hardwire ATX switch

  • Thread starter Thread starter Skeleton Man
  • Start date Start date
S

Skeleton Man

Hi,

What's the best method for hard wiring an ATX switch so the PC turns on after a
power failure ? I tried like a low value capacitor (first ceramic, then
electrolytic), but it didn't work (wouldn't turn on).

The board doesn't have any BIOS options for auto-power on.. (it's a s370, so
it's more than a few years old).

Regards,
Chris
 
Skeleton Man said:
Hi,

What's the best method for hard wiring an ATX switch so the PC turns on
after a
power failure ? I tried like a low value capacitor (first ceramic, then
electrolytic), but it didn't work (wouldn't turn on).

The board doesn't have any BIOS options for auto-power on.. (it's a s370,
so
it's more than a few years old).
Click to expand...

There's a simple way to do it, you cut one of the wires that goes to the
motherboard and solder it to the one next to it. I've done it with a machine
at home and it worked very well but I can't remember which wires. If you
don't get a reply by tonight I can have a look.

Michael
 
There's a simple way to do it, you cut one of the wires that goes to the
motherboard and solder it to the one next to it. I've done it with a machine
at home and it worked very well but I can't remember which wires. If you
don't get a reply by tonight I can have a look.
Click to expand...

Green wire (#PS ON) to black wire (ground).. I never thought of that ! Thanks !

Something that still allowed me to soft-off the machine if I needed to would be
better, but as it will be running 24/7 anyway, I guess I can get away with the
above.

Regards,
Chris
 
Green wire (#PS ON) to black wire (ground).. I never thought of that !
Thanks !
Click to expand...


Better have it on a UPS if you plan on this route. A brown out or several
quick blackouts can make short work of your PS.
--


"I don't cheat to survive. I cheat to LIVE!!"
- Alceryes
 
Green wire (#PS ON) to black wire (ground).. I never thought of that !
Better have it on a UPS if you plan on this route. A brown out or several
quick blackouts can make short work of your PS.
Click to expand...

How so ? My primary PC runs 24/7 and sometimes during a storm there will be 3 or
4 brownouts in the space of an hour.. It's coped with them fine.. so what
difference does it make hard wiring a PSU to accomplish the same functionality ?

Regards,
Chris
 
Skeleton Man said:
Green wire (#PS ON) to black wire (ground).. I never thought of that !
Thanks !
Click to expand...

That's it, it was green and black but one of them didn't go to the
motherboard, the green one I think.

Michael
 
Green wire (#PS ON) to black wire (ground).. I never thought of that ! Thanks !

Something that still allowed me to soft-off the machine if I needed to would be
better, but as it will be running 24/7 anyway, I guess I can get away with the
above.
Click to expand...


The only way to soft-off a system and have it stay off,
contrasted with one with loss of power coming back on, is if
your bios supports it.

Otherwise, logically there is no difference between it being
off because you told it to shut off, and being off because
power when out. It is likely impossible to solder any wires
to do what you want, it would have to employ a logic
circuit.
 
That's it, it was green and black but one of them didn't go to the
motherboard, the green one I think.

Michael
Click to expand...

The green would necessarily go to the board... black though,
could be any ground wire, any ground point in the system for
that matter.

This is a very low current connection, making another easier
option possibe- just use a slip-on tap splice connector
with the blanking partition cut out. For example,
http://www.ttieurope.com/microsites...ire_to_wire/splices/Splices_files/term2a6.gif
Some hardware stores stock such connectors though the
correct size may or may not be available.
 
kony said:
The only way to soft-off a system and have it stay off,
contrasted with one with loss of power coming back on, is if
your bios supports it.

Otherwise, logically there is no difference between it being
off because you told it to shut off, and being off because
power when out. It is likely impossible to solder any wires
to do what you want, it would have to employ a logic
circuit.
Click to expand...

A fairly simple circuit could do it, something that just flicked the switch
shortly after power was applied.

Michael
 
A fairly simple circuit could do it, something that just flicked the switch
shortly after power was applied.
Click to expand...


yes but the remaining issue is still that it must
distinguish the power-off event too, else it would just turn
the system right back on after it had shut off through user
command.
 
kony said:
yes but the remaining issue is still that it must
distinguish the power-off event too, else it would just turn
the system right back on after it had shut off through user
command.
Click to expand...

No, when it receives power on 1 line it should flick a switch on for 1
second and then off until power is applied again. That way it will only fire
when external power is lost. A simple circuit with a cap and relay should do
it.

Michael
 
No, when it receives power on 1 line it should flick a switch on for 1
second and then off until power is applied again. That way it will only fire
when external power is lost. A simple circuit with a cap and relay should do
it.

Michael
Click to expand...

That does not solve the problem of scenarios where the user
had turned the system off, it had been off when the AC power
was lost, cycled off and on again. In that scenario system
should remain off but will turn on instead using above
circuit.
 
kony said:
That does not solve the problem of scenarios where the user
had turned the system off, it had been off when the AC power
was lost, cycled off and on again. In that scenario system
should remain off but will turn on instead using above
circuit.
Click to expand...

Minor problem, if it's a server it should be on. Although it would be pretty
easy to solve that problem also.

Michael
 
Minor problem, if it's a server it should be on. Although it would be pretty
easy to solve that problem also.

Michael
Click to expand...

Sure but if you take that "should be on" argument, then the
entire circuit was pointless as it can simply be set to stay
on.
 
kony said:
Sure but if you take that "should be on" argument, then the
entire circuit was pointless as it can simply be set to stay
on.
Click to expand...

Not really because it will only come on if the power is lost, so you can
still power it off.

Also, it would be possible to build a circuit that got around this. You're
statement that this could only be done in bios is not correct, it would be
possible, you just need some way for the circuit to remember whether it was
turned off by the user or a power loss.

Michael
 
Not really because it will only come on if the power is lost, so you can
still power it off.
Click to expand...

True, but, the whole issue in the first place was that it is
to combat power loss from AC, so the very environment being
discounted is the same that was targeted for the circuit...
the AC power loss.
Also, it would be possible to build a circuit that got around this. You're
statement that this could only be done in bios is not correct, it would be
possible, you just need some way for the circuit to remember whether it was
turned off by the user or a power loss.
Click to expand...

In some kind of remote, unrealistic,
spare-no-time-nor-expense sort of way, yes you're right.

In a realistic, "quickly get the job done cheaply because
the whole point was to save $ and/or time reusing old
computer equipment", no it's not reasonably possible.
 
kony said:
In some kind of remote, unrealistic,
spare-no-time-nor-expense sort of way, yes you're right.

In a realistic, "quickly get the job done cheaply because
the whole point was to save $ and/or time reusing old
computer equipment", no it's not reasonably possible.
Click to expand...

Not true at all. A fairly simple circuit could easily achieve this with
AU$10 of parts and I'm sure someone with greater knowledge than me could
make a much simpler circuit.

Michael
 
Not true at all. A fairly simple circuit could easily achieve this with
AU$10 of parts and I'm sure someone with greater knowledge than me could
make a much simpler circuit.
Click to expand...


Relatively speaking, "maybe" AU$10 in parts and the time to
sketch the circuit, it's layout, order and receive parts if
they're not all in a junk bin, protoboard it (or whatever,
that just being the cheapest or quickest), test it and (if
not more professionally done) and get it integrated into the
system is substantial for an old board. The idea of "quick
and cheap" often turns out to an underestimate on both
counts, and has to be put into perspective of use on an old
board that may not be worth over 10 bucks.
 
kony said:
Relatively speaking, "maybe" AU$10 in parts and the time to
sketch the circuit, it's layout, order and receive parts if
they're not all in a junk bin, protoboard it (or whatever,
that just being the cheapest or quickest), test it and (if
not more professionally done) and get it integrated into the
system is substantial for an old board. The idea of "quick
and cheap" often turns out to an underestimate on both
counts, and has to be put into perspective of use on an old
board that may not be worth over 10 bucks.
Click to expand...

Give it up kony, you said it would not be possible which was a pretty silly
statement. This is a simple circuit and could be designed and created on
breadboard in very short time. The cost of the existing board is not
relevant, it's the cost of a new board/cpu/ram and I think I could create
this circuit in under the time it takes to install windows.

Michael
 
Give it up kony, you said it would not be possible which was a pretty silly
statement.
Click to expand...

Not at all, it was a realistic statement and you use the
term "impossible" only to further your arguement. What I
actually wrote was:

"The only way to soft-off a system and have it stay off,
contrasted with one with loss of power coming back on, is if
your bios supports it."

This was within the context of the system, not the system
PLUS homemade active circuitry added. I don't know about
you, but I find it rather foolish to suggest home-made
active circuits unless someone had expressedly asked about
them.

If we want to start talking about conceptually simple
circuits, then when someone asks something like "is my 200W
power supply enough", then someone could just respond "sure,
just swap in these simple parts changes to your power supply
like the transformer, diodes, inductors and caps". Simple
enough yes?

Someone posts that their monitor tube has gotten dim after 5
years. All they'd have to do is just unscrew it and swap
the tube. Simple enough, yes?

The list of "simple" things goes on, but it's it funny how
we don't see these things mentioned very often? Not really,
because something conceptually simple may end up taking more
time than to write about it, and that time and cost may
exceed the value of the equipment.
This is a simple circuit and could be designed and created on
breadboard in very short time. The cost of the existing board is not
relevant, it's the cost of a new board/cpu/ram and I think I could create
this circuit in under the time it takes to install windows.
Click to expand...

Actually, I have a feeling you don't do many of these
"simple circuits" at all, because if you did, you would
realize that a conceptually simple circuit still requires
wading through parts lists, ordering, receiving, sketching
the circuit or being naturally adept at layout-on-the-fly,
constructing it, testing and interfacing it. This is all
for an old board that may be worth $10 and may or may not
have much usable lifespan remaining.

Unless you've been regularly perusing the parts lists at an
electronics house recently, or out of random luck you happen
to have them all in a bin in front of you, odds are good
that I'd have windows installed in same amount of time it
took to _actually_ order, receive and unpack the parts...
not even build anything with them IF you had had a circuit
in mind and the method of attachment, which so far it
appears you don't.
 
Back
Top