What's In Your Dishwasher?
C'mon guys, this is fun - snap a shot of what's in your washing machine and post it up. :) Here's what's in mine at the moment:
That is a disassembled HP/Agilent technologies digital storage oscilloscope. The bugger's been built back in '92 - served god-knows-where in the army, and was surplussed. How long it waited out in the military storage facility before it waited in ours is completely unknown. Plug it in, turn it on, nothing happens. Usually they are filled with dirt, dust...shit that makes your skin creep - things that make you think of following D-Con procedures before going home for the day. I found what looked like a chicken getting sucked up by the twin 120mm fans in the rear and chewed to shit to be left to cake and dry in one of these scopes. Not only did it not start, but it smelled damn funny. Had to clean it out with gloves and compressed air before I dumped the whole works into the sink for a wash-down. But anyway, moving on along, spread one of these 60 pound monstrosities away from its chassis and built in color display looks a bit like this. My coffee mug serves as reference:
To the upper left is the "Video card" - more like a backplane manager as all data is dumped when strobed from data acquisition boards, held in RAMDAC and then turned into an analog, 640X480 RGB signal to drive a built in Sony Trinitron monitor on the front panel. Lower left is the actual passive backplane itself. Down the middle are the various DAQ (Data Acquisition) boards and in the lower middle, a pair of Pre-trigger, FFT dedicated boards. These are actually mini super-computers and require heavy duty cooling. The entire backside of the board it heat-piped to a massive, passive heatsink that relies on forced air through the back of the case to keep cool. It's a hot-rod! Cannot run the scope with the case off because it needs directed airflow. To the right is the actual CPU of the system. A Motorola 68XXX based beast, it's robust, proven RISC architecture makes them nearly faultless. Below the upper right, the actual boot ROM, BIOS and factory set FLASH-PROM is there. A few signal/sync-gen boards and two other DAQ boards round up the input processing for this monster machine. Total cost for all of that gold....yes, GOLD laying on my floor - $60,000 back in the day. Sorry, but this is not a home PC, but a piece of very quick scientific processing instrumentation. Signal paths are optimized and lain in gold to keep propagation times to as close to zero as possible. Even the coax headers and all of the pins are made of the stuff. End to end - there is a measurable 2 picoseconds signal propagation time, with actual capture/analysis done within a few milliseconds after the machine crunches sampled data. Moving on:
A close-up of the backplane component - that alone weighs like 5 pounds.
There's the other side of one of twin FFT boards. Man, imagine this running on the back of your over clocked ATI or nVidia card, eh? That thing weighs a TON....would probably break your video card down to a raw, bleeding stump left in the AGP socket. Again, Linux Pengi on my 12 Oz. coffee cup serves as reference.
Turning the passive 'plane over, there is trouble Houston...and probably why the scope was scrapped/non-functional/lost it's calibration:
Let's get a closer look here:
...and here....
The first damage by the screw hole would make one think, "Some guy's torqued its screw too tight". But the blistering/peeling on the other side of the board near the RGB connector to the RAMDAC board and damage in a similar spot on the RAMDAC indicates chemical attack. Sheesh, what on earth attacks polyurethane? Well, I know of a few things, and none are too nice. That is why I wash the entire chassis and scope in the sink before handling it further, begin repairs and perform final alignment. No telling what is in the scope that will be blown out when it's operating, or what outgasses when it warms up to Operating temperature. Here's a closeup of the damage to the RAMDAC too:
Luckily the exposed material is gold and thus is non-reactive when exposed to chemicals and atmosphere, unlike copper boards. However, dust, contaminates and other materials can become lodged underneath, changing the circuit path's inductance and other characteristics. If allowed to run after alignment/calibration, it would continue to decay in performance until it needs serious calibration, or randomly faults a specific path, DAQ board or throws timing circuits totally off. The cure? Simple. X-acto blade the loose edges of material away, careful not to nick the gold underneath. Mask it off with tape and paper and spray the area with automotive clear-coat acrylic. It will creep into the damaged area and seal it off with an impervious, dielectric compatible layer of protection. More to follow later guys and girls. :)
That is a disassembled HP/Agilent technologies digital storage oscilloscope. The bugger's been built back in '92 - served god-knows-where in the army, and was surplussed. How long it waited out in the military storage facility before it waited in ours is completely unknown. Plug it in, turn it on, nothing happens. Usually they are filled with dirt, dust...shit that makes your skin creep - things that make you think of following D-Con procedures before going home for the day. I found what looked like a chicken getting sucked up by the twin 120mm fans in the rear and chewed to shit to be left to cake and dry in one of these scopes. Not only did it not start, but it smelled damn funny. Had to clean it out with gloves and compressed air before I dumped the whole works into the sink for a wash-down. But anyway, moving on along, spread one of these 60 pound monstrosities away from its chassis and built in color display looks a bit like this. My coffee mug serves as reference:
To the upper left is the "Video card" - more like a backplane manager as all data is dumped when strobed from data acquisition boards, held in RAMDAC and then turned into an analog, 640X480 RGB signal to drive a built in Sony Trinitron monitor on the front panel. Lower left is the actual passive backplane itself. Down the middle are the various DAQ (Data Acquisition) boards and in the lower middle, a pair of Pre-trigger, FFT dedicated boards. These are actually mini super-computers and require heavy duty cooling. The entire backside of the board it heat-piped to a massive, passive heatsink that relies on forced air through the back of the case to keep cool. It's a hot-rod! Cannot run the scope with the case off because it needs directed airflow. To the right is the actual CPU of the system. A Motorola 68XXX based beast, it's robust, proven RISC architecture makes them nearly faultless. Below the upper right, the actual boot ROM, BIOS and factory set FLASH-PROM is there. A few signal/sync-gen boards and two other DAQ boards round up the input processing for this monster machine. Total cost for all of that gold....yes, GOLD laying on my floor - $60,000 back in the day. Sorry, but this is not a home PC, but a piece of very quick scientific processing instrumentation. Signal paths are optimized and lain in gold to keep propagation times to as close to zero as possible. Even the coax headers and all of the pins are made of the stuff. End to end - there is a measurable 2 picoseconds signal propagation time, with actual capture/analysis done within a few milliseconds after the machine crunches sampled data. Moving on:
A close-up of the backplane component - that alone weighs like 5 pounds.
There's the other side of one of twin FFT boards. Man, imagine this running on the back of your over clocked ATI or nVidia card, eh? That thing weighs a TON....would probably break your video card down to a raw, bleeding stump left in the AGP socket. Again, Linux Pengi on my 12 Oz. coffee cup serves as reference.
Turning the passive 'plane over, there is trouble Houston...and probably why the scope was scrapped/non-functional/lost it's calibration:
Let's get a closer look here:
...and here....
The first damage by the screw hole would make one think, "Some guy's torqued its screw too tight". But the blistering/peeling on the other side of the board near the RGB connector to the RAMDAC board and damage in a similar spot on the RAMDAC indicates chemical attack. Sheesh, what on earth attacks polyurethane? Well, I know of a few things, and none are too nice. That is why I wash the entire chassis and scope in the sink before handling it further, begin repairs and perform final alignment. No telling what is in the scope that will be blown out when it's operating, or what outgasses when it warms up to Operating temperature. Here's a closeup of the damage to the RAMDAC too:
Luckily the exposed material is gold and thus is non-reactive when exposed to chemicals and atmosphere, unlike copper boards. However, dust, contaminates and other materials can become lodged underneath, changing the circuit path's inductance and other characteristics. If allowed to run after alignment/calibration, it would continue to decay in performance until it needs serious calibration, or randomly faults a specific path, DAQ board or throws timing circuits totally off. The cure? Simple. X-acto blade the loose edges of material away, careful not to nick the gold underneath. Mask it off with tape and paper and spray the area with automotive clear-coat acrylic. It will creep into the damaged area and seal it off with an impervious, dielectric compatible layer of protection. More to follow later guys and girls. :)