Conversion Problems?

Add Your Comments

More stories have come out about the Prescott/Springdale/Canterwood “problem,” though not more information.

Whoever is leaking this story to various websites is being very coy about it, and not providing specifics.

The references to FMB (flexible motherboard standards) is the only viable clue to go by so far. We mentioned yesterday that Prescott had a number of additional requirements over and above that for Northwoods.

Disclaimer: Don’t take what I say below to the bank. This seems to be the most likely explanation of the problem, but I could be wrong, or even barking up the wrong tree; the problem may be elsewhere.

So what I did was to take the two 865/875 boards I had, and looked for the chips on the board responsible for handling current. These are called DC-to-DC converters.

On the Asus P4P800, I found an Analog Devices ADP3180 converter. The datasheet for that can be found here

The Abit IC7 uses an Intersil ISL6556BCB converter. Here’s the datasheet for that.

When you look at the datasheets, it very clear that both converters have the required voltages and processor sensing circuits needed to handle Prescott.

What is somewhat questionable is whether these converters and how they’re mounted on the motherboard (more on this later) can handle the 70A.

The Intersil datasheet has an interesting statement on the subject (page 17):

“The first step in designing a multi-phase converter is to
determine the number of phases. This determination
depends heavily on the cost analysis which in turn depends
on system constraints that differ from one design to the next.
Principally, the designer will be concerned with whether
components can be mounted on both sides of the circuit
board; whether through-hole components are permitted; and
the total board space available for power-supply circuitry.
Generally speaking, the most economical solutions are
those in which each phase handles between 15 and 20A. All
surface-mount designs will tend toward the lower end of this
current range.
If through-hole MOSFETs and inductors can
be used, higher per-phase currents are possible. In cases
where board space is the limiting constraint, current can be
pushed as high as 30A per phase, but these designs require
heat sinks and forced air to cool the MOSFETs, inductors
and heat-dissipating surfaces.”

To illustrate the potential problem with an example, let’s assume a mobo has a four-phase power conversion, and each phase is meant to handle 16A.

4 X 16 = 64. 64 is less than 70.

This statement also explains what the problem is: the problem isn’t that the circuitry can’t handle the electricity; it’s that the motherboard design isn’t meant to handle the heat.

Might a fan help? Maybe, maybe not.

Wouldn’t Gigabyte’s Six-Phase Power Solution Take Care of This Problem?

Don’t assume that. Ask Gigabyte that. So far, they haven’t made any claims about Prescott. I’m hardly a circuit board engineer, but the Gigabyte system uses some older voltage regulator chips (HIP6301CB) that do not support at least some of the Prescott standards, so again, don’t assume it’s OK.

Another Possibility

Is this a real problem, or is Mr. Mysterious Mobo Maker doing this to freeze sales until they can get their product out? We don’t know. This article is a best guess based on very little solid data.

What is most frustrating about this is that those who have been in contact with this unknown company apparently didn’t ask the right, or even any questions about this, like “What specific change obsoleted the current motherboards?”

It may be possible that Intel changed some Prescott specification at the last moment when it published the VRD in April. It could be something else.

For instance, if you look at the revision guide section of the 875 datasheet (page 19), there are all kinds of changes, including more than a few changes to resistor values. Maybe that’s the problem.

The reason why I had and still have some doubts about whether there actually is a problem or not is due to the poor quality of the information we’re getting from Mystery Mobo Man. If the information were specific, then it could be checked, and we could see how much of a problem this really is.

But right now, all we can do is make educated guesses.

Ed

Leave a Reply

Your email address will not be published. Required fields are marked *