Can a voltage drop affect a computer?

[curious]

I just set up an air conditioner in the same room as my computer. I've noticed that whenever the air conditioner turns on, the light in my room dims for a split second and then goes back to normal. The computer, lamp, and air conditioner are all plugged into different electrical socket panels.

I'm wondering if the voltage drop that happens whenever the air conditioner turns on could affect my computer in any way? Could it somehow corrupt the memory or anything like that?

 

[kony]

I just set up an air conditioner in the same room as my computer. I've noticed that whenever the air conditioner turns on, the light in my room dims for a split second and then goes back to normal. The computer, lamp, and air conditioner are all plugged into different electrical socket panels.

I'm wondering if the voltage drop that happens whenever the air conditioner turns on could affect my computer in any way? Could it somehow corrupt the memory or anything like that?

Please adjust your newsreader to use shorter line lengths.
80 characters is often considered the max but lower tends to
be more readable. That may result in more replies, too.

Yes the drop can effect a computer, but on most (better) PSU
several stages of capacitor filters will buffer it enough.
you might want to run some stress tests on CPU, memory,
video (if you're a gamer) to verify stability while the
heavy applicances like an AC unit are power cycling.

You might also consider a separate circuit for the AC, or
the computer, or even an online UPS (with line conditioner
built-in) if it continues to concern you.

 

[Paul]

I just set up an air conditioner in the same room as my computer. I've =
noticed that whenever the air conditioner turns on, the light in my room =
dims for a split second and then goes back to normal. The computer, lamp, =
and air conditioner are all plugged into different electrical socket =
panels.

I'm wondering if the voltage drop that happens whenever the air =
conditioner turns on could affect my computer in any way? Could it =
somehow corrupt the memory or anything like that?

No.

The ATX power supply has a store of energy in the capacitor
on the primary side of the supply. Even if the power is
removed for 15 milliseconds or so, there is enough energy
stored to keep the output of the power supply stable.
(And, in fact, that is why a UPS is able to operate computer
equipment without a problem - it takes time for the UPS to
switch to batteries, and the holdup capacity of the supply
is why the computer doesn't even notice.) When you buy an
ATX supply, some of the advertisements tell you what the
holdup time spec is.

The power supply is also rated to run at reduced voltage,
so if a 110V circuit drops to 90V, the ATX supply will continue
on. So the supply can resist electrical disturbances.

If a disturbance lasts long enough, then yes, the output
will glitch, but we're talking a pretty hefty load, and
likely not a load that can be sustained without tripping
a breaker.

Now, all of this assumes you have wiring that is up to code,
and the air conditioner is on a circuit rated to carry the
load.

Paul

 

[Vanguard]

I just set up an air conditioner in the same room as my computer. I've
noticed that whenever the air conditioner turns on, the light in my room
dims for a split second and then goes back to normal. The computer,
lamp, and air conditioner are all plugged into different electrical
socket panels.

I'm wondering if the voltage drop that happens whenever the air
conditioner turns on could affect my computer in any way? Could it
somehow corrupt the memory or anything like that?


--- REPLY SEPARATOR ---
(only needed due to OP using quoted-printable for newsgroup post)

The power supply in a computer is a switching power supply. It will
regulate its output voltages with the input voltage ranging from 90 to
135 VAC in 120VAC mode (180 to 265 VAC in 240VAC mode).

 

[Mike Walsh]

Yes, if the voltage drop is long enough or low enough it can corrupt memory and even hard disk data. If you can see the lights dim it is long enough to cause problems. The reason that your PC still runs satisfactorily is that a well designed power supply is able to supply the correct output voltage even when the input AC voltage is much below what it is supposed to be. If you are a little paranoid you can use a UPS that will kick in during such voltage sags.

 

[Richard Brooks]

I just set up an air conditioner in the same room as my computer. I've noticed that whenever the air conditioner turns on, the light in my room dims for a split second and then goes back to normal. The computer, lamp, and air conditioner are all plugged into different electrical socket panels.

I'm wondering if the voltage drop that happens whenever the air conditioner turns on could affect my computer in any way? Could it somehow corrupt the memory or anything like that?

When I used to teach we had the school heating system kick in at a
specific time and all the computers would lock up. Now, my Freeview set
top box drops the picture and sound when our household heating system
starts up and that's using a PowerClean plug system that's meant to take
out spikes etc, so I'd say "yes, it's possible!"


Richard.

 

[philo]

The ATX power supply has a store of energy in the capacitor
on the primary side of the supply. Even if the power is
removed for 15 milliseconds or so, there is enough energy
stored to keep the output of the power supply stable.
(And, in fact, that is why a UPS is able to operate computer
equipment without a problem - it takes time for the UPS to
switch to batteries, and the holdup capacity of the supply
is why the computer doesn't even notice.)

<snip>

That's why I always recommend a *ferrorsonant* based UPS...

as there is *no* switchover lag at all

(at worst, a minor decrease in amplitute)

 

[default]

I just set up an air conditioner in the same room as my computer. I've noticed that whenever the air conditioner turns on, the light in my room dims for a split second and then goes back to normal. The computer, lamp, and air conditioner are all plugged into different electrical socket panels.

I'm wondering if the voltage drop that happens whenever the air conditioner turns on could affect my computer in any way? Could it somehow corrupt the memory or anything like that?

It "could," but how likely that is would depend on lots of factors . .
.. Try it an see.

If it is on another circuit and the wiring is sized heavy enough
things should be OK. Having the lights dim would cause me to question
which circuit lights and AC are on, or the quality/size of the wiring,
or service capacity to the house.

If you get some specific symptoms post about those along with the same
questions, there may be an easy remedy.

My electric range was talking to my modem causing it to drop out - a
power filter on the range cured that. The 120 VAC switch on the
speaker amplifier for the computer would cause the computer power
supply to drop out and crash my system - a snubber across the switch
contacts and a ferrite bead fixed that.

 

[Paul]

I've noticed that whenever the air conditioner turns on, the light in my
room dims for a split second and then goes back to normal. The computer,
lamp, and air conditioner are all plugged into different electrical socket
panels.conditioner turns on could affect my computer in any way? Could it
somehow corrupt the memory or anything like that?

When I used to teach we had the school heating system kick in at a
specific time and all the computers would lock up. Now, my Freeview set
top box drops the picture and sound when our household heating system
starts up and that's using a PowerClean plug system that's meant to take
out spikes etc, so I'd say "yes, it's possible!"


Richard.

I believe some of the small switching power supplies ("wall wart")
have poor holdup time characteristics. If the wall wart used
the older style transformer, bridge rectifier, and filter capacitor,
it would be less likely to drop out.

This is an example of a switching power supply (wall wart):

http://www.usdigital.com/products/ps24/pictures/60-Watt-Regulated-Switching-Power-Supply-PS24.jpg

As another poster pointed out, there are other mechanisms that can
cause glitches in electronics. Common mode chokes, ferrite beads,
and other after-market devices can be used to defeat the problems.
But why should users have to reengineer the stuff they paid
good money for ?

Paul

 

[DaveW]

It could possibly burn out your power supply unit, and hence possibly your
motherboard.

--
DaveW

----------------
I just set up an air conditioner in the same room as my computer. I've
noticed that whenever the air conditioner turns on, the light in my room
dims for a split second and then goes back to normal. The computer, lamp,
and air conditioner are all plugged into different electrical socket panels.

I'm wondering if the voltage drop that happens whenever the air conditioner
turns on could affect my computer in any way? Could it somehow corrupt the
memory or anything like that?

 

[DaveW]

It could possibly burn out your power supply unit, and hence possibly your
motherboard.

--
DaveW

----------------
I just set up an air conditioner in the same room as my computer. I've
noticed that whenever the air conditioner turns on, the light in my room
dims for a split second and then goes back to normal. The computer, lamp,
and air conditioner are all plugged into different electrical socket panels.

I'm wondering if the voltage drop that happens whenever the air conditioner
turns on could affect my computer in any way? Could it somehow corrupt the
memory or anything like that?

 

[Jamie]

I just set up an air conditioner in the same room as my computer. I've noticed that whenever the air conditioner turns on, the light in my room dims for a split second and then goes back to normal. The computer, lamp, and air conditioner are all plugged into different electrical socket panels.

I'm wondering if the voltage drop that happens whenever the air conditioner turns on could affect my computer in any way? Could it somehow corrupt the memory or anything like that?

hmm, to be safe i would use a UPS on the computer.

 

[Electromotive Guru]

Okay for all of you not skilled in inductive appliances

Startup current foi rthe typical A/C compressor is about 115 amps
droppoing rapidly to 40 in the next few milliseconds, and the
finalizing at running current of 6-12 amps. Air-conditioners ar
highly-inductive loads

Surge suppressors are meant to resist high-spikes from electrica
interference, including power-station switching, not short-ter
brown-outs

The current stored in a copmputer's P/S cap is NOT enough to sustai
voltage for any duration, simply because it is only to filte
incoming supply ripple, not output ripple. Once you lose power an
line frequency, the power supply function is immediately halte
because there is no timing reference signal for which to switch th
supply at. The circuit collapses entirely and nothing in the circui
sustains any voltage to anything, since the oscillator signal i
obtained by an opto-coupler that reads crude DC ripple as
sync-pulse

A USPS (not just a UPS) is the best way to protect your machine fro
any voltage variation, since the batteries will make the ultimat
capacitors (equivalent capacitance of a 12V 5Ah battery is well ove
4 farads) and are always connected, so if this is a concern and yo
can afford it, this is the best way to protect your machine. A UP
can still hiccup enough to cause a potential error, especially durin
heavy processing

The difference between a UPS and a USPS is that a USPS is like a
always-on UPS, whereas a UPS simply tries to restore power as quickl
as it can when line voltage is lost. Some UPS's won't switch fas
enough because the trigger voltage ios set for 85VAC and no
95VAC...but they cost a lot more and tend to draw a little mor
power, even on standby

 

[default]

It could possibly burn out your power supply unit, and hence possibly your
motherboard.

How does that happen? The purpose of regulators is to keep the
voltage constant. If the input voltage drops below the 95 volts or so
that the regulator requires to keep the voltages up, the voltage
monitors should catch it and reset the computer or do whatever the
bios is programmed to do in the event of a power failure.

Likewise there's usually some provision to "crowbar" the output to
hold the logic supplies within limits - won't allow them to go too
high to protect the motherboard.

As a general rule, too high a voltage can cause damage, but low
voltage is only bad for induction motors and some poorly designed
audio amplifiers. . . . . as a general rule.

 

[Richard Brooks]

I wonder how many power supplies have over voltage protection. I bet a lot
of the cheaper supplies are stripped to the bone.

I've had the 5v regulator burn high (12v or more) rather than burn open
dropping the output to 0v, taking every single drive and board with it.
Basically I had a metal case to use. A vendor I told this to didn't
sound surprised as he'd had this happen two or three times in the year
previous!

The smell of Bakelite every time you switch on then becomes a scary
thing if you don't wash the case out and put new boards in.


Richard.

 

[kony]

How does that happen? The purpose of regulators is to keep the
voltage constant. If the input voltage drops below the 95 volts or so
that the regulator requires to keep the voltages up, the voltage
monitors should catch it and reset the computer or do whatever the
bios is programmed to do in the event of a power failure.

Keep in mind that DaveW often goes off half-cocked,
suggesting things not in evidence in all kinds of crazy
scenarios. To DaveW, everything is a problem and all parts
must be thrown away, the system rebuilt from scratch, and of
course the OS reinstalled else there will be imaginary nasty
ongoing registry errors, blahblahblah (nonsense).

If the input voltage to the PSU is too low, it won' t be
able to regulate properly and the PSU itself should shut
down if properly designed. It will never get to the point
of any motherboard monitors resetting the computer nor bios
code activating anything... and these kinds of bios
interventions are typically from overheating, not voltage
deviations.

Likewise there's usually some provision to "crowbar" the output to
hold the logic supplies within limits - won't allow them to go too
high to protect the motherboard.

As a general rule, too high a voltage can cause damage, but low
voltage is only bad for induction motors and some poorly designed
audio amplifiers. . . . . as a general rule.

If the PSU does not respond correctly, all by itself to the
AC power problem (correctly also meaning to shut off if
appropriate), the PSU should be removed from the system now,
not waiting for that problematic event to occur.

 

[kony]

I wonder how many power supplies have over voltage protection. I bet a lot
of the cheaper supplies are stripped to the bone.

They're all supposed to, but then they're supposed to be
able to output what's on the label, too, and we know that is
not quite true.

 

[default]

I've had the 5v regulator burn high (12v or more) rather than burn open
dropping the output to 0v, taking every single drive and board with it.
Basically I had a metal case to use. A vendor I told this to didn't
sound surprised as he'd had this happen two or three times in the year
previous!

Well that would be a problem all right, but we were talking about
momentary mains voltage droop do to an air conditioner turning on.
Mains voltage droop is unlikely to cause the 5 volt regulator to short
out (the problem you had - outputting 12 volts).

Well designed supplies take into account the price of the motherboards
they power. There should be an OVP (over-voltage protection) circuit
on the logic supply.

Many/most computer supplies only regulate the logic to a precise value
and depend on the others to stay in regulation solely by virtue of the
transformer turns ratios - the 12 V supply can't go over voltage
because there is no excess voltage there. The logic supply has active
components to regulate the output.

The OVP circuit triggers a thyristor (SCR) into conduction clamping
the 5 volts to zero (actually below a volt), if the regulator is
shorted the OVP may cause the fuse to blow, protecting the
motherboard.

If the OVP is triggered by a transient that puts a high voltage spike
on the logic supply, the inherent current limiting protection protects
the regulator and shuts down the drive to the oscillator. That's what
was happening when I'd turn the speaker wall wart "off" on my system -
the 5 V supply would crowbar because of the transient on the 120 mains
and take all the supplies down with it - resetting the computer but
protecting the motherboard. To reset it I had to turn off the power
mains for a few seconds then power back up. Reset switch has no
effect in that case.

To cure it I put a 100 ohm resistor in series with a .05 microfarad
cap across the switch. That absorbs the power that would normally
create an arc that causes the HV transient. A ferrite bead on the
supply line was redundant but should keep any fast rise time
transients off the line.

 

[curious]

Please adjust your newsreader to use shorter line lengths.
80 characters is often considered the max but lower tends to
be more readable. That may result in more replies, too.

Yes the drop can effect a computer, but on most (better) PSU
several stages of capacitor filters will buffer it enough.
you might want to run some stress tests on CPU, memory,
video (if you're a gamer) to verify stability while the
heavy applicances like an AC unit are power cycling.

You might also consider a separate circuit for the AC, or
the computer, or even an online UPS (with line conditioner
built-in) if it continues to concern you.

Thanks for the replies, everyone. I'm getting this Belkin UPS:

http://catalog.belkin.com/IWCatProd...t_Id=&Section_Id=77&pcount=&Product_Id=186802

Hope it works.


Now regarding the fact that I am using "quoted printable" which results in long line lengths... the author of my newsreader (Colin Wilson) actually recommended that we use the "quoted printable" (or "flowed") format. Here's a message he posted to a thread who's subject was the splitting of really long URLs:


"Everyone seems to be confusing the length of 'sent' lines and the
length of lines that their newsreaders display.

Wrapping lines - to 72 characters or whatever before sending them
should be completey unnecessary. However, because the NNTP protocol
restricts lines to around 990 characters, you've got to break them
somewhere.

The point of 'format = flowed' and 'quoted printable' is so that modern
clients can re-assemble these lines so that they are as long as the
author originally intended. Using '72' as the break point is just a
convenience in case anyone is using an old, teletype-era news client
(is anyone??). But every newsreader in the world these days will
correctly reassemble the lines.

So you should use format=flowed or format=quoted printable. The lines
in your messages will be as long as you intended. URLs won't break in
the middle, and everyone will be happy!

But of course some people don't like reading great long lines. They
prefer them to wrap, so that they don't need to move their eyeballs so
much. That's their decision. You - as the sender of the message -
shouldn't concern yourself with this. Some people might like their
messages to wrap at '40'. Some at '72'. Others at the margin of the
text display. You can't cater for them all.

So it's up to the *reader* of the messages - not the *sender* to wrap
them if that's what they want to do. You can configure XanaNews to do
this in Tools/Options/Message Pane, 'Wrap lines at 'n' characters'."

 

[kony]

So it's up to the *reader* of the messages - not the *sender* to wrap
them if that's what they want to do. You can configure XanaNews to do
this in Tools/Options/Message Pane, 'Wrap lines at 'n' characters'."

No, it is not.

Note the format everyone else uses.

One nut telling you otherwise is not a justification.