ON/OFF Switch for a Computer?

Hello,

This may be a newbie question, but i'd like some schematics for the ON/OFF switch of your standard computer. In other words, a solid-state switch that will allow the computer to do a soft OFF procedure, where the computer boots down before turning off.

Ideally, it would be 3-4 seconds push-to-hold to start the power-ON and boot-up sequence. And

3-4 seconds push-to-hold to bring up a "Do you really want to shutdown? YES/NO" prompt on the LCD screen, and after touching "YES" on the screen, the computer will boot down the OS, and shut down.

However, if the boot down somehow hangs up, i'd like to retain the option of a "hard-turn-OFF", like almost all computers now have, where the ON/OFF button can be depressed for something like 5 seconds, and the computer will shut down immediately (no "soft" OFF boot down).

If anyone out there has a schematic, that would be great. Thanks.

Reply to
Paul
Loading thread data ...

It's all done in software....

Reply to
TT_Man

formatting link

Easy - there you go.

--
Cheerz - Brownz
\'89 K100RS
 Click to see the full signature
Reply to
Brownz (Mobile)

I don't think the hard-OFF is. That has to work even if the CPU has stopped. Probably a big RC or digital time constant that bypasses everything. Tam

Reply to
Tam

It didn't on my computer for a while. So "has to work" is not quite true. For a while, the only hard-off option I had was to remove the power cable (there was no master on/off switch on the back, either).

Reply to
DJ Delorie

The ON/OFF switch is a simple, single pole MCNO, (momentary contact, normally open) switch. When pressed for a short time, this grounds the "power on" wire from the power supply. The mother-board circuitry holds this ground and the machine starts up. If the power button is held in for 2-4 seconds, the mother-board removes the ground and the power supply is turned off with the exception of 5V stand-by power. This is left on to allow the power supply to be re-started in the same manner.

A "soft-off" requires that at least part of the computer be working properly. In this case the normal software shut down is used, and the computer flushes buffers, closes files, etc., before the power is removed.

--
Virg Wall
Reply to
VWWall

Indeed it is. On my laptop sometimes the power button doesn't do anything because as you said it's software controlled, and rather buggy software under Windows.

This necessitates the occasional pulling of the power cord and yanking out the battery in order to perform a true cold boot.

Reply to
T

The momentary push button that turns the computer on and off can be overridden in software as can the delay to turn off. More often found in the BIOS settings as "Instant Off" and "Delayed 4 Seconds".

Google for "ACPI & ACP"

--
Best Regards:
                     Baron.
Reply to
Baron

Yes, I've observed that on most machines I've worked on... If there is some major problem such as an OS lockup, the power button will induce an orderly shutdown.

That is preferable to hitting the reset switch which will result in a "dirty" shutdown.

As a matter of fact, on most of the machines I setup... I generally do not even connect the "reset" switch

Reply to
philo

In my case, it was set to "Delayed 4 Seconds" - but it never worked. So at least in that machine, it wasn't something that relied on hardware that the BIOS set up, it also relied on the operating system to cooperate.

Plus, when it did work, it started a graceful shutdown of the OS, it didn't just turn the PC off. So, the button is a software off, not a hardware off.

Reply to
DJ Delorie

uestion, but i'd like

s push-to-hold

ow hangs up,

atic,

d.

True. The hard-OFF is separate from the software, completely.

Reply to
Paul

question, but i'd like

ds push-to-hold

how hangs up,

matic,

However, in my case, the batteries are NOT easily accessible by the user, so i will need a hard-OFF circuit that is completely separate from the CPU.

Most likely, i will use a PIC chip like this:

formatting link

This has multiple general purpose IO's, and i'll probably hook 2 of them to the CPU: one so the switch will know when the system is already ON, and the other for the "intialize soft-OFF" signal to the CPU, as well as for the final signal back to the PIC, to command the opening of the main MOSFETs (p-channel feeding an n-channel, most likely).

It appears most PICs have a sleep mode, which will be good for battery-life, as the PIC will be constantly waiting to be turned on, even when the system is off. Also, the 3-4 second delay can be handled with this IC.

Reply to
Paul

estion, but i'd like

push-to-hold

w hangs up,

tic,

I think you need to go back to the motorcycle group....

Reply to
Paul

I don't have or know where a schematic might be found but a basic description can be found in some Super I/O chip datasheets (look at enough of them and you might find a better description than the following but as an example...).

formatting link
Section 9.4, pg 70

Reply to
kony

everyone else uses"?

ics of just the switch

Unfortunately, motherboard schematics are all proprietary, so fat chance that i'll find anything on the web.

I may have to reverse-engineer an old motherboard....

Reply to
Paul

If you want a standalone solution that doesn't rely on software at all, the LTC2953 may do what you want:

formatting link

Currently evaluating this one for a soft power switch to do exactly what you describe, seems OK so far

Reply to
news

I can see you have taken some flak over this (mainly I think from people who have not tried to do it.) I have recently completed a job sorting out such a system which had never worked very well but is now fairly bulletproof.

There is a little board with an AVR which intercepts the signal from the PC on/off button. The AVR can communicate with the PC via 3 signal lnies but the actions available to us on these are constrained. On start up we run an application which signals to the AVR that it is working OK and looks out for the AVR requesting shutdown or for the UPS detecting mains failure. If the user presses the off button they get the dialogue box asking if they really mean it. If the mains fails they get a warning that the PC is going to shutdown. Then the PC app does its best to close everything (not easy in XP - things don't always shut down just because you send them the shutdown message !!).

Once the AVR is happy that everything is shutdown it switches off the PC.

There is a lot of detail missing from the above and I can't give away my customer's paid for ciruits etc.

Your job will be different any way - but expect it to tkae some serious effort to get it all sorted.

Good luck.

Michael Kellett

formatting link

Reply to
MK

I highly doubt it.

If the O/S is not responsive, your holding down the power button is probably just going to turn it off. No clean shut down, no syncing and unmounting disks and marking them as such, just out "like a light"

Reply to
cs_posting

If you had the datasheet for the Super I/O chip on the respective board, then using the power control pins you could trace that to another pad or lift the pin and air-wire to a separate control board - not needing the rest of the circuit on the motherboard on the input to the Super I/O chip. If you have the register data you then have something to program. I'm not suggesting this 100% resolves everything but is a place to start, and as you wrote, forcusing more on the specific board you need to control.

Reply to
kony

Wow! That's the closest thing anyone has posted yet!

Looks like a strong contender, actually.

This might be easier than having to program our PIC chips. I have ENABLE lines on both of our +3.3V regulators, so we should be able to use this.

This also has a "Forced Power-Off Sequence", for if the CPU hangs up during the soft-OFF bootdown.

Only one problem:

I'm not sure if we can program the 3-5 second delay with this IC.

Reply to
Paul

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.