Safe Power Supply Tinkering

  • Thread starter Thread starter Ewan Sinclair
  • Start date Start date
E

Ewan Sinclair

My power supply has just started to make ominous grinding noises, so it
looks like I'll want to whip out the fan and replace it before the thing
melts on me. However, I have to admit that I'm terrified of opening the it,
as I gather you can get quite a nasty (potentially fatal) shock if you don't
know what you're doing.

What precautions should I take before opening it up? Please err on the side
of paranoia in your advice; I only need to do this once and I *really* don't
want to go to hospital over it :-)

Ewan
 
Ewan said:
My power supply has just started to make ominous grinding noises, so it
looks like I'll want to whip out the fan and replace it before the thing
melts on me. However, I have to admit that I'm terrified of opening the it,
as I gather you can get quite a nasty (potentially fatal) shock if you don't
know what you're doing.

What precautions should I take before opening it up? Please err on the side
of paranoia in your advice; I only need to do this once and I *really* don't
want to go to hospital over it :-)

Just leave it unplugged over night. That is more than plenty of time.
 
Before you open it make sure you earth yourself by touching the side of the
case.

Steve Gill
 
What precautions should I take before opening it up? Please err on the side
of paranoia in your advice; I only need to do this once and I *really* don't
want to go to hospital over it :-)

There are two capacitors in the PS that are charged to about 320 V DC (in
series).
These each have "bleeder" rsistors, to insure that they will discharge in a few
minuites. Actually, about five minutes should be enough for them to discharge
to a very safe value. If you're really paranoid, unplug it overnight.

Be careful you don't burn yourself if soldering the fan wires!

Virg Wall
 
My power supply has just started to make ominous grinding noises, so it
looks like I'll want to whip out the fan and replace it before the thing
melts on me. However, I have to admit that I'm terrified of opening the it,
as I gather you can get quite a nasty (potentially fatal) shock if you don't
know what you're doing.

What precautions should I take before opening it up? Please err on the side
of paranoia in your advice; I only need to do this once and I *really* don't
want to go to hospital over it :-)

Ewan

After posting "misc." advice myself and still wondering exactly how
long someone should wait till opening their power supply, I have now
begun this test with all power supplies I open, again repowering them
and watching the voltage drop...

The voltage on every PC power supply I've tested so far, about 10 so
far, is below 50V in less than 60 seconds. 50V from a UNPLUGGED power
supply is not enough to do more than suprise you, and by the time
you've unscrewed and unplugged one installed in a PC, the voltage is
too low to even produce a spark.

Granted, there might be some really really badly designed power supply
out there that have no bleeder resistors, at some point in the furture
I will disconnect those and retest a power supply with atypically
large primary capacitors to see just how long it takes for them to
drain, but of course it would be good to get a general consensus on
the exactly voltage level we can consider "safe" for someone looking
to grab hold of a capactor and see what happens. Even so, an ATX
power supply will still power the control circuit and 5VSB as long as
it can before voltage drop shuts it down, so there is still a
safeguard to drain the power quickly.... A power supply doesn't
actually "know" it's plugged into AC power or not, it works until it
can't maintain the voltage then shuts off, after draining most of it's
capacitor-stored energy.

To more directly answer your question, there is very little danger
after a few minues.


Dave
 
There are two capacitors in the PS that are charged to about 320 V DC (in
series).
These each have "bleeder" rsistors, to insure that they will discharge in a few
minuites. Actually, about five minutes should be enough for them to discharge
to a very safe value. If you're really paranoid, unplug it overnight.

Does it discharge best when unplugged, or does it like to have a ground wire
to bleed down? if so, should I leave the little switch on the back on, or
off?
Be careful you don't burn yourself if soldering the fan wires!

Heh, thanks for the warning :-)

Ewan
 
Ewan said:
Does it discharge best when unplugged, or does it like to have a ground wire
to bleed down? if so, should I leave the little switch on the back on, or
off?


Heh, thanks for the warning :-)

Ewan

See kony's posting for some actual numbers. The bleeder resistors are
directly across each capacitor. It matters not what you do externally.
By the time you get the PS out of the case, and the cover removed, the
voltage will be low enough.

Even with open bleeder resistors, electrolytic capacitors have an internal
shunt resistance that would not let them hold a charge for "days" as the
rumors go. The most dangerous voltage in most electronic equipment is
the line voltage. In the old AT power supplies, it was brought right
up front in the computer case. Even before you've unpluged an ATX PS
there are no dangerous voltages in the case.

Be careful of sharp edges!;-)

Virg Wall
 
Ewan Sinclair said:
My power supply has just started to make ominous grinding noises, so it
looks like I'll want to whip out the fan and replace it before the thing
melts on me. However, I have to admit that I'm terrified of opening the it,
as I gather you can get quite a nasty (potentially fatal) shock if you don't
know what you're doing.

What precautions should I take before opening it up? Please err on the side
of paranoia in your advice; I only need to do this once and I *really* don't
want to go to hospital over it :-)

Ewan

Is this a quality psu, if not junk it and buy a good one, Antec solution
series (smartpower) is a good start.

http://www.antec-inc.com/pro_p_solutionpw.html

Lane
 
Well, I replaced my fan and I'm not dead or injured, so thanks for the
advice everyone :-)

Just in case I'll be doing this again though.....
Would shorting one of the positive leads to ground on the +5, 3.3 or 12v
pins through a resistor guarantee an empty pair of caps on the offchance
that the PSU was so cheap that they didn't put bleeder resistors on?

Ewan
 
Ewan said:
Well, I replaced my fan and I'm not dead or injured, so thanks for the
advice everyone :-)

Just in case I'll be doing this again though.....
Would shorting one of the positive leads to ground on the +5, 3.3 or 12v
pins through a resistor guarantee an empty pair of caps on the offchance
that the PSU was so cheap that they didn't put bleeder resistors on?

No.
 
Ewan said:
Well, I replaced my fan and I'm not dead or injured, so thanks for the
advice everyone :-)

Just in case I'll be doing this again though.....
Would shorting one of the positive leads to ground on the +5, 3.3 or 12v
pins through a resistor guarantee an empty pair of caps on the offchance
that the PSU was so cheap that they didn't put bleeder resistors on?

Nope. The capacitors are on the high voltage side of the PS. This voltage
is switched by transistors at a high frequency and the resulting waveform is
applied to a step-down transformer with windings to supply the low voltages.

These outputs do have filter capacitors, but they are charged only to the
voltages they filter. You can spot the HV capacitors as the two large,
cylindrical objects at the input of the PS.

Bleeder resistors are also needed to equalize the voltage across the two
capacitors, so it's unlikely they would be ommitted.

Virg Wall
 
Back
Top