Re: Industrial radiography and computers

["Scott D. Kniffin" <Scott.D.Kniffin.1@gsfc.nasa.gov>]

HPD and group,

No, actually almost no parts are made rad hard or are radiation tested any 
more.  The only people that still buy them in any quantity is the military.  
For civilian use, commercial off-the-shelf (COTS) parts are usually used
these days.  Over here at NASA, in keeping with the "better, cheaper, faster" 
doctrine, we take COTS parts and rad test them for flight.  The theory being 
that if the parts only need a 30 to 50 kRad lifetime for the mission, we can 
test a batch of parts for radiation durability using a low dose rate Co-60 
chamber, if they pass then they go up, if they fall a little short, add
some shielding.  The concept being that you spend about 1/2 to 1/10 the cost 
per chip and a few k$ on testing and you still come out on the cheaper side.  
Eg.: an interesting hybrid part that we've been testing costs about $3k in 
the COTS version, the rad hard (and virtually bullet-proof) part goes for a 
little over $15k each.  

The interesting thing we've discovered WRT the old 1MRad testing, is that
more 
parts pass if you lower the dose rate to something a little more realistic.  
The problem being that you encounter different failure mechanisims at very
high dose rate, such as physical degradation and embrittlement.  Remember,
we're talking about .8 to .5 micron technology here with plans to go to 
.25 micron in the near future.  Yikes!  Well, it keeps me in buisness.  

Scott Kniffin

mailto:Scott.D.Kniffin.1@gsfc.nasa.gov
RSO, Unisys Corp. @ Lanham, MD
CHO, Radiation Effects Facility, GSFC, NASA, Greenbelt, MD

The opinions expressed here are my own.  They do not necessarily represent
the views of Unisys Corporation or NASA.  This information has not been
reviewed by my employer or supervisor.  

At 23:03 03/06/98 -0600, you wrote:
>Interesting thread, many years ago I was reviewing the prototype testing of
>some instruments being installed in a commercial nuc plant for post accident
>monitoring.  Buried in the documentation was a description of the radiation
>hardening testing of the devices.  Evidently, they were exposed to over 1
>million rad of Cobalt 60 or Cesium-136 gamma radiation.  The location was
a pit
>in Taiwan, according to a systems engineer for the utility.  I am pretty sure
>that most electronic equipment is protype tested for radiation hardening, but
>not to the aforementioned levels.
>
>High Plains Drifter
>magna1@jps.net
>
>
>