• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Using +5V Standby in Mods

bisby

Weaksauce
Joined
Aug 28, 2007
Messages
110
So i was considering using +5V standby for quite a fancy little thing... I wanted some military switches (with LED indicators) that lead up to the power button (which would only be usable if all military switches are engaged).

I've been reading about ATX 20pin pinouts a bit... +5vSB seems like it should do the job. The stuff I've read claims anywhere from 10mA to 1A is availabe on this line. The supply I plan on using for this job is listed at 2A.

Seems like it would be enough for any ICs and LEDs i might need to make it fancy.

Which brings me to my real question: how many amps does an average motherboard draw from +5vSB while "off"? How much of these 2A are gonna be usable in my little project.

This all sounds feasible in my head. I have enough background in circuitry stuff to build the necessary LED features and what not... but I have no idea about drawing power out of a power supply (other than molex connectors)... especially when the power supply is "off".

Any help would be appreciated
 
When it's off, your computer draws practically nothing off the 5vsb line. If you're using WOL/WOR or have your BIOS set to start up the computer on KB/mouse activity, you'll get slightly more, but you'll still have the vast majority of that 2A to work with. Just tap off the 5VSB wire and a GND, and go to town.

As far as the switches are concerned--the bulbs inside switches will be fine if they're LEDs, but may draw too much power if they're incandescent bulbs. And I hope you get four pins on those switches, since you'll want the switch contacts separate from the contacts for powering the lights. You'll just need to put the switches all in series with your power button, and you'll be good to go.
 
awesome. thats exactly what i wanted to hear. ^_^

I pretty much have it figured out what i plan on doing. Using flip flops, using Q to power a red LED and Qbar to power a green LED. Might go the other way around or whatever (i havent drawn my schematics yet), but in general, i wanted to make sure there would be enough amps for a few minimalistic ICs and the LEDs... sounds like ill have way more than enough even if i get sloppy ^_^

thanks again
 
awesome. thats exactly what i wanted to hear. ^_^

I pretty much have it figured out what i plan on doing. Using flip flops, using Q to power a red LED and Qbar to power a green LED. Might go the other way around or whatever (i havent drawn my schematics yet), but in general, i wanted to make sure there would be enough amps for a few minimalistic ICs and the LEDs... sounds like ill have way more than enough even if i get sloppy ^_^

thanks again
So how, exactly, is this supposed to work? Red LED lit until you flip the switches, then you get a green LED and can turn on the computer?
 
G G G G
S S S B
R R R R

where G is green LEDs, R is red LEDs.
S is switches (DPST) and B is the power button.

When S is off, Red is lit. when S is on. Green is lit. when ALL S's are on, the G above B is lit. If ANY are off, it will be red.

I envision it like some sort of futuristic tank (if you ever played steel battalion, thats a pretty decent example)

flip switch, flip switch, flip switch, ignition.

I have the schematic for the LEDs completed and somewhat simplified and it has worked using test components. I assume the power button part will be as simple as attaching it to the switches to break the wire when off. I opted to use NAND's and open collector hex inverters to designate which LED is lit, rather than flip flops. I'll convert my schematics from pen/paper to computer and post them here later.



 
I'm a bit confused by your schematic--the LEDs appear to be wired backwards, methinks, and only four are shown. Are those bi-color LEDs? If you have it working with real parts, though, no big deal. I have to ask, though, why you are using NAND gates as inverters on top of the regular inverters. Or are you just using the vanilla inverters as buffers?

Did you ever consider doing this with DPDT switches? That would have eliminated the need for IC's almost entirely--all you would need is a single buffer.
 
yeah. its pretty sloppy. im just using the vanilla inverters as buffers. They are open collectors. Had some troubles trying to get the regular NANDs to power the LEDs.
I only drew 1 of the switches in the diagram. the top part is replicated 3 times. (so there are a total of 8 LEDs).

Im just sloppy i guess. ;) Im still early in my education and I find it fun to design unnecessary stuff... because I can. I have a few schematics in a notebook involving flip flops that would work. Although I must admit... I hadnt considered doing something as simple as that. Lol ^_^
 
Here's how you could do it if you just had DPDT switches (I left out the current-limiting resistors because I was too lazy to put them in ASCII schematics, but you should still use them):
Code:
MB+----||---------||--------||-------||------GND (or other side of MB header)
      +++        +++       +++   |   ++
      SW1        SW2       SW3   |   BTN
      +++        +++       +++   |
      /|\        /|\       /|\   +----to buffer for final "ok" LED
     / | \      / | \     / | \
    |  5V \    |  5V \    | 5V \-GRN LED
   RED    |   RED    |   RED
   LED   GRN  LED   GRN  LED
         LED        LED
 
That IS a lot simpler. >.<
the switches were the only thing I hadnt acquired yet. The ICs were easy to get. 50 cents a piece. so if i dont use them here, I know i can find use for them somewhere else.

I'll look into this for sure! Seems that DPDT are significantly more expensive, but also significantly easier to find (not being able to find DPST switches was one of the reason i didnt get them yet)
 
Actually, I just thought of a better way to wire it. On the top half (the side connected to the motherboard power button header), do this:

Code:
                      5VSB
                       |
                       |
                      RES
                       |
                      LED
                       |
GND----SW----SW----SW-----BTN----MB header
This way, you won't even have to use a single IC! See, the way the PSU works is that it waits for the motherboard to pull the PWR_ON signal to ground, which is typically done with the power button. Any ground will do. Several of the pins on the motherboard are grounds, just for convenience--it lets the case manufacturer to wire both wires from the button into a single connector and keeps things cleaner.

So if, by closing the first three switches, you connect that line to GND, it means that the final LED will be lit, and the signal from the motherboard will get pulled low when you push the button. Does that make sense?
 
indeed! I was thinking about how to keep the power button line clean, because I wasnt too sure of that part. but the fact that it is ground makes sense...

But that makes it have a green led... still dont have a place to wire in a matching red led without logic?
 
indeed! I was thinking about how to keep the power button line clean, because I wasnt too sure of that part. but the fact that it is ground makes sense...

But that makes it have a green led... still dont have a place to wire in a matching red led without logic?
Ah, I somehow missed the red LED below the button. Does that mean that when you flip the final switch, you see two red LEDs turn off and two green LEDs turn on?

In that case, you'll have to buffer the red LED. Fortunately, the ATX spec is helpful for your cause. In order to turn on the PSU, that pin on the motherboard has to be pulled lower than 0.8 volts, which is fine with the circuit I have above. And in order to be off, that pin has to be above 2.0V and less than 5.25V. So here's what you do: add a pull-up resistor from the 5VSB to the left side of the power button (in parallel with the green LED and its resistor). And then tap off that same line into a buffer for the red LED. That means that as long as any switch is open, that pull-up resistor will keep that line at 5V so that you get a good signal into your buffer. And as soon as all three switches are closed, that line will pull low and the green LED will light up.
 
Back
Top