Exact match. Not showing close matches.
'[OT] Pentium II Environmental Test'
|Another off topic, although I can see a PIC in the future, question for
this remarkably diverse and knowledgable group....
I need to test a Dell Pentium II based system for 1 hr at an ambient
temperature of 105 deg F, simulating a loss of a/c in a data acquisition
system. I have access to a test oven, but the question of determining how
to test operability has arisen.
Running our software would demonstrate the ability to operate at elevated
temperatures under the load our test software could provide, but how can we
be sure that the paths that the processor might take under abnormal
conditions (not practical to test all possible paths using our test tools)
How does Intel (read Microchip for list purposes) exercise the processor
for testing purposes? Is there a CPU diagnostic that could be run while at
the elevated temperature that would exercise the processor and memory?
In CPU diagnostics I work with (minicomputers), the diagnostics exercise
each instruction with all possible/practical permutations, starting with
the basic instructions and working towards the more complex instructions
and processor functions. Do these exist for the microworld?
Electronic Visions, Inc.
1650 Barrett Drive
Rockledge FL 32955
|I remember a software to test and stress IBM PC machines, 486 and
Pentiums (not sure). It used some special instructions loops that
used to heat more the cpu. There was special considerations and
recommendations about check the cpu fan and case air flow to be
adequade before tun those tests. I think it makes sense to say
that special instructions exercise more the internal cpu circuits,
and it consumes more current, overheating the component. I never
saw any study about what is the relation between current consume
per type of instruction or instructions group, but makes sense
to think it exist, even that not considerable useful.
It is not difficult to measure the current consumed by a PIC
executing loops for different group of instructions or just
one type of instruction during certain time. Measuring
temperature could lead to a non conclusive result, but the
consumed current is a positive factor and indicator.
I think this could be a good research program for the PIC
community, and a nice study subject.
Howard McGinnis wrote:
|If what you want is just to test the CPU and hardware - not your
software - for proper operation, take a look at QA Plus/FE, it's not
free. At all. Jump to http://www.diagsoft.com (See the retail products
link), or jump to www.sykes.com/prodserv/html/qaplus_fe.html
(apparently DiagSoft has sold it to Sykes? or merged? News to me.)
Only annoyance I've had with it is that it doesn't like RLL hard
drives much <G> Doubt that'd stop most people from using it, it likes
SCSI, MFM, ESDI, and IDE well enough...
If the software works correctly when the CPU's cool, and the CPU works
correctly when hot, then probably no software problems when CPU's hot
Howard McGinnis wrote:
> Another off topic, although I can see a PIC in the future, question for
> this remarkably diverse and knowledgable group....
> I need to test a Dell Pentium II based system for 1 hr at an ambient
> temperature of 105 deg F, simulating a loss of a/c in a data acquisition
> system. I have access to a test oven, but the question of determining how
> to test operability has arisen.
More... (looser matching)
- Last day of these posts
- In 1999
, 2000 only
- New search...