Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 2 of 4 FirstFirst 1234 LastLast
Results 11 to 20 of 36

Thread: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch




  1. #11
    Join Date
    Mar 2014
    Location
    Columbia, MD, USA
    Posts
    9

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    Thank you all for your interest and feedback. Youíve got a really helpful forum going on here. Thereís a lot here youíve written for me to process.

    Quote Originally Posted by profJim View Post
    Older systems like yours required most of the power to be supplied by the 3.3 volt and 5 volt rails. Modern power supplies provide very little power to these rails and the bulk of their power is delivered by the 12 volt rail. Looking at the Seasonic 860/1000 watt specs, there still isn't much power available for the 3.3 volt and 5 volt rails, where the maximum total output for both of these rails in only 125 watts.
    Yes, I agree. We can reasonably conclude that the system requires no more power than what is available from the Sparkle SPI 250G since itís a drop-in replacement for the original PSU. What we donít know is whether that power supply is over specíd for the system requirements; it may be capable of providing more power at the required voltages than the system needs. The Seasonic Fanless 400 supplies up to 20a of +5vdc whereas the Sparkle supplies up to 26a of +5vdc. It is my theory (wag) that the Sparkle is capable of producing much more current at +5vdc than the system will ever need. Of course the only way to know that is with an ammeter, which I have not done. This is a very lightweight system from a power consumption perspective.

    Quote Originally Posted by profJim View Post
    If your motherboard has a compatible 20-pin ATX connector, you should be able to replace the 120 volt AC power switch with a small momentary ON push button switch where these wires would go to the light green "power-ON" pin and a black ground wire. These are the same two connections that are used when doing a power supply paperclip test.

    The new on/off switches wires would need to make a permanent connection to the light green and black wires on the top of the connector.
    Iíve thought about this. I can probably find a momentary switch having the same form factor as the original AC rocker switch and there is actually an ATX-ON pin pair on one of the boardís connectors. I think I can use that pin pair (according to the motherboard manual) to accomplish the same thing as shorting the PS_ON and COM/GND pins on the 20 pin connector with a momentary switch.

    Quote Originally Posted by parsec View Post
    Assuming that the PSU in the link provided is an accurate analog of your current PSU, I have a few comments:

    The technical specs of the AT PSU described in the link does not list a 3.3V rail at all. Further research said AT style mother boards created their own 3.3V supply by deriving it from one of the other rails, +5V or +12V.
    Yep, no 3.3vdc produced by the original PSU.

    Quote Originally Posted by parsec View Post
    Next, we see a -5V rail on the PSU in the link. That rail does not exist on an ATX PSU. Whether or not your board requires the -5V rail, I don't know.
    I think not.

    Quote Originally Posted by parsec View Post
    Fortunately your mother board can accept a 20 pin ATX plug, otherwise adapting any ATX PSU for AT P8 and P8 jacks would be a major issue for you. Apparently your board does not need the -5V rail since it can use an ATX 20 pin connector.
    Agreed.

    Quote Originally Posted by parsec View Post
    Checking the specs of the latest Seasonic 400W fanless PSU, the specs show a max of 20A on the +5V rail, even if no power is drawn from the +3.3V rail. As Prof Jim said, the largest Seasonic PSUs supply a max of 25A on the +5V rail.
    Agreed. But my spidey senses are telling me 20a is way more than enough for this system. I know thatís not the responsible engineering approach to this problem but Iím still pretty confident in that theory. The responsible thing to do would be to measure the actual system current draw on +5vdc.

    Quote Originally Posted by parsec View Post
    I agree with Prof Jim's analysis of the power switch situation. AT PSUs had the AC power switch of the PSU on the PC case. AT PSUs did not have electronic on/off power control as ATX PSU's do. His suggestion of connecting a momentary switch to the appropriate pins on the 20/24 pin plug on the PSU should work.
    I think it will work too, but Iíve got another tweaky challenge to sort out, which Iíll get to in a moment.

    Quote Originally Posted by parsec View Post
    OTOH, the pin arrangement of the 20 pin socket on your board should probably be verified before using the new PSU, and it might be possible that it does provide electronic shutdown via the OS, and may even have a header for a standard ATX PSU/mobo style electronic power switch.
    Well, this freaks me out a little bit. Isnít this 20-pin plugís pinout supposed to be a standard? I actually do know the pinout on the motherboardís 20-pin connector since itís documented in the motherboard manual. I donít know the pinout on the Seasonicís 20/24-pin convertible power plug. I guess I should get that info for the Seasonic and compare it with the motherboardís pinout.

    I believe I have access to the ATX electronic shutdown/startup through a pair of header pins identified as ATX-ON in the motherboard manual.

    Quote Originally Posted by - wardog - View Post
    VERIFY that each of the wire/pin outs of you current MB are pin to pin compatible with whatever PSU you decide upon. First!!!

    "Dad, is that smoke coming from your computer?" Or worse ......
    So wardog, you definitely got my attention! I will certainly look into that once I locate a pinout for the Seasonic. I never knew about the Dell PSU non-standard wiring story; that seems just grossly irresponsible. But thankfully, the PSU is not a Dell product.

    Quote Originally Posted by - wardog - View Post
    Heaven forbid who ever manufactured this as of yet unknown MB/system decided to use the above OEM Dell parts in it's manufacturing to save costs.

    Old addage but applies here nonetheless, The devil is in the details. Please see #1 in my Sig.
    I thought I included the relevant info about the system in my OP, but Iíve taken your advice and populated my sig with as much information as I have about the system, up to the forum-imposed character limit.

    There are a couple of other significant things to mention that come to bear on the discussionÖ

    1) This is not a DOS or Windows OS environment, per se. There is no CONFIG.SYS or AUTOEXEC.BAT. There is no Windows GUI interface of any kind so certain system features you might assume are available, like shutting down the system power from the OS, are just not there. This system boots right from BIOS into the Mackie realtime operating system, which is essentially custom OS code wrapped around an invisible Win98 (I think, could be NT) kernel.

    2) I need the 115vac that the AT style PSU made available to the system by way of the 115vac cable and power switch. In this particular system, the 115vac makes a stop at a proprietary power distribution card on its way to the front panel power switch. This power distribution card and a number of companion linear DC power supplies use the 115vac to produce a variety of DC voltages to be provided to the digital audio mixer control surface. So whatever scheme I come up with for using an ATX power supply in this CPU/PSU chassis will have to also make 115vac available elsewhere in the chassis. The standard ATX PSU form factor has no such capability as I understand it. Iíd love to be wrong about that.

    Thank you all for your excellent support on this subject. Iím very grateful for your help.
    Subject System d8b: A circa 1999 purpose-built CPU/PSU controller for the Mackie Digital 8 Bus (d8b) digital audio mixer (d8b Database). The CPU/PSU chassis houses both a microprocessor-based contoller system and a number of power supply components that supply various DC voltages to a remotely located audio mixer control surface. Comms between the control surface and the CPU/PSU are by way of a parallel data cable. DC is supplied to the mixer control surface from the CPU/PSU chassis by way of a separate multiconductor cable. CPU/PSU features -

    • Itox CB50-BX motherboard 100 MHz FSB (link)
    • Intel 440BX AGPset chipset
    • Award BIOS v4.51PC - Windows (R) 95/98 Plug and Play Compatible
    • Celeron CPU Socket 370 at 300A MHz
    • 256 MB PC-100 SDRAM (upgraded from as-built 32 MB PC-100 SDRAM)
    • ATI Rage 128 MB AGP video adapter
    • Ethernet card - generic
    • MIDI card - Winman Midiman 1x1 ISA
    • 4 GB Compact Flash in IDE-to-Compact Flash adapter boot drive (upgraded from as-built IDE Maxtor 20 GB HDD)
    • 1.44 MB Floppy Drive
    • 1 AGP slot, 3 PCI slots, 1 16-bit ISA slot, 1 shared ISA/PCI slot
    • PSU - Sparkle SPI-250G 250 watt AT with front panel AC switching (link)
    • OS - custom realtime OS wrapped around a Win98 kernel (i.e., Windows is not accessible)

  2. #12
    Join Date
    Nov 2012
    Location
    USA
    Posts
    2,675

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    For those looking for the motherboard manual - http://www.sonido-7.com/d8b/files/CB...ser_Manual.pdf

    EDIT: Well for whatever reason I can't add hot linking w/o it somehow adding a UTF-8 LRM(E2 80 8E) to the end of the link. Copy and paste link if interested. Best I can do for now.

    EDIT 2: Upon re-visiting and re-editing this post it appears correctly now. Me thinks Cameron's fingers are busily working behind the curtain atm.
    Last edited by - wardog -; 03-16-2014 at 12:56 AM.
    #1 - Please, when seeking help, enter the make and model of ALL parts that your system is comprised of in your Signature, or at least the model #'s in your System Specs, then "Save' it.
    ____If you are overclocking, underclocking, or undervolting any parts, informing us of this and their values would prove beneficial in helping you.

    #2 - G.Skill RAM Configurator for your boardSamsung Memory for your boardLatest AMD Chipset Drivers/WindowsLatest AMD Graphics Drivers/WindowsLatest Intel Drivers

    #3 - Please use the eXtreme Outer Vision Power Supply Calculator found HERE to determine if it might be your PSU at issue.
    ____Consider your PSU to be the foundation from which all else is built upon. Anything built upon a weak foundation is poorly built.

  3. #13
    Join Date
    Nov 2012
    Location
    USA
    Posts
    2,675

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    And hopefully, here's the Sparkle SPI-250G PSU Manual/specs w/Pin Outs - www.sparklepower.com/pdf/SPI-250G.pdf
    #1 - Please, when seeking help, enter the make and model of ALL parts that your system is comprised of in your Signature, or at least the model #'s in your System Specs, then "Save' it.
    ____If you are overclocking, underclocking, or undervolting any parts, informing us of this and their values would prove beneficial in helping you.

    #2 - G.Skill RAM Configurator for your boardSamsung Memory for your boardLatest AMD Chipset Drivers/WindowsLatest AMD Graphics Drivers/WindowsLatest Intel Drivers

    #3 - Please use the eXtreme Outer Vision Power Supply Calculator found HERE to determine if it might be your PSU at issue.
    ____Consider your PSU to be the foundation from which all else is built upon. Anything built upon a weak foundation is poorly built.

  4. #14
    Join Date
    Mar 2014
    Location
    Columbia, MD, USA
    Posts
    9

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    Here's how one guy did it -

    D8b DataBase • Tips

    He actually modified his ATX PSU by installing the 115vac power switch cable he removed from his old AT PSU into his new ATX PSU on the plug side of the AC power connection so he had access to the 115vac for distribution to the front panel AC rocker switch and to supply 115vac to the proprietary power supply boards in the Mackie power portion of the CPU/PSU chassis producing +/- 48vdc, +16vdc, etc. He wound up with 2 switches on the front of the chassis, (1) the original DPST AC rocker switch controlling 115vac to the various Mackie power supplies and to the ATX PSU, and (2) a momentary toggle switch he installed adjacent to the AC rocker switch to switch ATX-SW on the motherboard signaling the ATX PSU to startup or shutdown. This resulted in a power up sequence requiring that he first flip the 115vac rocker to get power to the ATX PSU and then hit the momentary switch to actually power up the CPU from its quiescent state.

    I'd like to find a more elegant way to do this using combined functionality switch. I'm not sure it's possible. Is there such a thing as a ganged switch that will operate 1 circuit as DPST 115vac and another circuit as a normally OFF momentary ON for +5vdc? I may be relegated to the same 2 switch approach that this guy used.

    Does the ATX-SW switching function in the ATX PSUs respond to the momentary switch closure to both startup and shutdown? That is to say, if the PSU is running, will sensed switch closure at ATX-SW cause the PSU to shut down?

    Is the ATX PSU ATX-SW function oblivious to persistent momentary switch closure? In other words, in the paper clip PSU test, what are the consequences of leaving the conductor in place beyond just a brief closure? Are we talking fire hazard here or no adverse affect whatever? Is the ATX PSU just waiting for another ATX-SW closure event to change the powered up state of the ATX PSU to powered down? Thanks.
    Last edited by bitSync; 03-16-2014 at 03:24 AM.
    Subject System d8b: A circa 1999 purpose-built CPU/PSU controller for the Mackie Digital 8 Bus (d8b) digital audio mixer (d8b Database). The CPU/PSU chassis houses both a microprocessor-based contoller system and a number of power supply components that supply various DC voltages to a remotely located audio mixer control surface. Comms between the control surface and the CPU/PSU are by way of a parallel data cable. DC is supplied to the mixer control surface from the CPU/PSU chassis by way of a separate multiconductor cable. CPU/PSU features -

    • Itox CB50-BX motherboard 100 MHz FSB (link)
    • Intel 440BX AGPset chipset
    • Award BIOS v4.51PC - Windows (R) 95/98 Plug and Play Compatible
    • Celeron CPU Socket 370 at 300A MHz
    • 256 MB PC-100 SDRAM (upgraded from as-built 32 MB PC-100 SDRAM)
    • ATI Rage 128 MB AGP video adapter
    • Ethernet card - generic
    • MIDI card - Winman Midiman 1x1 ISA
    • 4 GB Compact Flash in IDE-to-Compact Flash adapter boot drive (upgraded from as-built IDE Maxtor 20 GB HDD)
    • 1.44 MB Floppy Drive
    • 1 AGP slot, 3 PCI slots, 1 16-bit ISA slot, 1 shared ISA/PCI slot
    • PSU - Sparkle SPI-250G 250 watt AT with front panel AC switching (link)
    • OS - custom realtime OS wrapped around a Win98 kernel (i.e., Windows is not accessible)

  5. #15
    Join Date
    Mar 2014
    Location
    Columbia, MD, USA
    Posts
    9

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    Well, I just answered a couple of my own questions. I tried an experiment where I wired up the new Seasonic fanless PSU to the CPU motherboard and peripherals and I left the old Sparkle AT style PSU 115vac hooked up to the auxiliary Mackie power supplies (+48vdc, +16vdc, etc.) and AC rocker switch, such that the chassis was using both PSUs. I turned on the chassis DPST 115vac rocker, made sure the Seasonic power switch was on, and shorted the ATX-SW header pins with a screwdriver. The system booted beautifully and I let it run for a bit. I then shorted the ATX-SW header pins again with the screwdriver and the Seasonic PSU shut down. I'll try this again and let the system run longer to see if there are any issues with the Seasonic fanless PSU but I am not expecting any problems.


    I need to figure out the whole 115vac distribution thing. I don't really want to modify the Seasonic PSU but I've got to get at that 115vac. A second power cord to the chassis might work but that's not ideal either.


    Also, still wondering about the best way to handle the power up / power down switching (i.e., both 115vac and +5vdc ATX-SW header). Might have to go with 2 switches.

    This thing is dead quiet, by the way, without the Sparkle PSU fan running (the Sparkle AT PSU did not run while only the 115vac was pulled through it; no DC current draw from the motherboard or peripherals).
    Subject System d8b: A circa 1999 purpose-built CPU/PSU controller for the Mackie Digital 8 Bus (d8b) digital audio mixer (d8b Database). The CPU/PSU chassis houses both a microprocessor-based contoller system and a number of power supply components that supply various DC voltages to a remotely located audio mixer control surface. Comms between the control surface and the CPU/PSU are by way of a parallel data cable. DC is supplied to the mixer control surface from the CPU/PSU chassis by way of a separate multiconductor cable. CPU/PSU features -

    • Itox CB50-BX motherboard 100 MHz FSB (link)
    • Intel 440BX AGPset chipset
    • Award BIOS v4.51PC - Windows (R) 95/98 Plug and Play Compatible
    • Celeron CPU Socket 370 at 300A MHz
    • 256 MB PC-100 SDRAM (upgraded from as-built 32 MB PC-100 SDRAM)
    • ATI Rage 128 MB AGP video adapter
    • Ethernet card - generic
    • MIDI card - Winman Midiman 1x1 ISA
    • 4 GB Compact Flash in IDE-to-Compact Flash adapter boot drive (upgraded from as-built IDE Maxtor 20 GB HDD)
    • 1.44 MB Floppy Drive
    • 1 AGP slot, 3 PCI slots, 1 16-bit ISA slot, 1 shared ISA/PCI slot
    • PSU - Sparkle SPI-250G 250 watt AT with front panel AC switching (link)
    • OS - custom realtime OS wrapped around a Win98 kernel (i.e., Windows is not accessible)

  6. #16
    Join Date
    Nov 2012
    Location
    USA
    Posts
    2,675

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    I don't promote this type of idea as it's not the nicest thing you can do for powering up a board. But nor do I like folks prodding around inside a psu. Knowledge or lack thereof, it's a safety issue. Physically, bodily, and electrically.

    You stated there's no way in the OS to power down so I assume the switch is simply killing power while the OS is yet "active'.??

    Look in the BIOS at "PWR Lost Resume State" for an idea that might allow you to squiggle around cutting in a switch and be able to be used with a power strip with its own switch.

    I'm thinking you can shut the system down by just flipping the strips switch top Off. To turn the system on I'm thinking just flipping the strips switch to ON might, I say might, work for you there.

    The above info is to be used/considered loosely as you mention 48v and what not. I haven't a clue what you're looking at and working there with and just how things are wired or even where that 48v line you mention comes from.

  7. #17
    Join Date
    Nov 2012
    Location
    USA
    Posts
    2,675

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    ^^^^ or am I forgetting something that that might not work

  8. #18
    Join Date
    Mar 2014
    Location
    Columbia, MD, USA
    Posts
    9

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    Quote Originally Posted by - wardog - View Post
    I don't promote this type of idea as it's not the nicest thing you can do for powering up a board. But nor do I like folks prodding around inside a psu. Knowledge or lack thereof, it's a safety issue. Physically, bodily, and electrically.

    You stated there's no way in the OS to power down so I assume the switch is simply killing power while the OS is yet "active'.??

    Look in the BIOS at "PWR Lost Resume State" for an idea that might allow you to squiggle around cutting in a switch and be able to be used with a power strip with its own switch.

    I'm thinking you can shut the system down by just flipping the strips switch top Off. To turn the system on I'm thinking just flipping the strips switch to ON might, I say might, work for you there.

    The above info is to be used/considered loosely as you mention 48v and what not. I haven't a clue what you're looking at and working there with and just how things are wired or even where that 48v line you mention comes from.
    Hey wardog,

    Actually, I wasn't "inside a PSU", unless you're meaning the proposed mod to the ATX PSU to access 115vac? I like being alive. Shorting out the header pins for ATX-SW is electrically equivalent to a momentary switch shorting those header pins, so I'm pretty comfortable with that; they're header pins on the motherboard, they're not inside a power supply. Insulated screwdriver, steady hand, and all that.

    The Mackie OS software won't electrically power down the system, but it will do some software housekeeping in advance of a shutdown. There's an option in the Mackie OS GUI to shutdown the system, which when selected, goes off and does some file saving, etc. and then informs the user that it is now OK to shutdown the system, which in the normal system configuration means to flip off the 115vac rocker switch on the front panel, thus removing power to the Sparkle CPU and in turn the CPU components, as well as the special purpose DC power supplies for the mixing console.

    "PWR Lost Resume State" - nothing like that in this 1997 Award BIOS v4.51PC.

    I'm sorry, but I have no idea what you mean by a "strips switch". I also don't know what this means in terms of a comprehensive startup/shutdown for the system.

    The other DC voltages I mentioned have nothing to do with the CPU; they are created in the same chassis the CPU resides in (on the other side of a sheet metal divider) to be sent to the console control surface. They are power supplies separate and distinct from anything to do with the CPU. The +48vdc and +16vdc are created by these Mackie power supplied from the 115vac. The +48vdc is actually what is known as phantom power that is supplied to condenser microphones from microphone preamps. The only thing that the CPU and these power supplies have in common is the same power switch, the same 115vac, and they inhabit the same chassis.

    Very grateful for your feedback.

    I'm at the point now where I've demonstrated to myself electrically how all this could work. I need to decide if I'm going to mod the new ATX PSU or find some other method of getting 115vac into the chassis, what kind of switching I'll use, etc. I think I've learned that I can't get an ATX PSU to behave like an AT PSU on provision of 115vac power from a rocker switch (what I was hoping for). Unless there's some undocumented jumper configuration inside the ATX PSU I'm going to have to use PS_ON momentary voltage at ATX-SW to power on and off the CPU. The rest is going to be figuring out the approach, parts, etc.

    Thank you!
    Subject System d8b: A circa 1999 purpose-built CPU/PSU controller for the Mackie Digital 8 Bus (d8b) digital audio mixer (d8b Database). The CPU/PSU chassis houses both a microprocessor-based contoller system and a number of power supply components that supply various DC voltages to a remotely located audio mixer control surface. Comms between the control surface and the CPU/PSU are by way of a parallel data cable. DC is supplied to the mixer control surface from the CPU/PSU chassis by way of a separate multiconductor cable. CPU/PSU features -

    • Itox CB50-BX motherboard 100 MHz FSB (link)
    • Intel 440BX AGPset chipset
    • Award BIOS v4.51PC - Windows (R) 95/98 Plug and Play Compatible
    • Celeron CPU Socket 370 at 300A MHz
    • 256 MB PC-100 SDRAM (upgraded from as-built 32 MB PC-100 SDRAM)
    • ATI Rage 128 MB AGP video adapter
    • Ethernet card - generic
    • MIDI card - Winman Midiman 1x1 ISA
    • 4 GB Compact Flash in IDE-to-Compact Flash adapter boot drive (upgraded from as-built IDE Maxtor 20 GB HDD)
    • 1.44 MB Floppy Drive
    • 1 AGP slot, 3 PCI slots, 1 16-bit ISA slot, 1 shared ISA/PCI slot
    • PSU - Sparkle SPI-250G 250 watt AT with front panel AC switching (link)
    • OS - custom realtime OS wrapped around a Win98 kernel (i.e., Windows is not accessible)

  9. #19
    Join Date
    Nov 2012
    Location
    USA
    Posts
    2,675

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    "strips switch" You know. An extension cord with an integrated/integral rocker switch that you can cut power to all the cords outlets. Leave the PSU switched on, but adjust/adapt the BIOS to start upon AC power restored with using the extension's switch instead.

    That "PWR Lost Resume State", look in the ACPI section of the BIOS. In the linked MB manual above I offered, see pg 42, "3.1.4 Power Management Setup". Maybe it's not there as described in that particular BIOS revision but I'm guessing something akin to it is and would serve as functionally equivalent to/for it.

    Meh, scrap the above there I just typed here. In re-reading I see you say the 48v supply is also controlled by the cases switch. That "strips switch" idea of mine won't work.

  10. #20
    Join Date
    Nov 2012
    Location
    USA
    Posts
    2,675

    Default Re: PSU Wizzards: Adapting a New Fanless ATX PSU to an Old AT Environment - AC Power Switch

    Quote Originally Posted by - wardog - View Post
    Meh, scrap the above there I just typed here. In re-reading I see you say the 48v supply is also controlled by the cases switch. That "strips switch" idea of mine won't work.
    Well, or would it? Can I assume the 48v supply has a separate power cord?

Page 2 of 4 FirstFirst 1234 LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •