running a computer in extreme cold

[Kirt Fitzpatrick]

Does anyone know what the affects of using a typical desktop in outdoor
temperatures would be? I am toying with the idea of putting my desktop into
my car and running it with a kiosk touch screen in the dash. I am pretty
sure I can make it all work accept that I am unsure about operating temps. I
know you shouldn't take a PC that has been sitting OUTSIDE in the cold and
bring it INSIDE into the warmer humid air and imediately turn it on because
of condensation. But if is stays out side then condensation isn't a factor
since it will be the same temp as the air and when turned on is will become
warmer than the air around it in a hurry.

I live in New Hampsire where it stays below freezing for months. The
processor should work at that temp (I assume since I have read articles
about people overclocking cpu's using liquid nitrogen to cool them). But
will the hard drive and DVDRom?

kirt

 

[Grunff]

Does anyone know what the affects of using a typical desktop in outdoor
temperatures would be? I am toying with the idea of putting my desktop into
my car and running it with a kiosk touch screen in the dash. I am pretty
sure I can make it all work accept that I am unsure about operating temps. I
know you shouldn't take a PC that has been sitting OUTSIDE in the cold and
bring it INSIDE into the warmer humid air and imediately turn it on because
of condensation. But if is stays out side then condensation isn't a factor
since it will be the same temp as the air and when turned on is will become
warmer than the air around it in a hurry.

I live in New Hampsire where it stays below freezing for months. The
processor should work at that temp (I assume since I have read articles
about people overclocking cpu's using liquid nitrogen to cool them). But
will the hard drive and DVDRom?

The cold isn't an issue - but humidity is. Depending how dry your car
is, you may or may not be OK. If the car gets quite damp, then you may
want to consider securing a bag of silica gel inside the PC case. This
will suck moisture from the air, and release it as it warms up when the
PC is on.

 

[Alien Zord]

Does anyone know what the affects of using a typical desktop in outdoor
temperatures would be? I am toying with the idea of putting my desktop into
my car and running it with a kiosk touch screen in the dash. I am pretty
sure I can make it all work accept that I am unsure about operating temps. I
know you shouldn't take a PC that has been sitting OUTSIDE in the cold and
bring it INSIDE into the warmer humid air and imediately turn it on because
of condensation. But if is stays out side then condensation isn't a factor
since it will be the same temp as the air and when turned on is will become
warmer than the air around it in a hurry.

I live in New Hampsire where it stays below freezing for months. The
processor should work at that temp (I assume since I have read articles
about people overclocking cpu's using liquid nitrogen to cool them). But
will the hard drive and DVDRom?

My company makes computers for public information systems, meaning displays
for railway and bus stations, shopping centres etc. We found that monitors
and hard drives do not work correctly bellow 10C. So all our systems have
heaters and do not allow computers to boot up unless the temp is above 10C.

 

[Pen]

Be aware that the typical guaranteed minimum spec for ICs
is 0 C(32 F). In my experience, any device with
an integrated circuit in it, at best, will not work until all
parts are above freezing.

 

[Grunff]

Be aware that the typical guaranteed minimum spec for ICs
is 0 C(32 F). In my experience, any device with
an integrated circuit in it, at best, will not work until all
parts are above freezing.

I'm sorry, but that's just not true. ICs are perfectly happy down to
liquid nitrogen temperature (-196C). I've run many circuits down at
those temperatures, using a wide range of bog standard ICs, with no
problems at all. There is no physical phenomenon which could impede the
operation of an IC at 0C.

 

[sdeyoreo]

hen why the different temp ratings between comercial and military
grade ICs?

 

[kony]

I'm sorry, but that's just not true. ICs are perfectly happy down to
liquid nitrogen temperature (-196C). I've run many circuits down at
those temperatures, using a wide range of bog standard ICs, with no
problems at all. There is no physical phenomenon which could impede the
operation of an IC at 0C.

True, the ICs aren't the problem, but the mechanical parts and
electrolytic capacitors may easily be.

 

[colinco]

I am toying with the idea of putting my desktop into
my car and running it with a kiosk touch screen in the dash.

Have you considered its impact resistance?

 

[Grunff]

True, the ICs aren't the problem, but the mechanical parts and
electrolytic capacitors may easily be.

No argument there.

 

[Grunff]

hen why the different temp ratings between comercial and military
grade ICs?

Max temp is far more relevant than min.

 

[Kirt Fitzpatrick]

what do you mean by impact resistance? The screens durability or how cold a
screen can be and still work? I have found some that work a little below
freezing. and all the ones I looked at can be stored down to at least -20
C.

kirt

 

[kony]

hen why the different temp ratings between comercial and military
grade ICs?

Might be a different epoxy used, or simply that they're guaranteeing
it over a wider range of temp because they did more testing, and
charged more $ to offset the extra liability, insurance.

 

[kony]

Have you considered its impact resistance?

Survival of the system might be a secondary concern if there's going
to be an "impact".

 

[Ben Fullerton]

kony ([email protected]) wrote:
: On Thu, 5 Feb 2004 10:02:56 +1300, colinco <[email protected]>
: wrote:

: >In article <[email protected]>,
: >[email protected] says...
: >> I am toying with the idea of putting my desktop into
: >> my car and running it with a kiosk touch screen in the dash.
: >>
: >Have you considered its impact resistance?

: Survival of the system might be a secondary concern if there's going
: to be an "impact".

...... and now for a minority opinion - or at least a question.

I have a solid state "ham" radio that disagrees with you, and an
explanation from an electrical engineer to give it a bit more weight.

Most conductors decrease their resistance as the temperature drops.
(Superconduction magnets (very close to zero resistance) are the most
powerful electromagnets known, and consume the least power.)

The problem that damaged my radio was that the lowered resistance of some
components (at about -15C or, if you prefer, +4F) allowed such a high
current that it cooked two LEDs and came close to doing damage to other
components.
The electrical engineer ran an Electronics and Amateur Radio shop. He
told me that they had far more rigs come in with damage from being turned
on at sub-zero temperatures than they ever saw from rigs that got baked in
the sun behind car windshields .... and explained why.

He explained that this was "inrush" or "start-up" current only, and that
if the rig was turned on at mild temps, and THEN moved to the cold
environment, there was very little risk.

How much of this translates to solid state computer electronics?
I have no idea - but I would not risk my computer to find out.

Ben F.

 

[colinco]

what do you mean by impact resistance?

The ultimate Windows driver error

 

[Grunff]

Most conductors decrease their resistance as the temperature drops.

But not as much as you think - nowhere near as much.

(Superconduction magnets (very close to zero resistance) are the most
powerful electromagnets known, and consume the least power.)

Superconductivity is a totally different physical phenomenon - none of
the conductors in a PC will superconduct at -100C.

How much of this translates to solid state computer electronics?
I have no idea - but I would not risk my computer to find out.

Not at all, because unlike conductors, semiconductors *increase* in
resistance as temperature drops.

 

[CBFalconer]

*** top-posting fixed ***

hen why the different temp ratings between comercial and military
grade ICs?

Because there are dinky little wires and other things bonding chip
pads to external connections, various glues and whatnots in the
packaging. These are subject to thermal cycling, differential
cooling, cracking, failure, whatnot.

Please post any replies AFTER the material you are quoting, after
snipping the portions that are not germane.

 

[DevilsPGD]

In message <<[email protected]>> "Kirt Fitzpatrick"

what do you mean by impact resistance?

Speed bumps.

 

[Alien Zord]

No argument there.

Yep, quite right. Its the thermal calibration of hard drives that is at
fault. Ever since voice coils replaced stepper motors as head actuators HDDs
had to perform thermal calibration every few minutes because voice coils are
very temperature sensitive. When an HDD is powered up at low temp it
performs its calibration straight away but its internal temp can rise 10
degrees inside first minute resulting in its heads simply reading the wrong
tracks and crashing the computer in the process. Our ECU powers up the
computer but stops it from booting by holding the reset line low until the
temp stabilises and is at least 10C.
Scanning coils and deflection circuitry of CRTs suffer from similar
problems.

 

[notritenotteri]

how's your anal retention ring functioning?

*** top-posting fixed ***



Because there are dinky little wires and other things bonding chip
pads to external connections, various glues and whatnots in the
packaging. These are subject to thermal cycling, differential
cooling, cracking, failure, whatnot.

Please post any replies AFTER the material you are quoting, after
snipping the portions that are not germane.