Interpreting PC Power Schemes 3

[herrEinsamkeit]

Hi all, i'm attempting to design+build a regulated power supply for an old computer, and i'm trying firstly to make sense of the old stock supply, and how the load components fit into the scheme. Taking the hard drive for example, the faceplate says:

5VDC --- 0.66A
12VDC --- 0.24A

and the power cable setup is as such:

R B B B
e l l l
d k k u
+5V COM COM +12V

So am i to assume that there are two diff. major circuits, working out something like this?:

? ?
/ \ / \
| | | |
5 G G 12

and so the impedances would be 7.6 and 50. ohms, respectively. I'm guessing that the 12V is for the motor and servos and whatnot, and the 5V is for the logic? But I thought that most ICs could tolerate no more than ~20mA at their pinouts.
Thanks for any help. Would esp. appreciate direction towards resources dealing with power supplies and/or power schemes for components.

 

[itsmoked]

herrEinsamkeit;

You are sorta correct on the two circuits. They aren't really *separate* as they do indeed share the grounds.

Usually they are RED/BLACK/BLACK/YELLOW though if memory serves me correctly. BLUE is usually -5V

The currents required have nothing to do with the logic gate ratings.....

Yes! The 12 volts is for the platter motor. The 5V is for the logic and drive-to-mutherboard communications.

The really big question though is why on Earth would you EVER consider building a PC power supply?!?!? They are probably the lowest cost for what you get item sold. They are fairly complex too. They have timed output signals to tell the motherboard "power ready" etc. etc.

 

[herrEinsamkeit]

(ooh crap, didn't know about the timed output deal)

but replying firstly to your question, I'm thinking of building it for the same reason that I'm making a gravity-driven rotary grinder out of wooden dowels and custom cut-and-filed gears and pullies: <i>there is simply no better way to learn.</i> Using rough figures, I'm thinking this system could top out at maybe around 1600 rpm. Now I could buy a good electric 3450 rpm rotary grinder for about $25, but what's to be learned that way? Sure, I payed attention in Physics class and I <i>know</i> about torque and angular motion well enough, but it was only once I started building and designing this sucker that I really started to understand either on a gut level.
Uh but that being said I wasn't aware of the timing scheme, I guess I'd always though that was coordinated once power was received at the motherboard. Well, I won't abandon it so easily. Do you happen to know of resources that detail this? Maybe there's a commonly available output-timing IC that would fit the bill.
But at any rate (I've really responded in a backwards fashion, haven't I?), thanks for the info. One of my power cables is red/blk/blk/yel, but all the others are as above.
Could you please elaborate on "The currents required have nothing to do with the logic gate ratings....."? I'm confused on that point— isn't there a lot of logic circuitry involved in a hard drive? Clearly I'm just misinterpreting what the values mean, or something like that, but elaboration would be helpful.
Oh an a minor note, no matter how haphazard my efforts may seem, I do actually try to be rather methodical. Right now I'm just trying to evaluate the system as a whole, figure out all of the major principles, and then I'll probably build some test modules like individual filters, PWM units, figure out power factor correction. If, by the way, you know of any resources for such things, that would be awesome. Thanks!

 

[OperaHouse]

Think of it this way. If each IC consumes only 2ma and you have 300 of them on a circuit board, the supply will need to supply 600ma. Whatever is connected to the IC more often determines what current is consumed. The IC will not draw more current just because the supply is capable of higher currents.

 

[herrEinsamkeit]

oh. haha. jeez, the more i think of that the more obvious it seems. thanks.

 

[itsmoked]

Stabbing at more of your questions..
The cables that are R-B-B-Y are for hard and floppy drives.
The two cables with a mix and blue wires are for the motherboard.

The drives have IC's with millions of logic gates in them.
This is what's consuming the 0.6A

 

[herrEinsamkeit]

I'm thinking this computer must have been some sort of exception. We bought it as a refurbished NEC Ready 9522 sometime in the earlier 90's, it couldn't have been any earlier than '92 or '93, because it runs off of a Pentium. I know what you mean with the R-B-B-Y scheme though, I've seen that in photos of systems online, and I believe also in the Dell system I'm using right now. For some reason NEC felt like being defiant with their color scheme; does anyone know if maybe this is just an older standard, something having to do with the AT power supplies? The Dell definitely uses an ATX, but I'm pretty sure that the Ready uses an AT; I can't locate any sort of "stand-by" feature, and all of the connectors look different from ATX connectors I've seen online. But yeah, all of my cables are verifiably red/blk/blk/blue for some reason, and the P1 and P2 (motherboard) connectors are blk-blk-yel-blu-red-whi, red-red-red-org-blk-blk (respectively). I can't spot any 3.3V connections either, further convincing me this is AT fare.

 

[johnwm]

Be aware that Dell uses a non-standard propietory power interface from PSU to the motherboard. Plugging standard PSU to Dell motherboard (or vice versa) is frequently terminal. google for Dell Power for more details. As far as I recall HD and FD power connectors have been R-B-B-Y back to the early AT days in the UK.

The sequence for powering up of the different rails is also specific on many multi-rail chips.

Good Luck
johnwm
________________________________________________________
To get the best from these forums read faq731-376 before posting

UK steam enthusiasts:

http://www.essexsteam.co.uk

 

[herrEinsamkeit]

weird. Well, unless I'm developing some rare breed of color-blindness, these suckers are what they are, and are marked as I've mentioned above. In response to johnwm, I'm not certain as to what my Dell scheme actually is, I only opened it up once or twice, and neither time was I concerning myself w/ the power supply. But I'm glad I mentioned it because that's an interesting little tidbit, thanks. Oh and a correction to my previous post, NEC had nothing to do with the color scheme, it's an Astec power supply, manufactured 1994. Come to think of it, the actual name of the company is "Astec Custom Power", maybe they chose a different scheme to distinguish themselves, a proprietary scheme similar to Dell's? Just food for thought.

 

[herrEinsamkeit]

Sorry, another correction, this time to my most recent:

"maybe they chose a different scheme to distinguish themselves, a proprietary scheme similar to Dell's?"

what I meant to say was similar as in both companies wishing to have proprietary schemes, not similar as in have a scheme actually like Dell's. Sorry. Is there any way to edit a post after it's been submitted? I've been on several boards where that's the case, it's really quite nifty.

 

[mrkenneth]

Astec probably just used blue instead of yellow for the 12V rail. Perhaps they ran out of yellow wire.

Since it is an AT power supply, I do not think there is any "smart" circuitry inside that allows WOL or such, so some power regulators with enough current should work in this case.

Remember too, that the switch is directly connected to the mains, unlike newer ATX power supplies that use a low-voltage rail for monitoring the power switch's status.

http://www.mrkenneth.com

 

[walker1]

One way to get an idea about how old a circuit board is:

Look at some of the chips. They often have manufacturing time stamped on them below (or above) the type numbering.

Ex.
"9534" would mean week 34, 1995. (I.E. late August or so 1995

Of course, the board assembly is newer than the newest chip, but usually not more than a few weeks or months.