Rating of PC power supplies?

Nothing new there !

PG *may* be protective ground. Orange seems a odd colour to use though !

Graham

Cool. If you ever get tempted to raise voltage on the 5V output, make sure all elco caps and rectifiers can handle higher voltage. Those fast diodes have pretty low maximum voltage.

--
Siol
------------------------------------------------
Rather than a heartless beep
Or a rude error message,
See these simple words: "File not found."

Just how do they do that? There is just one "switcher" in the box, with a couple taps at the output. I'd say they probably "stabilise" just one of the outputs (5V) and check the +12V for the sake of protection, making sure it does not go too high.

In a PC 5V probably runs much of the logic directly without any voltage regulators and is as such probably far more critical as far as stability goes than 12V.

--
Siol
------------------------------------------------
Rather than a heartless beep
Or a rude error message,
See these simple words: "File not found."
as far as stability goes.

couple

(5V) and

regulators and is

It's possible to take 'weighted' feedback fom both rails. Basically you sense both and scale the Rs accordingly. It'll provide a compromise 'cross-regulation'.

Graham

Silly me - I forgot that one.

Not sure if anyone actually uses it !

Graham

On Mon, 30 May 2005 18:03:53 +0100, Terry Pinnell put finger to keyboard and composed:

I have seen only one AT PSU (made by Skynet) that regulates the +5V rail. In fact I'm using it in this machine. All the others have used the same technique as in the example quoted elsewhere. My other machine uses an Antec 300W PSU which also appears to regulate the +12V and +5V rails.

- Franc Zabkar

--
Please remove one \'s\' from my address when replying by email.

On Mon, 30 May 2005 19:57:56 +0100, Pooh Bear put finger to keyboard and composed:

PG = Power Good. See the schematic diagram at the URL I quoted.

Here is Intel's ATX PSU spec (versions 2.2 and 2.0):

- Franc Zabkar

--
Please remove one \'s\' from my address when replying by email.

If you see a small inductor placed between the main transformer and the 3V forward rectifier, it likely is a set reactor regulator controlling the 3V output ( input will be the 5V transformer winding).

This was a common method of post-regulating 3V on commercial ATX supplies originating ca '98 (Delta in Taiwan etc).

RL

And I really only meant the three or four black leads that terminate in the two MB connectors. :-)

I have found, however, that a 10 ohm, 10 watt resistor on the +12 on one of the drive connectors was a suitable dummy load to power a bare MB. (old-style, of course.)

Thanks! Rich

On the old-style supplies and MBs, it was used as a system-wide reset. It's an open collector, that's held low until the supply decides it's regulating properly, then released.

Nowadays, with these newfangled PCs that don't even have a power switch, I'm pretty much out of the game. )-;

Thanks, Rich

Come on! Don't you remember that big red rectangular button, that we'd tell users was the "Any" key? ;-)

Cheers! Rich

A kind of a compromise than, probably a feature of the new units.

Interesting technique.

3.3V is loaded heavily by the CPU, with power consumption around 100W max. Now that takes some serious current at 3.3V.

--
Siol
------------------------------------------------
Rather than a heartless beep
Or a rude error message,
See these simple words: "File not found."

On Tue, 31 May 2005 02:31:13 +0100, Pooh Bear put finger to keyboard and composed:

I think you'll find that most (all?) motherboards use it. If PG is false, then the motherboard remains in the reset state. Some motherboards have the option of generating an on-board power-good signal.

The original IBM AT had no reset switch. However, it could be reset by shorting the PG pin to ground.

- Franc Zabkar

--
Please remove one \'s\' from my address when replying by email.

On Mon, 30 May 2005 23:54:30 +0200, "SioL" put finger to keyboard and composed:

Sorry, bad choice of terminology. As I pointed out elsewhere, and as the example PSU schematic shows, the one "switcher" gets weighted feedback from both the +5V and +12V rails. I should have written that the Antec PSU "appears to regulate by sensing both the +12V and +5V rails".

Having said the above, notice that in the example ATX PSU the +3.3V rail *is* independently regulated. The technique involves "bleeding off" excess charging current by making one of the rectifier diodes appear leaky, or at least that's how it looks to me.

- Franc Zabkar

--
Please remove one \'s\' from my address when replying by email.

Power good is a signal sent from the power supply, when the voltages are good, so the computer knows when it is okay to start reading bios and writing to RAM. If it started as soon as the cpu gets enough voltage to run we risk reading and writing faulty values in memory.

I built my own PC power supplies when building my first PC's. Old , heavy analog power supplies with modern motherboards.

I discovered that some motherboards needed a fast positive flank on the power good line, it was not enough to let the voltage rise and use the

5V line as power good, as I did for my first PC motherboards.

So I had to build a little circuit which sensed the 5V line, and tripped when it was close to 5V, and sent a positive flank on the PG line.

--
 Roger J.