one last easy power supply question

  • Thread starter Thread starter Matt
  • Start date Start date
Powersupplies do not take up much room. Keep it. You may be surprised what
you'll use the parts for later on.

I've fixed a few of them. They are simple electronically and are inexpensive
to repair.


Ed Cregger
 
Matt said:
My old dead power supply was modular... is there any benefit to keeping
the cords?
Click to expand...

Not really. Unless you intend to buy an exact replacement and the original
didn't have enough of a particular type of cord. -Dave
 
Powersupplies do not take up much room. Keep it. You may be surprised what
you'll use the parts for later on.

I've fixed a few of them. They are simple electronically and are inexpensive
to repair.


Ed Cregger
Click to expand...

Ed: A question for you, if you don't mind me nibbing in:

A friend who knows little about computers had a computer that went
south. I took it home and found no power anywhere on the MBO or on
the PS connectors. I took the PS out and took voltage measurements,
tried grounding the green wire to simulate the "On" button on the
computer. Nothing - no fan no heat, no volts.

I bought a replacement supply and before installing it I used the
voltmeter, and essentially had the same thing: Nothing.

I'm not lucky (or unlucky) enough to have a bad power supply on my
hands and walk into a store and buy a new on off the rack and find
it's bad, too. No, I'm doing something wrong.

I even put the new PS in the bad 'puter, just in case the gods are
toying with me. Nothing. The computer's a cheap eMachine and the
owner is ok with trashing it. But my engineering mind (now quite
dusty at 65 years old) is curious what makes a PS work. Shouldn't the
fan even twirl, or at least something show up on my DVM.

With either PS in the computer there isn't even any POST beeps. None.
And there's no voltage anywhere on the board. I even checked for a
faulty power button.

In my younger years I heard an expression that if nothing lights up,
you're better off that having something not working right. If
nothing's working the solution is usually simple - like a power
button.

What thinks you? I haven't returned it to the fellow, but if you have
a trick up your sleeve, I could resescuate the compupter.

Thanks,
Mike Wardreau - Ohio
 
Michael Wardreau said:
Ed: A question for you, if you don't mind me nibbing in:

A friend who knows little about computers had a computer that went
south. I took it home and found no power anywhere on the MBO or on
the PS connectors. I took the PS out and took voltage measurements,
tried grounding the green wire to simulate the "On" button on the
computer. Nothing - no fan no heat, no volts.

I bought a replacement supply and before installing it I used the
voltmeter, and essentially had the same thing: Nothing.

I'm not lucky (or unlucky) enough to have a bad power supply on my
hands and walk into a store and buy a new on off the rack and find
it's bad, too. No, I'm doing something wrong.

I even put the new PS in the bad 'puter, just in case the gods are
toying with me. Nothing. The computer's a cheap eMachine and the
owner is ok with trashing it. But my engineering mind (now quite
dusty at 65 years old) is curious what makes a PS work. Shouldn't the
fan even twirl, or at least something show up on my DVM.

With either PS in the computer there isn't even any POST beeps. None.
And there's no voltage anywhere on the board. I even checked for a
faulty power button.

In my younger years I heard an expression that if nothing lights up,
you're better off that having something not working right. If
nothing's working the solution is usually simple - like a power
button.

What thinks you? I haven't returned it to the fellow, but if you have
a trick up your sleeve, I could resescuate the compupter.

Thanks,
Mike Wardreau - Ohio
Click to expand...

--------------

Hi, Mike. It does, indeed, sound as though you did buy a dead power supply.
It happens. Fortunately, it doesn't happen as often in computers as it does
when buying auto parts store parts. <G>

Remember when all power supplies had a heavy transformer? Things were much
simpler to troubleshoot in those days. Today's switching power supplies are
a bit trickier to troubleshoot, but there have been/are articles on the web
explaining their detailed operation and repair. I would do a search for such
an article. The ham radio forums can usually lead you to links that are
useful for troubleshooting and repair. With the high current demand of ham
radio, switching power supplies were a blessing and have been adopted quite
well to that application. Computer folks tend to be more of the pull/replace
mindset. Like you, this runs counter to my nature.

I imagine that the switching power supply is using an SCR "switch" somewhere
in the chain to engage (turn on) the power supply. So far, I haven't found
one of those, or comparable devices, to be bad in my limited experience with
computer switching power supplies. I have found a few smoked toroid
transformers. The damage/failure point was obvious since the toroid (donut
shaped ring) was blackened. Have you had the case off of the older PS? It
might just be a bad solder joint that was finally narrow enough to burn out
under load.

If the device that is functioning like an SCR (silicon controlled
rectifier - solid state solenoid) isn't working, the power supply will not
operate. There will be no voltage at all visible at any output point of the
PS. Again, it is likely that if this device is damaged, there will be
visible signs of overheating, which should make it easy to locate and
replace.

Ed Cregger
 
Michael said:
Ed: A question for you, if you don't mind me nibbing in:

A friend who knows little about computers had a computer that went
south. I took it home and found no power anywhere on the MBO or on
the PS connectors. I took the PS out and took voltage measurements,
tried grounding the green wire to simulate the "On" button on the
computer. Nothing - no fan no heat, no volts.

I bought a replacement supply and before installing it I used the
voltmeter, and essentially had the same thing: Nothing.

I'm not lucky (or unlucky) enough to have a bad power supply on my
hands and walk into a store and buy a new on off the rack and find
it's bad, too. No, I'm doing something wrong.

I even put the new PS in the bad 'puter, just in case the gods are
toying with me. Nothing. The computer's a cheap eMachine and the
owner is ok with trashing it. But my engineering mind (now quite
dusty at 65 years old) is curious what makes a PS work. Shouldn't the
fan even twirl, or at least something show up on my DVM.

With either PS in the computer there isn't even any POST beeps. None.
And there's no voltage anywhere on the board. I even checked for a
faulty power button.

In my younger years I heard an expression that if nothing lights up,
you're better off that having something not working right. If
nothing's working the solution is usually simple - like a power
button.

What thinks you? I haven't returned it to the fellow, but if you have
a trick up your sleeve, I could resescuate the compupter.

Thanks,
Mike Wardreau - Ohio
Click to expand...

The power supply consists of two parts. A standby supply, with +5V output
(check +5VSB lead, to see if it is working). The +5VSB output must be
there, or nothing else will work.

+5VSB will exist, if the supply is plugged in, and the switch on the back
of the supply is in the ON position.

For a sample schematic, this is a nice example. The author of this site,
traced an ATX supply by hand, to make his schematic.

http://www.pavouk.org/hw/en_atxps.html

The main supply is controlled by "soft power". The PS_ON# signal, which
comes from the motherboard, pulls a control signal (+5V TTL logic levels)
low, and that causes the main rails to be powered. The fan starts to run
on the supply in response. You can manually cause this to happen, by
connecting PS_ON# to COM with a paper clip. For this to work, the supply
must have a working +5VSB, because otherwise you would not expect it
to work.

On the computer case, the power switch is a momentary contact device. It
doesn't latch in a position, and is used to give a short pulse when you
push it. The motherboard has some logic chips, which are powered by +5VSB.
They convert the pulse from the power switch, into a steady logic level,
and that is used to drive PS_ON#. The logic used should be open collector,
meaning a transistor pulls the PS_ON# signal to ground level, and otherwise
a pullup inside the power supply, pulls the signal to a logic one (disabled
state). So not only can the transistor drive PS_ON# low, when the PSU main
power connector is connected to the motherboard, but in a pinch, the user
can also connect PS_ON# to COM. (This is because open collector logic is
immune to being grounded - and that is the whole purpose of making PS_ON#
active low. It allows more than one open collector output to be tied together,
or even allows the output to be grounded to make the circuit it is driving,
active.

There are a number of power supply specs, and these will list wire colors and
signal names for you. The first is for supplies that had a -5V output on the
connector. The second has the -5V removed. The third is for newer 24 pin
power supplies.

http://web.archive.org/web/20030424...org/developer/specs/atx/ATX_ATX12V_PS_1_1.pdf

http://www.formfactors.org/developer/specs/atx/ATX12V_1_3dg.pdf

http://www.formfactors.org/developer/specs/ATX12V_PSDG_2_2_public_br2.pdf

A poster on these groups, has mentioned that on some Emachines, when the power
supply fails, it tends to destroy the motherboard as well. So it is always
possible, that connecting a replacement supply, won't remedy the situation,
if the motherboard also got fried. Many supplies now, have overvoltage and
overcurrent protection, so that such failures should be a reduced occurrence.

So the fan can twirl, if you connect PS_ON# to an adjacent COM pin. That
is how the "soft power" features of the computer, make it happen. The fan
will spin, and power will come out, for as long as PS_ON# remains connected
to COM. It is best, if the supply has some minimal load on it, as some
supplies like a bit of current to be drawn, to regulate properly. AFAIK,
ATX supplies can "push" current, but cannot sink it. A little load gives
them something to work against.

HTH,
Paul
 
Michael said:
A friend who knows little about computers had a computer that went
south. I took it home and found no power anywhere on the MBO or on
the PS connectors. I took the PS out and took voltage measurements,
tried grounding the green wire to simulate the "On" button on the
computer. Nothing - no fan no heat, no volts.

I bought a replacement supply and before installing it I used the
voltmeter, and essentially had the same thing: Nothing.

I'm not lucky (or unlucky) enough to have a bad power supply on my
hands and walk into a store and buy a new on off the rack and find
it's bad, too. No, I'm doing something wrong.

I even put the new PS in the bad 'puter, just in case the gods are
toying with me. Nothing. The computer's a cheap eMachine and the
owner is ok with trashing it. But my engineering mind (now quite
dusty at 65 years old) is curious what makes a PS work. Shouldn't the
fan even twirl, or at least something show up on my DVM.

With either PS in the computer there isn't even any POST beeps. None.
And there's no voltage anywhere on the board. I even checked for a
faulty power button.
Click to expand...

All PC supplies have short protection, but in the cheap ones doesn't
kick in until the PS has burned out. I suspect the motherboard's buck
regulator (takes 5V or 12V from the PS and drops it to 1-2V for the
CPU) developed a bad electrolytic, causing a pair of power transistors
to shorting and take out the first PS. Then the second PS either
failed from that or wouldn't start because ot detected the short in
the motherboard. You can test an ATX PS by measuring its purple wire
for 5V, and when the green wire is grounded, all the other voltages
should light up as well. If one is missing, its output diode is bad,
but if they're all 0VDC, then the high voltage section has probably
popped the fuse and shorted its transistors.
 
On Mon, 31 Mar 2008 10:14:48 -0400, "Ed Cregger"
Click to expand...
All three of you have provided some nice tidbits to chew on. I'll
work on them over the next day or several. (It's still cold in the
attic.)
 
--------------

Hi, Mike. It does, indeed, sound as though you did buy a dead power supply.
It happens. Fortunately, it doesn't happen as often in computers as it does
when buying auto parts store parts. <G>

Remember when all power supplies had a heavy transformer? Things were much
simpler to troubleshoot in those days. Today's switching power supplies are
a bit trickier to troubleshoot, but there have been/are articles on the web
explaining their detailed operation and repair. I would do a search for such
an article. The ham radio forums can usually lead you to links that are
useful for troubleshooting and repair. With the high current demand of ham
radio, switching power supplies were a blessing and have been adopted quite
well to that application. Computer folks tend to be more of the pull/replace
mindset. Like you, this runs counter to my nature.

I imagine that the switching power supply is using an SCR "switch" somewhere
in the chain to engage (turn on) the power supply. So far, I haven't found
one of those, or comparable devices, to be bad in my limited experience with
computer switching power supplies. I have found a few smoked toroid
transformers. The damage/failure point was obvious since the toroid (donut
shaped ring) was blackened. Have you had the case off of the older PS? It
might just be a bad solder joint that was finally narrow enough to burn out
under load.

If the device that is functioning like an SCR (silicon controlled
rectifier - solid state solenoid) isn't working, the power supply will not
operate. There will be no voltage at all visible at any output point of the
PS. Again, it is likely that if this device is damaged, there will be
visible signs of overheating, which should make it easy to locate and
replace.

Ed Cregger
Click to expand...
Well, at last I got around to getting the PS out of the attic and onto
the kitchen table. The new power supply has life after all:

3.9VDC on the green and 5.3 VDC on the purple. I grounded out the
green and the fan turned on.

Next step is to get this same set-up for the questionable compputer's
PS. That comes later. My wife and I have to go out for the evening
and by tomorrow I'll have another zillion things on my adgenda.
Retirement is a bitch!!

I'm going to repeat this message to the other fellows who provided
info to me, especially the links to the schematic and the pdf files.

Thanks again, Ed.

Mike
 
The power supply consists of two parts. A standby supply, with +5V output
(check +5VSB lead, to see if it is working). The +5VSB output must be
there, or nothing else will work.

+5VSB will exist, if the supply is plugged in, and the switch on the back
of the supply is in the ON position.

For a sample schematic, this is a nice example. The author of this site,
traced an ATX supply by hand, to make his schematic.

http://www.pavouk.org/hw/en_atxps.html

The main supply is controlled by "soft power". The PS_ON# signal, which
comes from the motherboard, pulls a control signal (+5V TTL logic levels)
low, and that causes the main rails to be powered. The fan starts to run
on the supply in response. You can manually cause this to happen, by
connecting PS_ON# to COM with a paper clip. For this to work, the supply
must have a working +5VSB, because otherwise you would not expect it
to work.

On the computer case, the power switch is a momentary contact device. It
doesn't latch in a position, and is used to give a short pulse when you
push it. The motherboard has some logic chips, which are powered by +5VSB.
They convert the pulse from the power switch, into a steady logic level,
and that is used to drive PS_ON#. The logic used should be open collector,
meaning a transistor pulls the PS_ON# signal to ground level, and otherwise
a pullup inside the power supply, pulls the signal to a logic one (disabled
state). So not only can the transistor drive PS_ON# low, when the PSU main
power connector is connected to the motherboard, but in a pinch, the user
can also connect PS_ON# to COM. (This is because open collector logic is
immune to being grounded - and that is the whole purpose of making PS_ON#
active low. It allows more than one open collector output to be tied together,
or even allows the output to be grounded to make the circuit it is driving,
active.

There are a number of power supply specs, and these will list wire colors and
signal names for you. The first is for supplies that had a -5V output on the
connector. The second has the -5V removed. The third is for newer 24 pin
power supplies.

http://web.archive.org/web/20030424...org/developer/specs/atx/ATX_ATX12V_PS_1_1.pdf

http://www.formfactors.org/developer/specs/atx/ATX12V_1_3dg.pdf

http://www.formfactors.org/developer/specs/ATX12V_PSDG_2_2_public_br2.pdf

A poster on these groups, has mentioned that on some Emachines, when the power
supply fails, it tends to destroy the motherboard as well. So it is always
possible, that connecting a replacement supply, won't remedy the situation,
if the motherboard also got fried. Many supplies now, have overvoltage and
overcurrent protection, so that such failures should be a reduced occurrence.

So the fan can twirl, if you connect PS_ON# to an adjacent COM pin. That
is how the "soft power" features of the computer, make it happen. The fan
will spin, and power will come out, for as long as PS_ON# remains connected
to COM. It is best, if the supply has some minimal load on it, as some
supplies like a bit of current to be drawn, to regulate properly. AFAIK,
ATX supplies can "push" current, but cannot sink it. A little load gives
them something to work against.

HTH,
Paul
Click to expand...
Paul,

Well, at last I got around to getting the PS out of the attic and onto
the kitchen table. The new power supply has life after all:

3.9VDC on the green and 5.3 VDC on the purple. I grounded out the
green and the fan turned on.

Thanks for the links to the schematic (which at my age doesn't mean
much anymore) and the Pdf files.

Next step is to get this same set-up for the questionable compputer's
PS. That comes later. My wife and I have to go out for the evening
and by tomorrow I'll have another zillion things on my adgenda.
Retirement is a bitch!!

I'm going to repeat this message to the other fellows who provided
info to me, especially the links to the schematic and the pdf files.
 
A friend who knows little about computers had a computer that went
south. I took it home and found no power anywhere on the MBO or on
the PS connectors. I took the PS out and took voltage measurements,
tried grounding the green wire to simulate the "On" button on the
computer. Nothing - no fan no heat, no volts.

I bought a replacement supply and before installing it I used the
voltmeter, and essentially had the same thing: Nothing.

I'm not lucky (or unlucky) enough to have a bad power supply on my
hands and walk into a store and buy a new on off the rack and find
it's bad, too. No, I'm doing something wrong.

I even put the new PS in the bad 'puter, just in case the gods are
toying with me. Nothing. The computer's a cheap eMachine and the
owner is ok with trashing it. But my engineering mind (now quite
dusty at 65 years old) is curious what makes a PS work. Shouldn't the
fan even twirl, or at least something show up on my DVM.

With either PS in the computer there isn't even any POST beeps. None.
And there's no voltage anywhere on the board. I even checked for a
faulty power button.
Click to expand...

All PC supplies have short protection, but in the cheap ones doesn't
kick in until the PS has burned out. I suspect the motherboard's buck
regulator (takes 5V or 12V from the PS and drops it to 1-2V for the
CPU) developed a bad electrolytic, causing a pair of power transistors
to shorting and take out the first PS. Then the second PS either
failed from that or wouldn't start because ot detected the short in
the motherboard. You can test an ATX PS by measuring its purple wire
for 5V, and when the green wire is grounded, all the other voltages
should light up as well. If one is missing, its output diode is bad,
but if they're all 0VDC, then the high voltage section has probably
popped the fuse and shorted its transistors.[/QUOTE]

Well, at last I got around to getting the NEW PS out of the attic and
onto the kitchen table. The new power supply has life after all:

3.9VDC on the green and 5.3 VDC on the purple. I grounded out the
green and the fan turned on.

Next step is to get this same set-up for the questionable compputer's
PS. That comes later. My wife and I have to go out for the evening
and by tomorrow I'll have another zillion things on my adgenda.
Retirement is a bitch!!

I'm going to repeat this message to the other fellows who provided
info to me, especially the links to the schematic and the pdf files.

Mike
 
Back
Top