Amber Galileo/Radiance HS MWIR 3-5um InSb

[MrSheep]

I wonder if it is because the Radiance 1 uses an entirely different microprocessor architecture. I will have to tear down my non-working Radiance 1 to see if it has the same TMS320C31 DSP chip.

Edit: Looks like it has the same microprocessor

[MrSheep]

Got an original copy of the DSP TMS320C31 Starter kit. Waiting on a floppy drive to come in. Thought it would come with a disk. But given it's from 1996 it makes sense haha

[Fraser]

I completed a teardown of the Radiance 1 this evening and have downloaded all of the IC's datasheets.

I will post the full set of teardown images in a new thread so that it may be found by anyone searching for such in the future. The camera contains some interesting IC's but thankfully nothing very exotic.
Fraser

[Fraser]

As a side note, during the teardown I checked where the Lithium backup cell was connected and traced it to two DALLAS branded IC's. One is the expected RTC+RAM (DS1485S) but the other is a microprocessor power management IC (DS1236S). The non-rechargeable battery is directly connected to these two IC's and my battery was completely discharged. I have to wonder whether this could be why my camera will not boot. Something to be investigated later.

[Fraser]

A quick list of the most significant IC's found in the Radiance 1 camera.

AD688AQ - Precision +/-10V reference
AD811AR - Wideband OP-AMP
AD7874 - Four channel simultaneous sampling, 12-bit ADC
ADS-118MC - 12bit, 5MHz ADC
AM26LS31SC - Quad differential line driver

BT478KPJ80 - 80MHz RAMDAC
BT858KPF - Video encoder

CY7C192-20VC - 64K x 4 Static RAM

DAC8413FPC - Quad 12bit DAC
DS1236S-10 - Micromanager
DS1336S - “Afterburner” Heavy duty power switch
DS1485S - Real Time Clock + RAM (8K x 8 )

HD63484CP98 - Advanced CRT controller
HM628512LFP-7 - 512K x 8 SRAM

IDT 7005-S35J - 8K x 8 dual port SRAM

MAX232 - UART/RS232  level converter
MT42C8128DJ - Specialist Television VRAM
MT5C1005DJ-25 - 256K x 4 SRAM
MT5C1008DJ-25 - 128K x 8 SRAM

SCC68692C1A44 - Dual UART (DUART)

TMS320C31PQL - DSP microprocessor system
TPC1280VB-160C - FPGA

[MrSheep]

Nice work. Looks like it uses the same DSP Chip. Also doesn't look like there is any damage to any of the boards. Does the fan turn on when u tried to power the unit?

(And wow I see a RadioShack battery, shame they closed most of the stores, use to be my go to for buying parts as a kid haha)

[Fraser]

Some developments on the Radiance 1 front. I am aware that this is not a Radiance 1 thread but Mr Sheep took an interest so I will post here for his and others information.

This morning I fitted a new CR123A Lithium cell in the ‘faulty’ Raytheon 1. I did this in the hope that it might resurrect the camera from its sleep as the battery is an important input to the Dallas micro processor management IC as well as the Dallas RTC+RAM.

I powered the ‘faulty’ Radiance 1 with no PCMCIA card installed and the cooler started. The fan also started. The cooler reached operating temperature after 6 minutes and dropped its speed to a gentle Purrrrrr….. like a cat  Of course no indicator lights lit (not even “power”) and the EVF appeared totally dead (no raster). I switched the camera off, inserted a programmed PCMCIA card and restarted the camera. There was no difference to when it was operating without a PCMCIA card fitted. I then installed my original PCMCIA card from my working Radiance 1 into the ‘faulty’ camera and….. it booted normally ! Yes it started and displayed a thermal image  It warned that all configuration settings are missing or corrupt, but that was expected. These settings and calibration tables may be rebuilt via control software. The very first images that I saw out of the camera were of the side of my PACE soldering iron. I then set up a hot coffee mug as my standard “proof of life” image  The dead pixel map and image calibration need to be sorted out, but the camera is definitely not faulty. I checked the Honeywell 8MB Linear Flash cards that I installed the V3.0 firmware on and neither work in the camera. I then installed them in my first, known good, camera and they work fine. Weird ! Further investigation is needed on this anomaly but I have several other Linear Flash cards of various makes and capacities to try in the camera. Hopefully it will like at least one of them ! I am sure I tested the ‘faulty’ camera with my known good PCMCIA card and the camera did not boot. That makes me think the Lithium CR123A cell 3V output is needed for a successful boot. Pictures attached.

Fraser

[MrSheep]

Holy this is huge! Good to know the battery is vital to the boot operation. I knew it was kind of important the way they gave a cut out on the PCB so you could service the battery. I am gonna try this on my other Radiance 1. (Disconnect battery to see if it boots).

Also I had trouble with cloning my card as well. A cloned card doesn't work but my original card works for both Radiance 1. I wonder if the card size and manufacture have to be the same.

Also how did you figure out that the battery was tied to the "Dallas micro processor management IC as well as the Dallas RTC+RAM"

[Fraser]

OK, some testing has been done.....

I programmed all of my Linear Flash cards with the data from my original AMBER provided card. The data copy was verified and my CSM Omnidrive is a very reliable programming tool. None of the cards would work in the second Radiance 1 camera that I thought was faulty I then tried all of the cards in my original, working, Radiance 1 and ...... most of the cards worked fine ! The ones 3 that did not work weer not a surprise as they are later technology and somewhat different to the original linear Flash cards. I attach a photo of the cards that I tested. The ones on the right work with my original camera. The ones on the left do not work in either camera. The question is.... why do the cards work in the original camera and not in the second unit that I thought was faulty ? My immediate thought was software revisions in the camera. There are several IC's in the camera that are programmed and I remembered one on the PCMCIA port PCB. I suspected that this IC might influence Linear Flash card compatibility. Mu original camera has V1.1 on this IC whilst the second camera has V1.0 on the same IC. The other programmable IC's show the same version numbers. I believe teh V1.1 configuration may provide wider Linear Flash card compatibility. Can the V1.1 IC be duplicated ? No idea. That is an investigation for another day.

The original PCMCIA PCCard is a FUJITSU 1MB Linear Flash card with the model number "81011-20". It works in both cameras without any issues.

As a side note, the behaviour of the cooler diffres at switch on depending upon whether the camera is booting or not. When the PCMCIA card is absent or not compatible, the camera starts the cooler as soon as the power switch is operated and the cooler just "does its thing". If the camera starts properly and boots with a compatible card, the cooler behaves slightly differently..... the power switch is operated and the cooler briefly starts, then stops and waits for the camera to get to a certain point in the boot sequence before restarting.

I discovered the direct connection of the lithium cell to the two DALLAS IC's using simple continuity checks. I expected the RTC+RAM to be connected to it, possibly via a diode (not present) and checked the other DALLAS IC after seeing that it is a microprocessor power management IC with a backup battery input.

Fraser

[MrSheep]

Let me check my cameras and see if I can replicate what you are experiencing. I will also check the versions of those chips. And my guess is yes they can be copied. If you look at my TFOV post I got a new T48 Programmer that can prob read that chip . We just need to know what that part number is first.

https://www.eevblog.com/forum/thermal-imaging/amberraytheon-triple-field-of-view-(tfov-60180500-mm)-lens/msg5541547/#msg5541547

[Fraser]

I have just identified the PAL used on the PCMCIA PCCARD PCBA. It is an MMI PAL16L8BCN.

I removed it to see if it can be read using my Wellon programmer and sadly the security fuse has been set. Copying such a protected PAL chip is beyond my current skill set. With a Google search I do see that there are well documented ways to do so with a PAL16L8 however.

Pictures attached

[MrSheep]

Interesting. Seems like a hardware hacking attempt will be needed. Do you have a link to said method. Seems like a brute force attempt of combinations will be needed. :0

[Fraser]

https://forum.vcfed.org/index.php?threads/newbie-needs-help-cloning-a-pal16l8.48690/

http://dreamjam.co.uk/emuviews/readpal.php

It does not sound too involved, though I note that another model of Wellon programmer is said to not work with the detailed "27C020 adapter".

[MrSheep]

I have the TL866CS . Maybe I can give it a go. The one u have is not a registered PAL so that is good news.

Might be a lower priority thought as I have a lot on my plate in regards to the other projects haha

Here is what my card looks like.

[Fraser]

Your card is a standard Intel product.

https://search.app/5HjWykkoLooxJhtp8

This is a low priority for me as well as I can just put the original card in the second camera (the fussy one!) and use one of the other cards in my original camera (the promiscuous one  ). I am curious as to what the issue is though as it is clearly not an attribute problem as neither the original nor the clone cards have attribute memory. The cards should all comply with the relevant PCMCIA standard so the issue with the second cameras PCMCIA port must be subtle. We can forget card based memory controller incompatibility as well as the Linear Flash cards have no such controller (unlike AT cards). A mystery that must wait until another time when I have more spare time. At least I now know that the second camera is working. I just need to investigate whether I can rebuild the calibration tables etc with the utilities that I have available to me.

[MrSheep]

Agreed haha. I want to get the "Non-working" stuff working first. And I think for you to rebuild the NUC u just need to use the RTOOLs Amber Radiance 1 controller software. Were you ever able to get that running. they work with both my Radiance 1's. However please do note that I think one of the menus is bugged. It won't let u leave the tab unless you open the program with the camera off and you hit NULL to switch tabs. After u switch tabs you can turn on the camera again and connect.

[IR_Geek]

this has been an excellent read.   Been a long time since I've done any kind of 'electronics' this deep.   The deeper you guys go the more I feel exchange between Oddball and Big Joe in Kelly Heroes
...
Big Joe: What's happening?
Oddball: Well, the tank's broke and they're trying to fix it.
Big Joe: Well, then, why the hell aren't you up there helping them?
Oddball: [chuckles] I only ride 'em, I don't know what makes 'em work
...

On that note.  My shell of a Radiance 1 has a Fujitsu card near close to Fraser's original 3.0 card but it's only printed in black and white and has "Standard" written instead of "Thermography".   The number on the side is different:  81011-20 9408 K38

There will be software differences between the thermography and standard.   

[MrSheep]

haha nice. And IR_Geek that is a good catch! I wonder what the thermography one looks like menu wise and whether it will work on my camera.

[IR_Geek]

Believe it added (or enabled) another menu group in DOS UserMenu for real time temperature on the video feed.   Unsure on the newer GUI's that came later.  Maybe the same thing.  They were there but wouldn't do anything unless you had the extra thermography in the firmware.

[MrSheep]

Ah ok that makes sense. Fraser might have the Radiance 1T that had a temperature calibration.

Until I get the amber working with RTools temperature calibration will have to wait for me. I already ordered a NI 1422 frame grabber.
This frame grabber was specifically asked for in the RTools manual

[IR_Geek]

One thing to remember is that sometimes a NUC is refereed to as 'calibration' in some documentation and sometimes gets mixed up with a temperature/radiometric calibration.   The IR world and terminology has evolved over the years and also from vendor to vendor. 

The NUC is just done to get all the pixels to behave in the same fashion for each integration setting and lens combination.   Also you can perform a NUC in the camera electronics and ALSO in computer side software.   Can get some weird results if both are enabled at the same time.    Not sure with this camera, but some would let you put the camera side NUC to just the analog video.  This would provide a clean analog video output for real time observation / alignment and then you could get a RAW digital video to a computer.   That way you could 'fix' the video later with better NUC's or calibrations if at some point you discovered an error.  Or if you had to change settings on the fly during a collection.

[MrSheep]

That is a great point. I have seen this before as there are camera nuc files in researchir for my sc6000. ResearchIR also allows for a (PC side) NUC.

BTW RTools is great software haha. Works well with the SC6000. Was a little tricky getting it work with a Phoenix camera as it required a specific chipset and needed the iPORT high performance driver to get it to work on my windows 7 Setup.

NIC that I bought:
https://www.amazon.com/dp/B07CGZD7YJ

I am glad I still have a PC with PCI slots haha. I need it for this card and the frame grabber that is coming in the mail

[Fraser]

Just in case of interest, here is an screen capture of the CSM Omnidrive card analysis for the original 1MB Fujitsu PCMCIA LF card. Note the IC type being a 28F010. I believe that the Intel card detailed earlier uses the same memory IC's. My other cards use the 28F008 and 28F016 memory IC's and I have to wonder if this is the issue with the older PCMCIA port configuration. It may be expecting 28F010 memory IC's ? I shall try to source a card with that chipset to test the theory but they are quite hard to find. I thought that the host system was effectively blind to the actual memory IC's used in Linear Flash cards as they just present an address and data bus to the host, with memory cell allocation on the card (not a true memory controller though). Maybe there is more to it than that 

I have attached the datasheets for the Intel 28F008 and 28F010 memory IC's.

Fraser   

[MrSheep]

Yeah I think I am convinced it has to be the exact same chip. I used an older HP laptop to clone the card I have and it doesn't work even though I was able to read and copy over my firmware. I will eventually test this theory by buying the same 1Mb card.

[Fraser]

I believe we are witnessing the challenges of using Series 1 and Series 2/2+ PC Cards cards in a host. The host has to know what to do with the three different series of cards. Two different software setups are required to cope with both Series 1 and Series 2/2+ cards. If a host is only set up to recognise Series 1 cards, a Series 2 card will not be recognised. Sound familiar ?

The original linear flash card that I have is a Series 1 type containing 28F010 memory IC’s. All of the cards that I bought are at least Series 2, with some being Series 2+ or even Series 3/5. The Series 2 Intel cards use the 28F008 Memory IC and This may be used to spot whether a card is Series 1 or Series 2 etc. Intel Series 2+ apparently uses 28F016 memory IC’s.

What I believe we are seeing with the fussy Radiance 1 cameras is a PCMCIA port configured for Series 1 cards only. When we insert a Series 2 card, it does not know what to do with it. The more promiscuous Radiance 1 camera that I have clearly has a  later configuration that supports both Series 1 and Series 2/2+ cards but it does not recognise Series 3/5 cards, such as those containing StrataFLASH.

A very enlightening Intel document detailing Inter Series PCMCIA card compatibility may be found here……

http://intel-vintage-developer.eu5.org/DESIGN/FLCARD/APPLNOTS/29215801.PDF

Fraser