Argus 3 (BST) teardown and repairs

[Ultrapurple]

I second what Fraser said about giving back to the community (and also saying thanks to the people who help you along the way).

I expect I will find this thread useful in the future. I have a working Argus 3 sitting in my shed that I intend to 'have a go' at in due course; it's highly likely that I'll refer back here. Thanks all (and especially to Bill for sharing his unique insight).

[Bill W]

Honestly    , this is the least i can do for the community and you guys who are so helpful and good willing.
Here, I just try to re-draft a bit on my own "visual" manner the teaching of Bill.

Also thanks for the proof reading of the forum posts & my website so I know to fill in the bits I have skipped over too lightly.

Bill

[Lambda]

Bill, i can imagine the time you invested in your website for documenting it....

Honestly, i think this is the only source in this specific field and product (on my knowledge) of practical informations reachable for public on a worldwide level...
In advance, thank you for your coming update....

I keep on digging in the guts of this "beast". i tried to imagine two configurations which would allow to control manually the gain and offset/contrast of the video delivered by the core.

- First "hypothesis", see the first scheme.
- Second "hypothesis", see the second scheme.

Here is also the datasheet for the DAC MAX510 (AEWE) from MAXIM (there is the 509 and 510, the document has to be scrolled for reaching the pinout part corresponding to 509 and 510 in correspondance...)

https://nl.mouser.com/datasheet/2/256/MAX509-MAX510-1292568.pdf

For the first hypothesis, i am a bit concern if i cut the analog DAC A and B output (pins 1 and 2) without more precaution: i do not know if there is not somewhere a kind of feedback send back to the SECCA... cutting pin 1 and 2 ("floating" them) will not induce some messing with the SECCA?

I have the impression that the second hypothesis is closer from what you explained to us, Bill, till now.

In both "hypothesis", the configuration of the pin 4 in relation with the divider and "os", "gn" stays the same...
Please, just tell me your thoughts and if i drive on a wrong path....

Thank you again.

Best regards.

Stéphane

[Bill W]

Option 1 will work, but is excessively destructive.  Just remove your choice of resistor to render that DAC output ineffective:
R49 - Offset enable
R50 - Gain enable
R51 - Gain DC level
R52 - Offset DC level
The pad at the SECCA end of the resistor can then be used to inject the required DC.  See attachment

Option 2 is the same DAC condition as you get with no EEV control board in place, so makes no difference.

It will be clearer once I have posted the circuit around the DAC, and the Argus2 equivalent, so maybe do not worry too much for now.

Bill

[Lambda]

Ok Bill, i store it preciously. 

While waiting your update, i stay tune, no hurry.

Just for the discussion, why do you consider the "option 1" I proposed excessively destructive?
Because, at the end, it consists just in cutting pin 1 and 2 (i mean: cutting the PCB terminal tracks/paths connected respectively to pin 1 and 2).

In your way, you have also to act on the hardware by physically extracting resistors (unwelding/cutting)....

Thank you for your comment on this point.

Best regards.

Stéphane

[Bill W]

Yes, they are electrically equivalent, but the resistors are easier to put back than it would be to uncut tracks / pins if you change your mind, that's all.


Bill

[Bill W]

Circuit for the DAC bit attached below

Bill

[Lambda]

 

Thank you Bill !! Now the mist is gone!

Here is a summary about all your explanations (remarks are "copy and paste" from your remarks from previous posts):

Best regards and thank you again!

Best regards.

Stéphane

[Logan]

Hi, thanks a lot for sharing your design file!
I now get a few printed, but my s300a's chopper wheel is glued to the shaft (instead of the screw in some other designs). How can I remove it safely? I'm afraid just pulling it out may damage the motor bearing...

[dalittle]

Bill,
It was such a pleasure to read your post on the Argus 3.  I haven’t heard the term SECCA in over a decade! My goodness.

At ISG we also designed our own FPA board so it would fit in a smaller enclosure. The Raytheon one was huge. However, we didn’t have Raytheon put the FPA on our board, we did it ourselves (saves on the air-miles.)

Most people will not know this albeit Bill probably does: The early 200 series BSTs had Peltiers set to stabilize at 22C while the 300 series BSTs stabilized at (I think it was) 35C and thus that core used more power most of the time.

The iris assemblies on both the EEV units and ISG units were kinda a throwback to the old PEV days. And they were unreliable as hell.  In later PEV variants, ISG removed half the leaves and it helped. Then in BST, changed to a cat-eye design that never ever failed.  It was good enough for firefighting application.

So, Raytheon killed the BST product line because they had a new program to replace it – and I can’t for the life of me remember the name of program – it wasn’t the ASI, it was something else ferroelectric also.  Bill may recall.  In any event, the new technology never came out of engineering at Raytheon.

Raytheon used to bump-bond the pixels to the readout and in the early days, I understand they lost half of them to yield. But over time, and towards the end, they really did perfect that process and yields were in the very high 90’s I was told.  And, in typical fashion, when everything was going great, they killed the program.  Chris Bade was the RTI boss man at the time. We did not agree with their decision. Seems bigger companies have a cancer of bringing people in from outside the infrared space to hold very key positions and they end-up making really bad decisions, then they move on to something else leaving the wreckage behind.

Off topic comment: I was in the UK when 9-11 happened. As I was to leave the following day, I was stuck at the Gatwick Hilton with a Raytheon guy and had an additional 3 days or so because all the flights were cancelled.  Raytheon and I would take the train in from Gatwick and go to a pub on Queens Gate in South Ken called Kavanaugh’s.  We would be wearing jeans and trainers. Every time we walked in, patrons would look at us – ah, Americans? Let me buy you a pint! And we would sit there for hours talking about what had just happened.  The hospitality was bar-none the best I’ve ever experienced.  Mind you, we were in Central London so… these same guys were the ones running you over in the Tube station if you were even the least bit unsure!  Really nice blokes during very unusual times.  Our two countries really do share a lot of common values (although our common language doesn’t seem to be that common.)

Thanks again!

[Bill W]

David,

I was not aware of a 'low temperature' BST, but then ISG were into BST sooner than EEV as you neither had the profit margins that in-house Pevicons gave EEV nor had to go on a year or so wild goose chase after being told to use the GMEO sensor (the 256x128 sucessor to the Cairns Iris PZT types ex Plessey) as that was GEC-Marconi in house.

Low temperature makes sense if your intended use was outdoors at night.

The sucessor chopped sensor from Raytheon was going to be the 'TFFE' thin film ferroelectric.  As you say it never saw the light of day.  As well as the competing ASi (would be the AS3500 at the time) another problem was the EU lead-free solder directive.  Until we sat them down and read it out to them they did not believe it applied to anything other than consumer equipment.

There was also the issues around export licensing which was getting worse each new annual order.  That was why Argus had sensors fitted at Raytheon, it meant we imported 'cameras' not 'sensors'.  Yes the PCB crossed the atlantic two times (or 3 if the customer was MSA USA), but the Argus2 shutter motor went across 4 or 5 times.
Made in Switzerland
Raytheon insisted on selling them in the kit to get pricing - so over to USA
Over to UK to get assembled
Back to USA for the sensor fitting
Back to UK to build camera !



regards
Bill

[Logan]

Hi, thanks a lot for sharing your design file!
I now get a few printed, but my s300a's chopper wheel is glued to the shaft (instead of the screw in some other designs). How can I remove it safely? I'm afraid just pulling it out may damage the motor bearing...

Okay I removed the glue by screwdriver...
I just made a modified version of the file, attached. Rename the ".txt" to ".stl" after download.
The total thickness is reduced to 2mm, the "film" thickness is 0.6mm.
Reason for this is, sometimes the thin "film" is not perfectly flat, it will bend towards outside, together with the total 3mm thickness, I cannot avoid it scratching other parts.
I tried Blender for the first time, just because this task...
I also wanted to center the "film", to make it not start from one side, but the polygon is fused together to the frame, so I haven't success yet.
It would be really nice to balance the weight, but I don't know how to do it.

[dennis1573]

Hi Logan,

i have printed a few of these, and never had problems with warping/bending.
If you have a heated printbed, make sure its completely cooled down to room temperature before you remove it.
I dont think its a good idea, to center the Spiral. That way, it wont be in contact with the printbed and you need supports...
I think you should print it in two parts, 1st part the Wheel with the Spiral on the bottomside and the 2nd part just the Wheel then you glue the 2 halves together.
I have attached the Chopperwheel with the Spiral centered (2mm Frame 0.6mm Spiral) and the same in 2 parts so you wouldnt need supports.
I cant help you with balancing, i always printed it with 0.3mm for the Spiral so that was never needed...

[Logan]

Thanks a lot dennis!

[Logan]

Well I tested the solid version.
The dynamic range is really bad, and there's NO improved sensitivity as I hoped.
Also there's fixed pattern noise(?!) and uneven brightness with solid chopper.
No wonder why Raytheon just choose the dotted film version...

[Bill W]

The uneven brightness might be a slightly out of phase due to the greater weight of the solid chopper.
Check if the detection pulse is in the same place with each.

It should be the same delay after video field pulse.  On Argus3 there is a significant delay between them as we moved the chopper sensor from the Raytheon expected location.
If you can get at 'CHFB' signal that might be zero delay to video - I'd need to check

Bill

[Gareth79]

Seeing this thread prompted me to look at the 2x Argus 3s I have, I thought both I had needed repairs, but seems one is ok.

I bought 3x of them a few years ago, one was in perfect order and I resold it (for more than I paid for all 3).  One had a dead screen which I replaced with an LCD monitor, the other used to take quite a while to warm up.

I just powered them up though (connecting directly to the contacts) and one seems fine. The other has a permanently grey image (even after leaving it several minutes), I think this is one I left on overnight and which had the grey image when I returned (oops!), is that likely to have damaged something?  I have stripped it down and am awaiting a Nintendo screwdriver to be able to remove the control board, with a view to making a more compact device. (A friend thinks it might be an RS System Zero screw?)

Pic of the working one!

[Bill W]

I just powered them up though (connecting directly to the contacts) and one seems fine. The other has a permanently grey image (even after leaving it several minutes), I think this is one I left on overnight and which had the grey image when I returned (oops!), is that likely to have damaged something?  I have stripped it down and am awaiting a Nintendo screwdriver to be able to remove the control board, with a view to making a more compact device. (A friend thinks it might be an RS System Zero screw?)

Yes, they are RS 'tamperproof' types.  I'll advise the RS parts when I can.

Leaving on overnight should not be a problem, although long term use was never a consideration in design.  If you can quantify 'grey' it might help.
A noisy grey image suggests all the electronics is working and the cause is more a closed iris, which would also generate the 'warning triangle of doom'

A flat noise free grey image would suggest zero gain is being applied.  Then you look to the gain biases and similar.

A blank (ie video black) image can have a lot more causes. 

A scope on the BNC out will show up which if it is not obvious.

Note that removing the control board also removes all the gain / offset leaving the core set to manual control, no input' which is zero gain and offset.  You can use the information in the tear-up thread / fire-tics project to convert the stripped down core into manual mode.
http://www.fire-tics.co.uk/project3/index.htm

Bill

[Gareth79]

I just powered them up though (connecting directly to the contacts) and one seems fine. The other has a permanently grey image (even after leaving it several minutes), I think this is one I left on overnight and which had the grey image when I returned (oops!), is that likely to have damaged something?  I have stripped it down and am awaiting a Nintendo screwdriver to be able to remove the control board, with a view to making a more compact device. (A friend thinks it might be an RS System Zero screw?)

Yes, they are RS 'tamperproof' types.  I'll advise the RS parts when I can.

Thanks, I sent a pic to a local friend who has a set of those drivers so I may not need to wait for the Nintendo one to arrive even.

Leaving on overnight should not be a problem, although long term use was never a consideration in design.  If you can quantify 'grey' it might help.
A noisy grey image suggests all the electronics is working and the cause is more a closed iris, which would also generate the 'warning triangle of doom'

A flat noise free grey image would suggest zero gain is being applied.  Then you look to the gain biases and similar.

It's the proper video with overlays etc. but flat grey and no noise that I could see, it basically stays like the grey from startup and doesn't 'clear' to reveal the thermal image. The triangle didn't appear within several minutes, I have seen it on one of them before. I'll get it all hooked up later and check all the test points in detail.

Note that removing the control board also removes all the gain / offset leaving the core set to manual control, no input' which is zero gain and offset.  You can use the information in the tear-up thread / fire-tics project to convert the stripped down core into manual mode.
http://www.fire-tics.co.uk/project3/index.htm

I've already had a good look over that, very useful many thanks! I will put it all in a more compact box for sure,

When I'm done I can post you all the spare bits for your parts collection!

[Gareth79]

Minor update - the friend dropped over their RS System Zero tools and I have a solution early

[Gareth79]

Ok, I finally found time to have a futher look at the "grey screen" Argus 3.

I stripped it down and replace the power/video header, and then added the RS232 header, and bridged R66. I could connect fine using the configuration software, and figured out the controls, eg. changing to the lower speed (9,600 I think) and setting it to "Computer".  Basically whether the Auto or slider mode is set it just has the same solid image, with a hint of noise and banding (pic below).

I think stripped down the working camera I had and repeated it, to check I was doing it correctly, and indeed with that one I get the correct image (pic below).

Any suggestions welcome! I prodded a few test points but am not yet sure what to look for. I will have a look over the circuit diagrams. Thanks!