Intel's new ATX12VO computer-supply spec

With a 12v single prime power, a PC motherboard might just have a wall-wart connector.

And maybe we'll see super-cheap power converter modules, like the LTM-series from LTC, but a fraction the price.

Of course, 12 is the wrong voltage. It should have been 24.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"

technology.com:

f-power-

le-rail-

yeh, but. I think most MBs are already designed to get most of their power from 12V and there is a long way down from 24 to 1-2V

We were just talking about that on Friday: best way to get from 24 volts to +1. Single buck? Cascaded bucks? More complex magnetics?

One of my guys asked an ADI applications engineer, regarding an LTC synchronous buck chip to do it. His reply was typically profound: it depends.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"

When a PSU dies, you can just buy a new one. When something critical fails on a motherboard, it usually requires a new board, and then you find that one compatible with your existing CPU and memory is no longer available, so that means buying new ones of those, and throwing out perfectly good hardware because it's been rendered obsolete.

And the increased heat dissipation on the motherboard makes it more likely to fail.

Sylvia.

Sylvia Else wrote in news: snipped-for-privacy@mid.individual.net:

Well, then crank up the clock!

Just place it in a fluorinert bath first. :-)

of-power-

gle-rail-

What a load of hooey. The only failure I've ever had in a PC was the caps in the voltage regulators on board for the CPU. Some years back there were a bunch of poor quality electrolytic caps made that found their way onto q uality motherboards.

Paranoia about the power supplies on a motherboard failing is not very real istic. PSUs that fail are designed poorly to be sold cheaply. Buy cheap m otherboards and they can fail as well. Buy quality stuff and it will outla st your need for the computer as so many of them do.

--

  Rick C. 

  - Get 1,000 miles of free Supercharging 
  - Tesla referral code - https://ts.la/richard11209

er-

My mobos mostly fail due to bad caps. They tend to boot right up after replacement of the three to four dimensionally largest cap families on the mobo.

Thank you,

--
Don Kuenz KB7RPU 
There was a young lady named Bright Whose speed was far faster than light; 
She set out one day In a relative way And returned on the previous night.

of-power-

gle-rail-

why would the heat dissipation increase? all the regulators are already on the motherboard, a few of them might have another input voltage that's i t

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

I had a single CPU AMD in their early years and all it took was one bad heat sink remount and it fried in seconds, literally, and took the motherboard with it!

My ONLY ever fail like that. AMD search up showed that it was a very hot CPU despite not being that fast.

My Dual CPU AMD ran hot too, but not as much. My subsequent Intel purchase had 6/12 cores and it NEVER has run hot.

Heat fails do happen.

But the most common failure in ALL electronics is at power on and power off events. That is one reason some folks leave their machine up 24/7/365.

e-of-power-

ingle-rail-

it

I don't think so. There are 5 and 3.3 volt circuits which will now be on t he Mobo. The 3.3 volt circuit likely doesn't need much, but the solid stat e disk drives run off of 5 volts, no? Actually, I don't know that for sure . Many SSDs are now on special format plug in cards rather than the old di sk drive form factors. They may have dropped the 5 volt input since they a ren't spinning rust and be using 3.3 volts which they in turn drop down to a lower voltage to power the Flash internals. I'm assuming Flash uses a lo wer internal voltage than 3.3 volts. Nothing else uses 3.3 volts internall y these days. But then Flash needs internal charge pumps to generate a hig her programming voltage, so the circle is complete.

--

  Rick C. 

  + Get 2,000 miles of free Supercharging 
  + Tesla referral code - https://ts.la/richard11209

Your crapping up of a heat sink mount has nothing to do with the idea that adding the 3.3 volt regulator to the mobo will cause premature heat death on mobos.

But then I'm breaking my rule in replying to you. It's not like I expect a reasonable response.

--

  Rick C. 

  -- Get 2,000 miles of free Supercharging 
  -- Tesla referral code - https://ts.la/richard11209

Am 15.03.20 um 02:57 schrieb Sylvia Else:

At the moment, I can barely smile at this. I had it happen this week. After years of flawless service, the workstation below my desk went kaboom. It made a blinding light on the opposite wall.

The disks have survived, that's all I have checked up to now.

cheers, Gerhard

The specifications are quite picky allowing only small voltage tolerances and tight ripple requirements. You couldn't use a float charged standard 12 V battery across the 12 V line (requires 12.6 to

13.8 V). After all, there are regulators on the MB that should handle the 12 V input voltage variation.

Preferably 20 - 28 V voltage range to allow a battery in parallel to act as a short time UPS.

Why not use isolated on board DC/DC converters. This would help with ground plane noise issues, when no high currents would run all around the MB.

In large server farms the 380 Vdc (+/-190 Vdc) feed is sometimes used and isolated converters are used to generate lower voltages.

OK, that is a bit extreme. Still, perhaps it was the result of a crow-bar circuit protecting the rest of your system. You can hope.

Sylvia.

Maybe we should get Skybuck's opinion on why motherboards fail?

High efficiency standard-footprint single-inductor GaN based buck switchers would make replacement infrequent and practical. And we could use them too.

Probably polyphase for the big stuff.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"

A more modern switcher module would use all ceramic caps.

There would need to be some bulk storage for gross load steps, like whan a CPU comes out of sleep, bit the single prime source could handle that.

Really, we're ready for a new ps architecture, for PCs and other things.

--

John Larkin         Highland Technology, Inc 

The cork popped merrily, and Lord Peter rose to his feet.  
"Bunter", he said, "I give you a toast. The triumph of Instinct over Reason"

Small Form Factor (SFF) PCs manufactured by Intel, Dell, HP, and others have used ~19VDC wall warts for nearly a decade now. It's obvious to me why they chose that particular voltage.

Thank you,

--
Don Kuenz KB7RPU 
There was a young lady named Bright Whose speed was far faster than light; 
She set out one day In a relative way And returned on the previous night.

Gerhard Hoffmann wrote in news:r4kqi4 $vji$ snipped-for-privacy@solani.org:

Thankfully, they USUALY fail open. But not always.