Two new acquisitions by Fraser - for fans of older professional thermal cameras

[Fraser]

This past week was interesting for me on eBay as I spotted three thermal imaging cameras that were of interest. I bought two and missed out on the third.

Those who know me from my past posts will be aware that I no longer actively hunt for thermal cameras on eBay, but when I see something I like in passing, I can still be tempted ! On this occasion I saw a Raytheon X50 thermal scope, a Bullard Eclipse 160 fire fighting camera and a Raytheon NightSight 200 Pan/Tilt mounted thermal camera. I communicated with the sellers of the Bullard Eclipse 160 and Raytheon NightSight 200 which lead to a deal on each of them The Raytheon X50 had suffered severe battery leakage damage and sold for more than I was willing to spend on such an unknown patient. I wish luck to the successful bidder as important PCB's reside along-side the battery bay 

The Bullard Eclipse is a firm favourite of mine due to its unusual 'organic' case shape. I absolutely love that camera design. I will bid on any Eclipse that crosses my path and I do not have the Eclipse 160 in my collection. This example is located in the UK, so a no-brainer for me to buy. A mutually agreeable price was agreed with the seller and the camera is now mine. She is in nice condition with only a scuff on the polycarbonate screen to be sorted out. Apart from that she just needs a good clean.

Now the purchase of the Raytheon NightSight 200 took quite a lot of thought and pondering on my part as I would not normally jump to buy a Pan Tilt heavy duty thermal camera system. I already trod that path with the Bosch MIC412, Ganz AllView and FLIR M324 Pan Tilt cameras. I do not have a lot of need for pan tilt head thermal cameras ! The Raytheon NightSight 200 does have some appeal for me however. She is an 'Old Gal' from the late 1990's and was created for use by the military and law enforcement agencies. These cameras also found their way into civilian use on board boats for maritime use. The NightSight 200 is not a common thermal camera to find on eBay. She was deployed to organisations that likely destroyed them before disposal. I got lucky and bought a unit that is in very nice condition from a UK Police Force disposals outlet ! She does not appear to have seen much use so was likely a unit bought for evaluation. This camera was originally designed by Texas Instruments Circa 1997 and went to Raytheon when they bought the Texas Instruments thermal imaging business at the end of the 1990's. Raytheon then marketed the NightSight under their brand to the military and US Police Forces. It has a respectable resolution of 320 x 240 pixels from its high frame rate BST core and the lens provides a 12 degree HFOV which makes it a telephoto lens. The focus of the lens is remotely controlled and the camera has a human detection range of ~1500 feet. The Camera head is quite a beast, made of metal and measuring 12" x 12" x 10". She is no 'beauty queen' in the looks department but I have an interest in the unit as an early ruggedised pan/tilt thermal camera for military and law enforcement applications. I have seen pictures of this camera in some of my thermal imaging text books.She has a place in history at the beginning of the 2000's when both BST and microbolometer thermal imaging systems were being deployed for all manner of tasks, including Law Enforcement.

Once the cameras arrive, I shall update this thread with more pictures of the units. The NightSight 200 is coming to me without its little remote control box so I may be facing a challenge when it comes to controlling that camera head. I have written to a company (Aspects) who still support the NightSight 200 to see if they will help with information, but that may not lead anywhere.

If you are a fan of the older thermal imaging technology, I think the inside of the NightSight 200 may be interesting 

Fraser

[Fraser]

Both cameras arrived in todays post 

The Bullard Eclipse 160 is in good condition and fully operational. I think it will need new AA rechargeable cells in its battery pack but that is all on the electronics front. The camera has the optional electronic thermal throttle feature enabled which is nice to have.
The case just needs some gentle polishing and the scratch on the polycarbonate screen is shallow so should polish out relatively easily.

The NightSight is a beast ! The unit appears built to survive a war zone ! I could not resist taking a look inside the head to see what hides within. The answers is ….. a BST core with communications PCB and an unusual rectangular meniscus profile objective lens.
I will post images later 

Not a bad ‘haul’ and the prices were definitely too good to miss

Fraser

[Bill W]

The NightSight is a beast ! The unit appears built to survive a war zone ! I could not resist taking a look inside the head to see what hides within. The answers is ….. a BST core with communications PCB and an unusual rectangular meniscus profile objective lens.

Fraser

Analogue or digital BST ?

The original BST analogue 'dev kit' came with an olive green machined from solid chassis and a rectangular lens.  Might be a picture in one or other of the old BST threads (not from me / Argus).

the Marconi one got binned        :'(

[Fraser]

Bill,

The BST core is analogue and an early version. It is marked up as Texas Instruments and the IC’s are from 1996. The PCB looks like those found in the Argus 2 analogue version. I have not done a detailed comparison yet though. The BST FPA PCB is also marked up with Texas Instruments. From what I understand, what we know as the Raytheon BST core kit was, in fact, a Texas Instruments product that was incorporated into the Raytheon stable when they bought the Texas Instruments Infrared Department.

I have done a teardown of the NightSight 200 and will create a new thread for that in case anyone in the future is searching for NightSight 200 information. The teardown revealed a pretty standard BST core, chopper wheel and lens system design. No surprises there as it was a well designed core kit. The objective lens is an interesting item….. it was made as a standard circular meniscus lens and then cut down to create the unusual rectangular lens shape. The edges where the lens was cut are scary in terms of damage but as only the central area of the lens is being used, it does not matter. There is a darned great chunk of lens chipped out on one side, but it is out of the FPA’s view so of no consequence.

The build quality of the NightSight 200 is very good but not up to the standard that I found in the MIC412 or military gimbal mounted SeaFLIR.

With regard to control systems within the head assembly of the NightSight 200, there is a microprocessor PCB adjacent to the core and a larger, more complex, microprocessor PCB in the base of the Pan/Tilt assembly. There are two composite video outputs (monitor and VCR ?) and the video contains an overlay showing the camera Pan/Tilt direction of view relative to the monitor.

I suspect the small control box contains a microprocessor so I may have trouble finding the right instructions to control the Pan/Tilt and BST core functions. I can always hot-wire the core to get it running though 

[Fraser]

Some pictures from the NightSight 200 teardown showing the interior of the camera and core .....

[Bill W]

The BST core is analogue and an early version. It is marked up as Texas Instruments and the IC’s are from 1996. The PCB looks like those found in the Argus 2 analogue version. I have not done a detailed comparison yet though. The BST FPA PCB is also marked up with Texas Instruments. From what I understand, what we know as the Raytheon BST core kit was, in fact, a Texas Instruments product that was incorporated into the Raytheon stable when they bought the Texas Instruments Infrared Department.

That is the lens and mechanics I recall from the 'dev kit'. 
Yes Raytheon bought TI commercial infrared division - I visited them in Dallas and they were, and a Raytheon site still is, on a TI branded site/campus with a Raytheon 'enclave'. 

Of course the IR business itself was later bought out of Raytheon by L3 in 2004.

[Fraser]

Having slept on the matter of the NightSight 200 controller, I recalled how an old 1980’s VHS VCR wired remote worked. It used different resistances associated with each button on the remote and an ADC in the VCR read the different voltages which were on-passed to the system control processor as commands. I know that the later NightSight 4000B cameras joystick controller was a passive device so maybe I will get lucky with the NightSight 200 ? There appear to be three wires and two coaxial cables  that are not associated with the power side of the camera, monitor and keypad. We know what the coaxial cables are for so I am hopeful that some reverse engineering of the three wires connectivity on the pan tilt base PCB will reveal the inputs that are expected….. data or voltage levels. The challenge will be accessing both sides of the PCB as the rotary contact assembly passes through the middle of it. It could be quite a task so may have to wait until I have the spare time to do it. I will look for IC’s on the PCB that contain an ADC as they might be involved with reading the remote controller signals if voltage level control is in use here. A passive remote controller would certainly meet the requirements of the KISS principle and this is a 1990’s design after all.

I tracked down the original identity of this camera when produced by Texas Instruments. It is the Texas Instruments P100 vehicles mounted pan tilt thermal camera. It’s sister unit is the M100 for maritime use. That may have greater weather sealing in its design to address the harsh conditions experienced at sea. It looks the same, but weighs more. I found the web pages for these cameras from 1997 but Raytheon bought the TI IR team soon after so not much information is present. Raytheon appear to have quickly moved on to the NightSight 2000B pan tilt camera unit so I found nothing on their web site in 1997/1998 about the P100/NightSight 200.

Fraser

[Fraser]

OK, for those who wish to follow along with other techies thought processes, here are my thoughts on the challenge of providing a replacement remote controller for mt NightSight 200. I have zero knowledge of the original remote controller, beyond some pictures of it. I am approaching this problem from the standpoint of a designer wishing to achieve an objective in a manner that is both simple and low cost to the company. I am in my mid 50's so was designing electronics in the 1990's when this unit was designed. This helps me to think in 1990's electronic component 'mode' rather than in the modern world of inexpensive Atmel microcontrollers that abound in modern electronic designs ! Back in the 1990's microcontrollers were not so cheap !

So lets lay down the known facts of the situation to set the scene......

1. The NightSight 200D is a pan and tilt camera system that provides bearing indication to the operator via the display.
2. There are limited options for the thermal scene. These are: White Hot, Black Hot and focus.
3. The Pan Tilt head has x and y axis of movement plus the ability to scan a preset range of bearings.
4. A windscreen wiper assembly is present on the camera head to clean the IR window.
5. The NightSight 200 will only operate with its dedicated remote controller that provides a X-Y joystick plus push buttons.
6. The NightSight 200 cannot be controlled from a PC (this is an important piece of information taken from an FAQ document)
7. There are a limited number of wires between the remote controller and the NightSight 200 head (via the Monitor) A 5 pin DIN plug is used
8. Continuity tests revealed that there are only three unidentified wires coming from the NightSight 200 head to the Monitor connector.
9. The remote control connects to the Monitor unit and not directly to the Head assembly.
10. Main Power for the Head assembly is fed to it on a dedicated cable and connector and is not part of the monitor or remote controller.
11. The Pan Tilt function of the NightSight 200 is dual speed, but not continuously variable speed.


So there we have some facts that I collected together using various data sources on the Internet, including test reports, FAQ's and the limited manufacturers data that is available. I was fortunate enough to find some decent images of the original remote controller unit so can identify the functionality. Pictures attached. From the remote controller images I can identify the following functions:

1. On/Off rocker button
2. +/- Rocker button (Likely for focus)
3. Up Arrow button (possible a "return to Home position - forward looking" button)
4. Left right sweep arrows - Auto scan mode where camera head continuously sweeps across a pre set bearing range.
5. 50:50 Black & White circle - Black Hot and White Hot display selection.
6. Windscreen wiper symbol - IR window wiper activation to clear debris, rain or snow.
7. X & Y movement joystick - Pan & Tilt control. Whether switches or potentiometers are used is not known. Pan & Tilt has 2 speeds of movement.
8. The buttons on the remote controller appear to be backlit translucent types.
9. The remote controller uses a 5 pin DIN plug to connect to the Monitor.

Thinking about remote control from the 1990's designers perspective for a minute........

1. Is pc control of the NightSight required ? - NO
2. Is cost of materials important ? - This is a value commercial product so YES, the BoM is important.
3. Is it justifiable to include a microprocessor/microcontroller and associated support components in the remote controller ? There is already a microprocessor system within the base of the pan tilt head so an additional microprocessor/microcontroller for a relatively few button functions is not justified in this case. As no PC control is required, there is no need for a RS232/RS422/RS485 link for remote control.
4. The KISS principal applies to provide low BoM and reliability in the field.

You will see from the above that no PC control was provided on the NightSight 200 because it was intended to be deployed as a self contained solution with the operator watching a monitor and using a dedicated remote controller. no complications of PC control or associated software were desired. This is a very important factor for me as it removed the likelihood of a complex data communications link between the remote controller and the pan Tilt camera head. Flir cameras of a similar nature do contain an RS232/RS485/ethernet link as they provide the user with a sophisticated controller and the option for pc control of the head assembly. The FLIR remote controllers are both complex and expensive to produce as a result. Even my PM5xx/6xx series cameras used remote controllers that contained a microprocessor for RS232 communications with the camera as the camera offered that data link for control in its standard design. Thankfully the NightSight appears to be a simpler design.

To achieve the desired remote control functionality, I would be able to create a simple resistor network keypad for the buttons on the remote controller, just as is common these days when connecting keypads to Arduino's etc ! Button presses would present differing voltages to the remote control line to the head unit and the voltage would be converted in a ADC to a format that could be understood by the Microprocessor control system within the Pan & Tilt head unit. The microprocessor could then command functions in the thermal imaging core or Pan & Tilt head assembly. That covers the button functions, but what about the X & Y joystick function ? There would be the option to use a microswitch type joystick with dual position switches in each direction to provide dual speed of Pan & tilt movement. A resistor network, similar to that of the button functions could be used to send the joystick position to the Pan & Tilt head units position microprocessor via an ADC. If a potentiometer type joystick were used, the varying voltage coming off the X & Y wiper contact could be converted by an ADC in the Pan & Tilt head and the produced output values interpreted by the Pan & Tilt control microprocessor program. 

So could one ADC be used to read the buttons and the Joystick position ? Hmmm, good question. In theory, yes, but much depends upon the type of joystick used....switched (digital on/off) or Potentiometer (analogue - variable). There are three unidentified wires on the monitor cable of the Pan & Tilt head. It would be possible to use one for the buttons, one for the X axis and one for the Y axis. Such would require the use of three ADC's though. It has yet to be established if there are control related ADC's on the NightSight 200 head, and if so, how many.

There will also be a 0V and +Ve power feed to the remote controller to provide bias voltages and the supply to the button backlight LED's, if fitted.

Well there we have it. My current thoughts on the remote controller for the Nightsight 200 unit.

I attach pictures of an original remote controller sold in an auction and an image of the Controller PCB housed within the base of the Pan & Tilt Head. There is certainly plenty of electronics in that base PCB ! Note that this PCB likely detects the heads position and creates the video overlays for the displayed image on the monitor as well as control of the head unit.

Fraser

[Fraser]

For anyone wondering what the major IC's are on the Pan & Tilt head base PCB ........

MC68HC705B5 - Single chip Microcomputer (MCU) an early type of SoC
XC3042 - FPGA
XC1755 - PROM
AM29F010 - 128K x 8bit Flash Memory
MACH210A - EE CMOS Programmable Logic
MAX622 - High side power supply (VCC +11V output)
LT1016 - 10nS comparator

[Fraser]

The MC68HC705B5 MCU contains an ADC 

It has 8 multiplexed analogue inputs on this little beauty 

Block diagram and details attached.

[Bill W]

PCB layout with 90° bends 

Bunch of amateurs LOL

[Bill W]

Suspect the labels are Raytheon internal part ID's (cage codes) for the programmed devices holding the FPGA & Micro code

3198103 ver B
3198104 ver B

Would need to check but look close to the part ID's used on the analogue BST parts

[Fraser]

Bill,

Yes the PCB track layout does look very old fashioned.

I did a quick bit of investigation into where the three unidentified wires go on the base unit PCB....... I found that two of them go via the rotary contacts to another MCU PCB mounted in the upper head casing. The third wires destination currently eludes me due to lack of access to the rear of the bottom PCB. The PCB is lacquered which prevents easy continuity tests.

The small PCB mounted adjacent to the BST core in the upper head casing is populated with another MC68HC705 MCU plus two LM18293N dual motor drivers. The UC3610 chips are just Schottky diode bridges. This PCB is responsible for controlling the tilt motor and window wiper motor. It also handles connectivity to the BST core so has the potential to control that core rather than being just a connection node. I shall investigate this little PCB further and thankfully it is easily removed so that I can access both sides of it.

Could it be that the remote controller is dealt with by the MCU in the upper head casing ? It is certainly looking that way at the moment. I attach pictures of the upper head casining and the small PCB mounted on one side of it.

Fraser

[Fraser]

As a side note, I found an image of a BST "200 series" Core Kit on the Texas Instruments site. A terrible image but better than nothing 
The PCB does not appear to be the same as that used in my NightSight 200 BST core, but it is similar. These kits were sold to OEM's to enable them to create a thermal imaging system of their own design but using a known good core component set. Many OEM's of the time preferred this path to designing their own cores from scratch.

There is no lens system shown in the the kit but I found an image of the early Texas Instruments automotive night vision thermal camera that appears to use the same lens as that found in my NightSight 200. As Bill has stated, it would appear that a rectangular meniscus objective lens was favoured by texas Instruments for some reason. As previously stated, the rectangular lens is just a cut down circular meniscus lens. Could it be that this lens was already commonly available from another project ? It seems to waste a lot of Germanium real estate though.

Fraser

[Bill W]

The PCB does not appear to be the same as that used in my NightSight 200 BST core, but it is similar. These kits were sold to OEM's to enable them to create a thermal imaging system of their own design but using a known good core component set. Many OEM's of the time preferred this path to designing their own cores from scratch.

Odd that the 'kit' photo does not include the Raytheon detector PCB, but only the mating half of the 80 way Airbourne connector.  Would have expected a Rayhteon image to be fully furnished.
Each OEM usually negotiated their own kit anyway to suit their build, losing the big, empty detector PCB being the obvious one.  As EEV/Marconi we had it as shown but without the motor spider.  Raytheon insisted we bought the motors from them, no doubt to keep up volume and price down for internal camera use.

At the time there were few if any viable options for buying a detector.  All the suppliers (both sides of Raytheon, Lockheed, BAE UK, Indigo)  wanted to send a pre-calibrated kit as that was far better value for them.  Only ULIS later departed substantially from that approach.

Bill

[Bill W]

Another popular DIY control method would be using TV remote control signals.
Hardware encoder in the handset and put the decoder in software in the micro(s)

Bill

[Fraser]

Bill,

Yes I have seen the Philips RC5/6 IR control system integrated into many products. It was well documented and understood, so an obvious choice for OEM’s. I have also seen the venerable old SL490 IR encoder used in simp,e control applications as the matching decoder IC was an easy integration into a system. I have seen the SL490 encoder also used in wired remote controls using a simple Ignacio and 0V communications cable that also provided power to the remote controller. The SL490 signal was superimposed on the DC feed and recovered using a simple circuit before passing to a microprocessor.

I remain hopeful that the NightSight is using a very simple resistor network type remote controller and the ADC’s in the MCU. If not, it will be head scratching time and I will need to consider other possibilities. I will learn more once I trace the unidentified wires to their destinations as if it is the ADC pins…..great, if a GPIO pin…damn it likely has an encoder chip in the remote control. The MCU has integrated RS485 connectivity and if that is being used, I am in trouble as I have no way of knowing the command protocol and reverse engineering the firmware of a MC68HC705 could be a challenge !

My one bit of light in this remote control situation is a passing comment by a 2000B owner who stated that the remote control joystick he had was passive and so simple that it could be easily duplicated.

Fraser

[Fraser]

I now know how the NightSight is controlled !

When reverse engineering a piece of electronics we often hope for a simple and logical approach to the design and that is what I was hoping for in this case. Sadly I was disappointed on this occasion as, despite the camera not having a PC controlled variant, it uses RS485 for its communications. I traced two of the unidentified wires to the small MCU PCB located inside the head assembly adjacent to the BST core. They did not directly terminate anywhere on the MCU so I removed the PCB for visual inspection. On the rear of the PCB there are a few IC's, amongst which is the LTC485. The two unidentified wires terminated at pins 6 and 7 of the LTC485......... a RS485 transceiver IC I knew then that my hopes for a very simple passive remote controller were dashed. The LTC485 drives pin 50 (RDI) of the 68HC705B5 MCU. To determine the communications protocol I would have to reverse engineer the firmware and that is if the IC does not have firmware protection in place. That comes under the "too much work" heading for me at the moment. I could fire up the BST core and get it working without too much bother but this core is little different to those that I already have in the EEV Argus 2 and Talisman Wasp cameras so there is little to be gained. For now, this NightSight 200 camera is a nice addition to my collection and part of the Texas Instruments/Raytheon vehicle borne thermal camera story. I have no regrets about buying it and it is in very nice condition. Just a pity that I cannot get it "rocking and rolling" with a simple remote controller.

I attach pictures of the second MCU PCB.

I must now return to other, higher priority, matters and then the Bullard Eclipse will be getting a full service and restoration.

Fraser 

[Fraser]

I have not forgotten the third unidentified wire that terminates on the lower MCU PCB in the base of the Pan & Tilt head. I will trace that to its destination another day when I have more time to disassemble the PCB/rotary contact assembly from the casing. I am not expecting that wire to save the day though.

Fraser

[Bill W]

that is a real shame Fraser.

You could try sending an ASCII 'U' to the head at 9600 and see if it goes up.

[Fraser]

Good idea Bill. I doubt that I will be that lucky but it is well worth a try  

Fraser