (0V for Positive Rail)
Pin 6 +15V @ <600mA
Pin 7 Spare (Available on Power Input PCB but not connected to Processor PCB.
Pin 8 RS232 TXM (TX)
This information was taken from the silk screen of the Power input Board and agrees with my LEMO connector standard pin numbering.
Fraser
-
Very nice work Fraser.
I wonder if the designers of the board are around to be embarrassed. Perhaps if they are on this forum they could comment on the pressures that caused it to be the way it is.
-
Thanks CatalinaWOW,
This thread has turned out to be a bit of a ‘personal diary’ of my adventures with the DIOP lens. I was working on the controller PCB today to tidy it up a bit. It is yet more poorly thought through electronic engineering. More like a prototype that needed a re-spin to sort out the errors. A perfect example is the use of Molex connectors and the late discovery that the normal Molex PCB plugs, with their locking tab, would not fit in the small spaces provided. The production solution ? Solder the plug in place and then pull the plastic base off of the pins and bend the pins to suit….finally a blob of RTV to hold the Molex connectors onto the pins. The RTV did not work as all connectors easily slide off of the pins.
The lens does work though ;D
-
Excellent breakdown and pictures! Good to know about the power draw.
You got me curious as to 'why' they packed those electronics so tightly. DIOP did make a whole bunch of DFOV lenses and they all appear to have used used the "Smart" electronics. I'd hazard a guess that the electronics were designed for a much larger housing and were forced into smaller packages. The optics designers probably didn't care what it looked like as long as it worked. See pictures below I raided from the old website.
-
Some of the soldering on those boards is shocking, Fraser. I've seen better done at home on kitchen tables.
(https://www.eevblog.com/forum/thermal-imaging/new-arrivals-amber-raytheon-radiance-1-camera-and-diop-75250mm-dfov-lens/?action=dlattach;attach=1242306)
(And, just for total clarity, I am not suggesting Fraser soldered any of the above).
-
It certainly looks like either a 'Issue 1' PCB used as a final development model, or perhaps a conversion of an existing PCB to do something else.
In both cases hopefully before getting some made properly ! :palm:
Bill
-
I have made a decision on the power supply front. That may sound trivial but I want to balance size and current rating against the needs of the lens rather than make some huge power supply that is massive overkill.
The highest current load is the controller PCB that runs from the 5V DC-DC converter. The motor drives are +15V at around 200mA current draw. I already have a neat compact Meanwell desktop power supply that can provide 5V@2.5A, +15V@500mA and -15V@300mA. I intend to use this triple output power supply and make use of the spare pin (7) on the LEMO Connector to provide the controllers 5V rail from the external power supply, rather than the DC-DC converter. Why ? Well using a triple output power supply ensures a quality supply to the controller PCB as I am noting 5V rail fluctuations with the present DC-DC converter when the motors are running :(
The down side of this plan is that it deviates from the DIOP standard build but I can live with that and if another owner, in the future, wants to revert it to standard build, my changes can easily be reversed. The changes are simple. I will move the 5V output wire on the power input PCB to the “Spare” position that is connected to Pin 7 of the power connector. This may have been an official option as the PCB track from pin 7 is decent width so likely intended to carry current.
Once completed, I will have confidence that the controller is running on a decent power rail, drawing around 380mA and the motors will have a dedicated +15V 500mA power rail to call their own ;D The -15V power rail is well catered for with a current rating of 300mA.
I could possibly get away with using just the +15V and -15V power supply outputs but that leaves the +15V rail on the limit of what the power supply is rated to supply in terms of current. It also leaves the rinky dink DC-DC converter powering the. Controller which I am not too keen to do.
So here I am, modifying a $35K professional lens because of a poorly designed electronics package…… not what I was expecting to do but as long as it works well, I can live with it. The alternative would be to replace the electronics package with something more modern as this is just common lens controller stuff. I do not feel the need to do that though. I do not think I will be using the RS232 remote control feature so I will not be bringing that out to a connector for use. I could always use an in-line breakout box to access the RS232 later, if required.
Fraser
Edit: For anyone wondering, when I modify a piece of equipment I always document the modification and place the information in the user manual and on the equipment (or even inside it) on a label. In the case of the DIOP lens, I will be creating a nice connector pin-out label on my Brother laminated label printer using 24mm ‘black on clear’ label media. This avoids future owners becoming confused and provides the information needed to revert the unit to ‘stock’ if desired.
-
Going to try to create a standalone cable for RS-232 and power using that MeanWell power supply. Also will update in a new thread. I have obtained the very rare Triple Field of View - TFOV (60/180/500mm f4) Lens that you mentioned. :)
There were never really any pictures besides one low resolution image in the Indigo Merlin brochure.