Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
×
Hardware

Ask Slashdot: Reliable Powering of ATX Systems? 173

Posted by Cliff from the coping-with-soft-power-switches dept.
David Muir Sharnoff writes in with this question: "ATX power supplies don't supply power unless the motherboard says to turn on. Many motherboards do not have a setting of: 'supply power all the time.' This means that ATX systems cannot safely be used a servers. Anyone know of an ATX power supply that ignores signal? " More appropriately: does anyone know of an ATX motherboard/power supply/case combination where powering is similar to AT systems?
This discussion has been archived. No new comments can be posted.

Ask Slashdot: Reliable Powering of ATX Systems? More

Ask Slashdot: Reliable Powering of ATX Systems?

Comments Filter:
  • The SuperMicro motherboards are nice from
    this point of view.

    BUT, they have a stupid BIOS that will not
    boot without requiring you have a keyboard
    in place!!! PLEASE - all BIOS manufacturers -
    HAVE AN OPTION IN THE BIOS FOR CHECKING
    KEYBOARDS!

    Grrr ... sorry, it's been a long day :-)

  • OK, how about off? (Score:1)

    by Anonymous Coward
    How about turning them OFF? I have seen new machines that can't be turned off without unplugging.

  • Re:Use a UPS !!!! (Score:1)

    by Anonymous Coward
    Sometimes it is useful to remotely restart a
    crashed server by cycling its power. Like in
    colocation situations.

  • ATX: change the bios (Score:2)

    by Anonymous Coward
    I had the same "problem", my atx mobo at
    home would not bootup after a powershutdown.
    After some investigation I found that most
    ATX mobo's have a setting in their bios to
    powerup after a powercycle.


    -aj-

  • Silly question, but, what does ATX mean? (Score:2)

    by Anonymous Coward
    I thought ATX was a PC case description (I see it in ads). Now it seems its a power supply and what does this have to do with being able to be a server? This proto nerd is confused.
  • ATX is more of a Mother Board Description.

    It includes:
    Physical card dimensions(HxWxD, PCI/ISA connector position, CPU position).
    Power supply connection and function (pins, volts, amps, power on).
    Cooling (P/S blows into case).

  • have just done this (Score:2)

    by Anonymous Coward
    I had a motherboard (DEC AlphaPC 164) that didn't produce the DC-OK (aka PC-ON)signal. This is pin 14 on the ATX motherboard connector. You can cut this wire (usually green colored) and the adjacent black (pin 15) wire above the connector and wire them together with a wire nut so that it is always on whenever the power cord is plugged in. Or connect these two wires to a new switch and replace the ATX style switch on your case with a standard AT-like switch (this is what I did). I bought the switch at radio Shack for under a dollar. Works great. The only thing the DC-OK wire does is tell the power supply to go from sleep state to full power state. It does this when the pin is grounded. The black wire #15 is one of those unneeded ground wires that can be redirected to this task. I was educated to this method by PC Power and Cooling (www.pcpowercool.com) which is probably the finest producer of power supplies out there. Good luck.

  • Related: Soft Power& Reset Switches?? (Score:2)

    by Anonymous Coward
    I hav a related question.

    Is it possible to have the ATX power switch act
    like an SGI power switch(I press the button, I
    runs shutdown then turns off).

    Likewise with reset. If I press reset, can I have it run 'reboot'? Can I even control what happens with the reset button?(I just unplugged the thing, since it does nothing good(safe))

  • WTX will appear *very* shortly. (Score:4)

    by Anonymous Coward on Thursday May 27, 1999 @08:51AM (#1876896)

    WTX will replace ATX for server boxes. It appears to be a good spec to me.

    WTX.ORG [wtx.org]

  • ATX specification URL (Score:5)

    by Anonymous Coward on Thursday May 27, 1999 @08:17AM (#1876897)
    The ATX specification can be found here:

    http://www.teleport.com/~atx/spec/index .htm [teleport.com]

    (Boy I hope this message get moderated up.
    That would be keen.)

  • Many motherboards do not have a setting of: 'supply power all the time.' This means that ATX systems cannot safely be used a servers.

    Not where I come from - where I come from it just means that those motherboards can't be safely used as servers.

    If you get a decent motherboard which has the required power setting, why would you need to play silly buggers with the power supply unit?

  • check under the APM options, if you're doing a make menuconfig it will have an option of 'power off computer on shutdown' or something similar to that. As long as your M/B supports it, the machine does magically turn off on shutdown then.
  • On the Asus mainboard I use there's a setting in the BIOS that will cause the powerswitch to act as a suspend switch. A daemon could watch for suspend requests and run shutdown when it sees 'em.
  • The Asus mainboard (don't have the model number handy; the machine and docs are about a mile away) I use can be configured to switch on whenever
  • The Asus mainboard (don't have the model number handy; the machine and docs are about a mile away) I use can be configured in the BIOS to be 'always on', but it also has a setting so that the power switch must be held down for four seconds in order to switch the machine off. It's a nice feature for servers because it protects you from accidentally pressing the power switch, and if you really do want to turn off the power it's easier than untaping the metal plate that's covering the switch (we really do use little metal plates to protect switches at work).
  • What I want to know is, what does ATX get me that makes it worth having to toss out all of my venerable AT form factor hardware in order to upgrade my machines? And is it going to be possible to buy an AT K7 motherboard?
  • Of all those features, the only one that seems to require changing the form factor of the case is putting the ports directly on the board... and it's not entirely clear to me that that one is a net win, anyway. While fewer cables is a good thing, I think the modularity that the old design provides is a better thing. It's nice to be able to choose 9- or 25-pin serial ports at will, for example.
  • When I say "AT", I mean "AT". :) I love those big old monster boxes. I do much the same thing you do, but I prefer the 286 desktop machines - the original ATs - over any tower I've run across.

    And as far as PC2000 is concerned, M$ can blow me. They can have my ISA slots the day I find a PCI (or AGP! ;) ) monochrome vid card to drive my secondary displays...

  • ASUS P2BF & Intel Nightshade (Score:4)

    by Eric Green ( 627 ) on Thursday May 27, 1999 @07:18AM (#1876906) Homepage
    I know for a fact that the ASUS P2BF has a "AC Power Fail Auto-Restart" BIOS option and the Intel Nightshade has a "Restore Power State" BIOS option that will turn the power on automatically if it was on when AC power was lost (i.e. power was not turned off via the front panel switch).

  • Posted by oNZeNeMo (guns'n ammo):

    most of the time, you can fix this by editing the init scripts, but I can't give you the speciffics without knowing your distro. On my system which is roughly based on 5MB of Slackware 3.5, I had to edit the init scripts to enable power off for the new kernels.

  • Mod to keep power on all of the time (Score:4)

    by gavinhall ( 33 ) on Thursday May 27, 1999 @07:23AM (#1876909)
    Posted by el_steevo:

    From the power supply wires coming from the power supply in the case to the MOBO, short the green wire to ground. This will give you power all of the time.

  • OK, this leans slightly toward the "off topic" category but...

    I have an epox kp6bs board in a supermicro 750 case. Under Windows, the date / time is reflected fine, even when the machine puts itself to sleep. Under linux, however, it appears that date / time stops when the machine goes to sleep. Therefore, while the machie has been running continuously under linux since May 20th, it thinks that the (current) date is May 21st ('cause it only runs the clock when I'm actually using it).

    I don't recall, but I believe that I have power management set to something like 1 hour but suspend completely disabled. If I reboot the machine it picks up the correct date just fine. Anybody got a clue what's doing this to me and how to fix it? :)


    Who am I?
    Why am here?
    Where is the chocolate?

  • ATX corrected several porblems of AT motherboards, but added the stupid soft power switch.

    With the old AT cases it was possibal to mix up the power connectoers (they were 100% identical visually, other then label and wire color, if the manufature bothered to get those factors right) Mix up the connector and at least you need a new motherboard. There were also problmes of screw spacing never quit standardised.

    ATX cases have a different power supply connector.

  • Well, get a recent gs and gv (ghostscript, ghostview)...
  • You should short the jumper that receives the signal from the case's power button.
  • I've got an ASUS P2B-L. It uses the soft-power current to control all the wake-on-signal functions. This includes the space-bar trick you mentioned (which is real neat-o), wake-on-lan, wake-on-serial and the switch on the front. There is a jumper for the wake-on-spacebar because some power supplies can't supply enough standby power to work it. The manual explains this.

    Wade

  • For instance, my Macase ATX case (at home) has a reset button but the Compaq ATX case in front of me now (at work) doesn't. Manufacturers like to cut corners. Skipping a reset button is a cost-saver.

    Oh yeah: sometimes the power-off-after-five-seconds happens when you release the button. With my home PC (ASUS P2B-L), the 4-second power off happens as documented. But with my Compaq, it powers down only after I release the button.

    Wade.

  • No, using NTP (network time protocol) alone won't work because (1) between the time that the system resumes from sleep and the NTP daemon realizes that something is wrong, the clock will be off; (2) the NTP daemon will try to calculate the clock's drift and calculate a correction factor, and that factor will keep changing as the system sleeps for various lengths of time, "skipping" different-sized chunks each time.

    apmd, on the other hand, is wired into the APM system and restores the time from the RTC immediately upon resume. apmd+NTP could work very well together.
  • damn, will they ever learn? [useit.com] the specs are in 'pdf' format.

  • Use the force, luke... er... Lucas Digital? (Score:3)

    by tgd ( 2822 ) on Thursday May 27, 1999 @08:27AM (#1876918)
    This might be related to the Linux "use" at Lucas Digital that caused some discussion last week. Then again it might not and it just made me think of it.

    If I had a rack of Linux servers with ATX supplies (which, coincidentally, I do), and wanted to make sure they came back on when power was lost (doh, didn't think of that, what are you running a server without a UPS for anyway???), then I might use a spare 386 or 486 I've got kicking around to do it.

    Its a piece of cake to wire a relay to a parallel port, there's schematics all over the place on the net to do it. Trim the power supply's "on" pins, wire them to the relay, and you could easy control the power of 8 systems from the parallel port with maybe $40 worth of hardware. Its not exactly what the question was asking, but it might be useful. Maybe Slashdot needs something like this for when the machine misbehaves. ;)

    Also useful in case of an errant init 0 instead of init 6...

    Alternately, I think those Matrix Orbital displays with the keypad interface also have eight digital outputs on them, you could wire one of those to a serial port, and trigger the relays with those outputs. Running some monitoring software to put the status of the machines on the LCD, you've got a nifty monitoring/control system.

  • Re:OK, how about off? (Score:3)

    by tgd ( 2822 ) on Thursday May 27, 1999 @08:36AM (#1876919)
    Virtually all of those will hard power off even if the system is hung if you hold the power switch down for five seconds...

    Sometimes they don't make that clear in the manual though. Its a problem if you have Linux on them and don't know that, and don't compile the kernel to shut the system off on halt.

  • Here's what it will cost to run a computer all month for your electric bill at home if you pay $0.08 per kilowatt hour:

    40 watts * 0.08 cents/1000watthour *24hours * 30 days
    =$2.30

    If you get the industrial rate at $0.02 per kilowatt hour, that will be about $0.58 per box.

    UPS supplies seem to draw much more. They seem to have a type of saturation transformer where the iron core is intentionally saturated with excessive current to give a well regulated and safe voltage to your computer including isolating it from the mains.
  • Perhaps your AB-BH6 is an older revision.
  • This works with an Award BIOS. Press RESET and press Delete key to enter setup. Open the "Power Managmement Settings" menu. Set the "Power Button Override" to "Enable" Save and Exit.
  • Should be. I'm running an AT motherboard. In Windows 95 (I'm at work, but I'm demonstrating that it can be done) on a standard AT clone PC, when you go to Control Panel, then go to Power. "Allow Windows to Manage This System's Power". "Add Suspend to Start Menu". Etc. Under Display, Go to the Screen Saver command, "Energy saving features of Display", etc. All doable under Linux using the APM drivers in the kernel and apmd. Look in the documentation in your motherboard's or bios's documentation under Power Management. This road's been traveled. Just do a search for Power Management.

    Two key things:
    1. Your BIOS. Assuming post 1994 (probably before, just a guess) , you should have power management options in there.
    2. APM and APMD. Might require a kernel recompile to enable APM in Linux, but that's extensively documented. See first post about link to Battery-Power-HOWTO (similar techniques, section 3.x)
  • Look for APM (advanced Power Management) support in the BIOS and Linux. Should be doable. I've never looked into it, but it's the basic principle. Most modern bioses support sleep mode, which is functionally idle-type mode. Of course, you'll want an EnergyStar-certified monitor. Most modern monitors are. This sort of information was really big around six years ago, with major manufactuers shipping low-power systems, such as the IBM EV1(?) which had an LCD monitor (a rarity at the time, and expensive). If you look through PC Magazine there should be some information about Green PC's around 92 or 93. Additionally, I believe there's some info on power managament at: ftp://sunsite.unc.edu/pub/Linux/docs/HOWTO/mini/Ba ttery-Powered

    Particularly Sections 3.1-3.4 (which describes APM)

    Also HDPARM, which can be used to reduce the spin rate on the Hard Drive to reduce the power used.

    But the question is why? Aside from the good for the environment angle, there isn't a pressing need for it. A computer only draws about as much power as an light bulb (last spec I heard about PC power consumption). If you REALLY want to save power, shut off the system. Sleep mode draws ~2-4 watts afaik.

    Hope this helps,
  • You need APM support in the kernel, and you need to run apmd. It'll reset the clock automatically after it resumes. Works well on my laptop. One of the options under APM support, is the autopower off the computer on halt feature.
    I setup control-alt-delete to run shutdown -h, instead of shutdown -r, so I can turn it off just by hitting control-alt-delete.
  • Doesn't the power supply have to be on for the motherboard to do anything?

    In an "Ask Slashdot" that wasn't posted I asked about making my computers power efficient. Is ATX relevant here? Since I'm out of the house for 12 or more hours at a time it would be great to have my computer turn its power supply down to idle when I'm gone.

    Is that what ATX does? I've never come across any comparison of AT vs ATX so I've just stuck with AT upgrading piecewise from my first AT case. If ATX allows variable control of the power supply its worth the switch.

  • Well I do want to save power, so I do shut off the system. It's really a matter of convenience thought. Right now it's a small inconvenience to wait several minutes for the system to boot up if I just want to look up something quickly on the web. If I ever get around to running ethernet down to the kitchen and moving the 486 there it will be a big inconvenience to have to go upstairs and turn on the server in order to use the 486 downstairs as an X-terminal.

    The reason I want to save power is because there's way more idle time than active time. Two computers with 200 Watt power supplies running 24/7 is wasteful when you consider that they're idle 10-12 hour straight when I'm at work and 6-8 hours straight when I'm asleep.

    Sleep mode of 2-4 watts each would be great. Is that achievable with an AT motherboard/power supply?
  • I get power from my old AT even when nothing's connected?!?!?...
  • off topic? Is it possible to tell win{9[5,8],NT}
    not to shut the power on shutdown?
    Can you set this in other Unices(!Linux)?
  • Bzzzt, wrong. This is not for saving money on CPU fans - they are still needed. The reason for blowing air in is that the PSU can filter the air coming in. If it blows air out, like the original PC, then it will suck air in through any other hole in the computer (e.g. through the floppy drive) where there are no filters, and thus clog the system with dust.
  • Care to give me a little more to work with?
    HIT.net maybe? :)
  • We use apcupsd. If the power goes out it calls a little bash script that sends email to my pager. Very simple, actually.
  • Thank you! At least one other person caught that! :)
  • The case has to have the screw holes the correct distance from each other. "ATX case" is a bit of a misnomer in this regard; "case for ATX motherboards" would be more accurate, but also more unwieldy. All you have to worry about is "will this motherboard fit in this case properly?" when buying stuff.
    --
  • While we're on the (admittedly off-topic)subject, I've got a Giga-Byte Socket 5 board with a "Dallas" type RTC chip that keeps time beautifully--until I power up the board. I can set the time in the CMOS setup screen, save the changes, boot to some or no operating system, turn off the system, disconnect the power supply, come back a month later with a different power supply or the same one, different or same keyboard, video card, CPU, etc., and the time shown on the Cmos setup screen will reflect the passage of that month, but any time the board is under power the clock "freezes" (stops counting seconds) for as long as power is applied. I've downloaded and flashed the BIOS to no avail. Anybody got a clue?

  • This is the correct answer-for the "speed" LED's. If it was for fans it would be 12 volt and not 5.
    Those LED displays ,by the way, aren't exactly precision frequency counter readouts since you can set them to read any number between 00 and 199--I've got one set to switch between "HI" and "LO".
    A better use for those LED displays would be to use it as a readout of the temperature inside the case. Anyone got a (cheap, low parts count) circuit handy?


  • The only linear computer supplies I ever saw in surplus catalogs came from discontinued mainframes, not bankrupt PC manufacturers.
    Linear power supplies take the current from the wall socket and run it through the primary winding of a transformer. The secondary winding (in vacuum tube days a voltage step up, current step down--for solid-state circuits a voltage step down, current step up) passes the same wattage (voltage times current, sort of--it's a little more complicated with alternating current than with direct current)--minus what gets lost, due to less than 100% efficiency, and turns into heat--to rectifiers which "re-route" the alternating cycles so that the current flows in the same direction instead of reversing 60 times per second. This makes it direct current, but now the voltage goes from zero up to whatever the peak is and then back down to zero 120 times per second, so various passive (and sometimes active) components are used to filter and regulate the voltage. If the incoming frequency is higher than 60 Hz then smaller, lighter (less expensive) transformers and passive filtering components can be used. This is the reason for 400 Hz generators on aircraft, to save weight and space, although, in the case of aircraft, not money. The higher frequency = smaller, lighter, cheaper components relationship is also taken advantage of in switching supplies where the alternating current from the wall socket goes straight to the rectifiers, is changed into pulsating direct current, filtered to "average out" the pulsations, and sent to transistors which turn on and off somewhere in the neighborhood of 20,000 to 40,000 times per second. This results in a "pulsed" direct current which can be stepped up or down in voltage with a transformer just as with alternating current, but at the much higher frequency a smaller, lighter, cheaper transformer can be used. In both cases (linear and switching)the transformer electrically isolates the power supply's output from the wall socket supplied input.

  • Do it in hardware (Score:5)

    by Booker ( 6173 ) on Thursday May 27, 1999 @07:21AM (#1876939) Homepage
    Pin 14 on the ATX power supply connector is the "PS-ON" signal. From the ATX spec: [intel.com]
    PS-ON is an active low signal that turns on all of the main power rails including 3.3V, 5V, -5V, 12V, and -12V power rails. When this signal is held high by the PC board or left open circuited, outputs of the power rails should not deliver current and should be held at a zero potential with respect to ground. Power should be delivered to the rails only if the PS-ON signal is held at ground potential. This signal should be held at +5VDC by a pull-up resistor internal to the power supply.
    So, you should be able to hack it together to have that pin permanently grounded, so that the PS is always on. I haven't tried this; I think it should work fine, but don't blame me if your motherboard burns! :-)

    There's also more info on ATX at http://www.teleport.com/~atx/ [teleport.com]

  • I have once or twice disconnected my motherboard and run my power supply with no problems (accaching my drive to another machine). AT does not have a "smart" power supply, so it should not know whether it has a motherboard hooked up or not.
  • I run a voicemail server on an ATX system and yes, I do have a UPS attached. But if the power outage is a long one, I still want the system to power back up on restoration (and I can't afford a huge UPS farm!)

    All I'm saying is this information is useful, and I'll be acting on it!



  • I have had poor luck with OEM power supplies that come with brand-named computers. I have also seen lots of cheap ATX power supplies that are rated 140 watts peak. Remember, the rating you're seeing is the peak rating, not the optimal power rating which is usually 20-40% lower than the peak rating. AT power supplies are always cheaper and always offer higher wattage ratings per dollar.

    One major problem that I have experienced wth ATX power supplies is that since the fan is mounted on a horizontal plane it seems to be more prone to wear and "chuffing."

    ATX motherboards are great since they reduce ribbon cable clutter. However, if you must use ATX you should ALWAYS get a motherboard that has AT power supply connectors (most aftermarket boards do). You'll save at least $30 and be able to get higher-rated power supplies for less money. You also won't have the annoying power button problems.

    Kris

    Kriston J. Rehberg
    http://kriston.net/ [kriston.net]

  • Dont simply short it - use a small capacitor! This way you can successfully emulate the button press on most Motherboards.
  • UPSs are just supposed to keep you up and running until your 'tested weekly' backup generator is online. :-)
  • ..and to answer the second part of your question, what does this have to do with being able to be a server? I would presume that it is because the auto-shutdown function of ATX power supplies/motherboards results in the machine being in the 'switched-off' state after a power failure, for example.. and of course a server has to remain 'up' all the time :)

  • It was told here that the ATX power supply looks at pin 14 of the power harness for the control signal and that anything buy pulled low is OFF. It was also told here that the pull up for that pin is located in the power supply and the mobo pulls it low on software or it is physically pulled low with the front switch. I'm just putting 2 and 2 together here though I was a ET for some 6 years before dumping EE for CS degree.

    As far as the power supply idling, what you really mean is the mobo/CPU idling. I think the PS will use less power if it is asked to supply less power and most ATX mobo's have APM built in. IIRC. If you want to use that feature, install APM on your OS and enable in your mobo BIOS. I did this with my IWILL DP6NS server at home til I started crunching RC5-64.
  • True, and in these cases, there is still some sort of a switch that the Motherboard uses to turn on the PS. As others have said, you just need to find out which wire to ground.
    ------------------------------
  • Actually, I have done this myself many times. IMHO, the best way to do it is to find the two-wired connector that connects to the mobo and controls the power. You simply short these two wires, and the PS stays on.
    ------------------------------
  • Yes, but for the time being, I think this is a Win98 Second Edition feature only. It's called ACPI and it's an all encompassing *thing*. Aside from using an unshareable IRQ (#9), it also gives you a soft power switch and another means to provide resource enumeration.

    It is actually a rather nice idea, as it tries to sidestep some of the hackish nature of APM.

    Keep in mind even the shipping version of Win98SE has horribly buggy ACPI support, even though it tries to be very conservative about when to enable this. My guess is W2K will also have some form of ACPI support.

    You can find some more information somewhere on telebit (I think).
  • Every power supply I have ever worked with, other than an actual AT has had tabs on the power connector that prevent you from puting the connectors on backward with out forcing them.
  • The problem is with the way *nix keeps track of time. The date problem happens because *nix systems read the date from hardware when they start only, they then keep track of second sense some date in the mid-sixties. This why *mix can run on systems with out hardware clocks, and why they are not going to even blink at the whole Y2K thing. Without the system running the time dosen't get updated.
    This is in fact a good idea because it keeps cron from getting confused when the time changes.

  • I always thought that was for a cpu fan.
  • Yeah, I have the same problem. I have a P166MMX ATX gateway system that runs Debian and winNT (I only ever go into NT for Quicken and some Photoshop (GIMP's antialiasing of text SUCKS)). I never turn off at school - but try to not leave on at home (no network, no computing). The problem is, after a few hours off - the damn thing just turns itself on! The first few times it really confused me, because i'd wake up and the computer be'd on. Then I caught it in the act, and have been wondering ever since.

    Haven't been able to figure it out, but I put it on one of those control base things that let you CUT the power to your box - and that obviously fixed it.

  • What kind of software do you use to make your server page you when it is about to go down?
  • For all intensive purposes, an ATX motherboard and power-supply combo can "power-off" a system. This "feature" can be set up (usually) by using a combination of the BIOS settings, and the operating system.

    However, there will always be a small amount of electricity flowing through the components from the power supply when it is off. With that in mind, the power supply is never fully, completely off. This is why it is best to unplug the power cord from a system before any tinkering inside of it is done.
  • I got the impression that the author wasn't worried about outages, but about the motherboard failing to 'ask' for power... It's just one more thing that can go wrong, and isn't worth the convenience (the ability to automatically power-off like a Mac) on a server system.
  • Even if it is Windows.

    The NEC Versa LX notebook I just got performs a Windows shutdown when the power switch is cycled. If the power switch is held down for three seconds it does a normal power off without the shutdown. Kind neat.

  • The Intel AL440LX has an auto-power on option in the same place as the restore option. So I wouldn't be surpised if the N440BX has the same option.

    Of course with the restore option (and the P2B-F's in that matter) it only works if you never use the ATX front-panel power switch... but if you never wire that it's not an issue :)

    As to the original question... I'd just advise using Intel/Abit/Asus etc... mobos or something like that which DO have the always power on feature.

  • Its not worth it. Most if not all cases come with power supplies. Also, replacement power supplies are relatively inexpensive.
    Finally, the old power supplies were/are very inefficient; they work by using transformers to step down the voltage. The newer power supplies, including the ATX supply, are switching power supplies. This means that more of the power actually gets to the motherboard and less is wasted in the conversion.
  • In Short, ATX refers to both the form of the motherboard AND the power supply it requires. An ATX system typically allows the OS to power off the system, much like a laptop. ATX motherboards (again, referring to the arrangement of components, including the plate near the top that houses a range of things from USB to serial connections, AND the different type of power supply) are increasingly more common, but don't forget than many AT-form factor motherboards now support ATX power supplies (I have one at work that's a Soyo Super-7 K6-2 based AT board that has both AT and ATX power supply connectors).
  • The GA686BXD (Dual PII) doesn't have this jumber, but the GA686C (Single PII) does.
  • Why should the P/S blow IN? I would think that would be worse as the "cooling" air would be heated by the P/S.
  • Most ATX motherboards should have a BIOS setting for "restore power after power failure" or some such. Usually there are three settings; "on", "off", and "previous state".

  • It depends on the BIOS you have - If you happened to buy an AT/ATX compliant motherboard, you may be out of luck, because they often ship with BIOSes that don't know how to handle ATX power. On ATX motherboards, I have seen that Phoenix BIOS has the easiest to understand options, in the Boot section:
    - Whenever power is detected, turn the system on, regardless of how it was shut down.
    - If power is detected and the system was not shut down by software, turn the system on. If it was shut down properly, then leave it off
    - Whenever power is detected, go to sleep until someone presses the power button. (this is the default setting for most ATX machines)

    What I did for my servers was to set them with the first option, and disconnected the frontal power button... The rear (real) power switch is harder to press by accident - and... Well, servers are meant to stay up, am I wrong? :)
  • When you power on an ATX system's power supply, you see the LEDs on your keyboard and case flash for a fraction of a second. This is enough to tell the computer "Hey, do you want to wake up?", let it check in the BIOS settings how you programmed the boot process to be, and take the right decision.
  • Where on the Kernel compile options can you find that? Do you have to recompile, or you can use some kind of "insmod poweroff"?

    Each time more people are using ATX motherboards, and I think that most of us, when typing shutdown -h now mean I want to turn off the bloody system - why is it not enabled by default?
  • Nope - that doesn't work, at least on the motherboards I have tested it. I don't know why, but I think that a braindead engineer preferred the machine not to boot at all than to boot at power up in case the button got stuck :-/
  • ATX has a cleaner, more logical, less crammed design than AT. The changes (at least, the ones that spring to me now) are:
    - The CPU is at the top, not in front of the cards. Many (most? [all?]) current processors will discourage you from using full-size cards on an AT motherboard because they are just in the middle.
    - RAM is in a much more convenient place than just under the power supply, it is actually possible to fill your motherboard with RAM without unmounting the power supply :)
    - Printer, serial, keyboard, mouse, USB, video (optional) and audio (optional) ports are soldered on the motherboard, so you won't have 5 extra cables on your case, which makes up a cleaner design, easier access and maintenance.
    - IDE, SCSI and floppy ports are located actually near the disks, so your cables won't run all across the place - once again, cleaner.
    - The boxes are a bit wider. I don't know what is that for.
    - The fan on the power supply doesn't suck the hot air from the case, it blows fresh air in instead. What's the effect? I don't know.
    - ATX motherboards can implement wake up on ring, LAN activity, keyboard activity or port activity, they are continously powered - I don't know if this can damage your computer in the long run. To keep safe, if you plan not to turn on the PC, turn off the UPS or regulator.

    I tend to prefer AT because it has been the standard for many years and still works very good, and I don't think most people will need this features. However, I just bought an ATX systems (they were out of AT motherboards and I needed the machine immediately), and assembling it was much easier. Also, it's easier to spot a misconnected cable. I think I will start recomending ATX now.

  • Do it with hardware (Score:5)

    by SEWilco ( 27983 ) on Thursday May 27, 1999 @08:15AM (#1876970) Journal
    At hardware and auto parts stores you can get solderless connectors which tap a wire into an existing wire. You can use two of them to add a wire which connects the wires for pin 14 and a neighboring wire. They're usually blue plastic with a metal tab which you push into the two wires. Faster than soldering and can be removed for maintenance.
  • I'm using an alton 747 bx board that has AT and ATX power connectors and also has a hardware jumper to disable ATX power functions. I ended up with this setup just by coincidence after I had done some upgrades and had enough spare parts for another server (except cpu). I have and old 386 tower sitting around so I got an AT-slim ps from pc power and cooling and then got a p2-450 to put on the board. Since the board accepts an AT powersupply, there is no danger of it powering down through some ATX function.

  • Hmm.. that sounds vaguely familiar. A friend of mine's computer mysteriously powers on during lightning storms and come to think of it, it's an ATX mobo. Of course, he *should* unplug his computer from the outlet during these situations, and it could be the cause of his current computer woes (or maybe it's just Windows 95). Anyone else have this "problem" (strange powerups)?

  • Yep. works well for me, as that is what i am doing
  • An ATX supply provides a standby power output (called "+5VSB") which the motherboard can use for soft power control. This standby power can power the circuit that decides if the main supply rails should be switched on or not.
  • An AT-style power supply connector usually cannot be installed "backwards" in that the connector is rotated 180 degrees, but "backwards" in the sense that the two connectors are swapped. Simple rule for AT power supplies: "Black to Black" - the five or so black ground wires should always be in the center of the two plugs.
  • Just a little note: Not all ATX power supplies will work with this setup. At work we use ATX supplies for bench-testing some of our equipment (non-PC) and at least one ATX supply will not power up with the PS-ON connected to ground. The power supply will only switch on when the PS-ON signal goes from being open to being grounded. I don't think anyone will have a problem with most of the ATX supplies out there, but some of them won't work with this mod.
  • All of the Asus P2B series boards that I've used do power on when the cord is first plugged in; in other words, when power is first applied to the PS, the board turns on. So that takes care of power failures, etc. I haven't checked for relevant BIOS settings, but if there are any, they apparently default to this behavior.
  • Because the power switch goes through the MB instead of to the power supply it makes it more difficult to narrow down problems with a computer. Is it the Power Supply? Is it the Motherboard? You can't tell without popping out the power supply and trying a known good one.

    Additionally, is it just me or are the ATX power supplies more fragile than AT's? I've had to replace two of them myself in the period of one year, and my brother had to replace his a few months after he got his. I don't know about him, but I have a UPS (APC) so I know my power is clean. And I've been buying good quality power supplies. Argghh, I like the ATX case design thing-ma-bobs, but I REALLY don't like the power supplies. Is it just me? Am I cursed? Or have other's experienced my travails?

    BTW, this is very recent so it's not because I'm using pre 2.01 compliant power supplies.

  • A couple options (Score:5)

    by Zachary DeAquila ( 31195 ) on Thursday May 27, 1999 @07:16AM (#1876979) Homepage
    There're a couple of options

    Some ABIT motherboards can be strapped with a
    jumper between the reset and power switch connectors to make them act in an 'always on'
    fashion like AT motherboards do.

    If you feel like experimenting a bit, you can
    look in electronics parts catalogs for POST
    transistors... these basically act like delayed
    momentary-on relays. One lead goes to +5V,
    the others go to the power-switch jumper.

    --Z
  • I am curious to know how this works. If you, say, wanted the reverse behavior, for the power enable line coming from the MB to the power supply to disable power supply startup when power is supplied. How does the system power up far enough into the powerup sequence for the Motherboard to read bios settings and know it should not continue to power up? Is there a bistable latch circuit in there somewhere?

    **offtopic alert**
    It's a phenomenon I have encountered in embedded designs in the past- a device has to be on in order to know if it should turn itself on. It certainly sells more MOV and Gas-tube surge protectors to have everything powered up and ready for a soft-powerup, or a lightening jolt. In the old days when you turned something off, it stayed off and was physically disconnected from the power grid. (shaking head and muttering about how glad I am my Stereo uses a Harmon-Kardon tube-type integrated amplifier)
    **end of offtopic alert**

    Awhile back I needed the additional 3.2 volts that an ATX supply provides, to power a DEC Alpha motherboard (Cabriolet). The motherboard used the traditional AT-type power supply connector but had additional connectors for the 3.2 volts it needed. I ended up chopping off the connector on an ATX supply and splicing on a traditional AT connector, then hooking up the 3.2 volt connectors to the 3.2 volt lines. Then I had to ground the control line that the MB uses to control the power supply. It worked out okay, after a bit of reverse engineering the ATX power supply spec.

    Melbert
  • It isn't possible to power an ATX motherboard with an AT power supply. The ATX power supplies produce an additional 3.2 volt output not present on a standard AT power supply. That is why modern motherboards that work with AT power supplies always have a big heatsinked part on them, to regulate and produce the lower voltage not availabe from an AT power supply (which only supplies +-5 and +-12 volts). A cord adaptor to allow you to use an AT supply on an ATX motherboard would have to incorporate a pretty fancy inline DC-DC converter to give the MB the voltages/current it needs.
  • Reset is a pretty cruel and unforgiving beast.

    I suspect that a power-switch event could be linked to trigger a shutdown somehow, maybe in the inittable (?) like control-alt-delete is. It seems like pretty low-level hardware stuff, though. Does anybody know if power-switch handling is part of the ATX specification (which I believe is available at Intel somewhere) to make it worth the effort of something like this? (it would be a nightmare to have it be architecture dependent with tons of variations)

  • Paging (Score:2)

    by schon ( 31600 )
    What kind of software do you use to make your server page you when it is about to go down?

    Try [maclawran.ca]
    Big Brother... it works wonderfully for this.. (assuming your UPSd logs things like that :o)


  • Re:Mod to keep power on all of the time (Score:4)

    by aressa ( 35705 ) on Thursday May 27, 1999 @08:23AM (#1876990) Homepage
    Make sure you drop a 10kohm resistor in there, or you will reduce the life of the power supply...
    We ran into this same issue with some boxes that we wanted to boot right up when power was applied, soldering gun and a handfull of Radio Shack Resistors later we were golden...

    A

  • Re:Do it in hardware (Score:5)

    by komet ( 36303 ) on Thursday May 27, 1999 @07:38AM (#1876991) Homepage
    We do this with all our servers (medium-size ISP in Europe) and it works without any problems at all.

    Just take a bit of the isolation off the wire of Pin 14 (usually, bit NOT ALWAYS green). There's a GROUND signal on either side of this; take the isolation off that as well. Then solder together and wrap with insulating tape. Works like a charm. The only thing you shouldn't do is tell the mobo to power off; it gets a bit confused then.

    This is where pin 14 is:
    Looking down onto the contacts: (wires on rear)
    v--(plastic tab)
    ===
    + + 0 # 0 + + + + +
    + + + + + + + + + +

    # is pin 14. 0 is ground signal, should be a black wire.
  • Unfortunately, even in the M/B's "suspend" mode most AT power supplies still stay on and supply the x00 watts to the M/B. The ATX M/B can control the ATX power supply, however, which is very convenient for this.

    --bdj

  • Some of the comments I've seen hit it on the head: The Brand + Model + Revision of Motherboard you buy is key for ANY system (esp. a server).

    Myself I don't like buying anything I haven't researched to death; I also really want to hear what other people's experience has been with the stuff. (note: people I trust... not some script kiddie; though on /. you can usually tell which are which)

    Checking Tom's Site [tomshardware.com] and Ars Techina [arstechnica.com] to see what they think doesn't hurt either.

    As far the Power-Always-On feature you want well, 3 things:

    1) I would not recommend any soldering at all as your warranty will most likely go out the window. This can be a very bad thing in a corporate environment.

    2) Most high quality Motherboards come with either a CMOS or Jumper setting (or combo) which tells the PS what to do when power is applied.

    3) A good case goes a long way... both in ensuring you have a quality PS, and that you won't slice your fingers when you need to work on it. (Can we have a poll on how many times we've done that? :) ) You'll want to make sure the PS is compliant with ATX 2.01 to be on the safe side too.


    As far as personal choices, I'm a big fan of both Supermicro [supermicro.com] Motherboards and Cases - my personal fav is the SC-701a style case - it's a beauty! but unfortunately its soon to be discontinued... :(

    Anyway, just my $0.02 (which is worth even less as its Canadian...)
  • I have a Supermicro P6GDE (Dual PII 400, GX chipset) that has a jumper on the MB to choose, "Bios Power Stater, or Always on..."

    As soon as I through the switch on the back of the powersupply its on, and if that switch is left in the on position then it comes on when the power returns after a black out, etc.

    Note the case is a supermicro ATX750 or something(which is really an Antec Case and Powersupply) and it seems to be designed just for this...
  • [ This is an expanded version of an earlier posting. It also has an useful subject. ]

    A while back, I did some current measurements on my ATX PC. Here are the results:

    PC off, monitor off: 0.10A (12.0W)
    PC off, monitor in "power save/standby": 0.17A (20.4W)
    PC off, monitor "full on": 0.70A (84W)
    Everything on: 1.44A (172.8W)
    PC on, monitor off: 0.82A (98.4W)
    PC and monitor in "green/power save" mode: 0.60A (72W)

    The eye-opener is the fact that, even with everything "off", the system is still eating 12 or so watts. I was expecting maybe 8W, but not 12W. ATX systems can eat significant power, even when off.

    Note that the "full on" current was measured just after Win98 had finished booting. A CPU-intensive program like Quake or Unreal wasn't running. It would be interesting to see how much of a difference, if any, there would be.

    It's also interesting to note that, if I disconnect all cables from the power strip and then measure the current, the power strip is drawing 0.01A (1.2W). This is probably leakage current from the surge suppression in the power strip (or measurement error in the ammeter).

    These numbers are probably on the high side, as my system has a Voodoo 2 and two hard disks. Detailed specs:
    • ABIT BH6 motherboard w/64MB CAS2 PC100 SDRAM.
    • Celeron 300A overclocked to 450MHz (100MHz FSB).
    • Two IBM 10GB IDE drives (7200RPM).
    • Creative Labs Graphics Blaster Riva TNT.
    • Creative Labs SB Live Value.
    • Creative Labs 40X CDROM (hey, I've got all this CL stuff because it was cheap).
    • Generic "made in China" Voodoo2 card.

    Side note: from measuring my girlfriend's PC, which is virtually identical to mine except she has only one IBM hard disk and no Voodoo2 card, the Voodoo2 card and extra hard disk use up 0.27A or 32.4W. Most of this is probably from the Voodoo2 card.

  • Re:Example ATX power utilization (Score:3)

    by kral ( 182123 ) on Friday May 28, 1999 @07:48AM (#1877020) Homepage
    Initial note on credibility: I have been in the employ of the leading manufacturer of power supplies for more than a decade, in a highly technical role.

    Your conclusions are accurate, but your actual data is wrong. The power is way too high to be wasting on every PC on the planet, but is in fact significantly lower than you measured. The power into a PC power supply cannot be estimated by measuring the AC current and multiplying times the known AC voltage. This would work with in-phase sine waves or DC, but not with the highly distorted current waveform present at the input of the power supply. Most of the measured current is in odd harmonics (3rd, 5th etc.) of the 60 Hz line. Multiplying a 180 hz current times a 60 Hz voltage will give alternating positive and negative power over time, with zero net power. (i.e. the 180 Hz power alternately flows from mains to computer, then from computer to mains). Your current meter measures this as part of the total current, though it doesn't reflect power. (Your local Utiliy's Wattmeter is not fooled - it reflects true power).

    A Wattmeter is required, and the error can be 3x. (Actual power dissipated is lower is lower than you calculate.) This has been a real pain for a non obvious reason. Typically, we design equipment for use on a 10A service (assumed minimum size of fuse/circuit breaker; the ratings on wiring, wall sockets, etc.), which would imply that we could go up to 1200 W without problem even in a consumer environment. In fact, we will exceed 10A RMS on significantly lower power units. High power systems (750W and above) will almost always need an additional power stage, a Power Factor Correction (PFC) front end, which pulls power from the mains in a clean sinusoidal waveform to allow us to extract the maximum power at a given RMS current level.

    Two upcoming related issues will increase the cost of power supplies: In Europe, there is already a new requirement on the books that will require that the level of harmonics in the current draw be reduces (same as saying that we will be required to make the input current look more sinusoidal). This will be a new requirement on low power (consumer sort of power level) equipment. Second, the "low power" energy savings modes will be made significantly more stringent, which will have significant impact not only on computers, but on all that consemer gear that stays alive waiting for your IR remote to tell it to power up. Even little wall-warts will be affected. the power drawn by old fashioned steel transformer based warts is on the order of a Watt or two - multiply that number times the number of cell phone chargers hanging off people's wall sockets 24 hours a day and you'll see staggering costs to society.

Slashdot Top Deals

HELP!!!! I'm being held prisoner in /usr/games/lib!

Close