I'm thinking about a new PC build and I want to have 2 switches to control the power/reset and to allow me to turn on the WC pump while the computer is off. I think I figured out a way to do it, but are there any elec junkies out there who can tell me (a) is it safe? and (b) is there an easier way to do it? Thanks in advance. For reference the 3PDT is two poles with an off position in between (ON-OFF-ON), and the SPDT is two momentary poles with an off position in between ([ON]-OFF-[ON]).
Problem... The first switch (the one you're planning on using to control the pump / fans / lights) is wired to the PSU's 'on' signal wire. No matter where the source comes from, be it the motherboard's power button or your switch, when the power supply turns on, the rest of the computer will turn on as well. The only two ways I can think of to do what you want to do, are... 1. Add a secondary auxiliary power supply (they make small ones that you could hide in a 5.25" drive bay), and set up a switch / relay system that only allowed the aux PSU to turn on when the computer's power supply is turned off. 2. Find some way to disable the ATX connector on the motherboard, possibly using an extension cable and some form of relays or otherwise. Problem with that, is everything else in the computer will still be powered, including the DVD drives and hard drives. Easiest option would be number 1.
One major problem that I can see immediately is that you're connecting Fan-Gnd on the mobo directly to 12v from the molex, that will result in tears. For that matter you're also shorting 12v and Gnd from the molex with the switch in the position as shown on the diagram. Also, if you have the cables plugged into the Mobo and turn the PSU on for what ever reason, even if the PS-on signal never reaches the mobo then the mobo will turn on, you would have to switch every power supply to the mobo, 12v, 5v 5vsb and 3.3v. I'm in a rush at the moment but I'll take a closer look at it this evening. Moriquendi
If all you want to achieve is being able to turn on your pump and fans without starting up your computer you'd be better off getting a wall wart or something like a laptop adapter for the pump etc. What you're suggesting above is an awful lot of work and potentially risk for not much gain. Moriquendi
Hrm, I got that idea about shorting the green wire (PS_ON) from this thread (specifically posts #7 and #10). I was under the impression that shorting the green wire turns on the PSU, but you had to short the power switch to actually turn on the mobo. If that doesn't work then I might as well give up the whole idea of a second switch, really, or maybe just wire them in series so that both have to be on for the on/off/reset switch to work, and ignore the pump part entirely. Small secondary power supplies are so danged expensive, and I don't really have a good place for one in my case. I'm ok with stuff being powered on (as long as it won't hurt them), but this option seems like something that's a little outside of my ability to design, and probably more expensive/complicated than it's worth. D'oh. I redid that part several times (you can sort of see my eraser marks) but I guess I still didn't figure it out yet. Well, I'd rather run the pump off of the mobo, because it has RPM sensing and I can use that to detect a problem. If that means that I can't just use a switch to run it when the computer's off, then I guess I'll have to live with that. It's seeming more and more like I should just get a SPST switch and just put in in series with the on/off/reset switch, so that I can't accidentally hit the reset button (and if I make the switch a key switch, I can keep nosy little brothers out ) (Oh-- if I do this, is it ok that both the on/off ground and the reset ground are tied together? Or should I make it a DPDT switch?) That's where I originally started, but then I thought heck, if I'm already using a switch to lock the power, why not also have one that runs the pump? At the time it seemed easier than unplugging the PSU every time, but I guess I was a little too hopeful... Thanks for your help, all.
Im getting kinda confused here... I'll do a little explanation then maybe you can explain what it is you want to achieve. Reset, Power and PSU-ON are both active low signals, you ground them to make then do whatever it is they do, provided the grounds are all connected (which in a computer they always are, with the possible exception of audio grounds) you can use any ground. Forget the reset switch for the moment, essentially it just tells the mobo to start again at the beginning but without the power sequencing bits, it only really deals with the software. When the computer starts up this is what happens. Parts of the mobo are active from 5vsb standby power. You press power switch(momentary) Mobo grounds PSU-ON signal. PSU starts up, noisy power to begin with so... PSU waits till voltages are stable then... PSU asserts Power-OK signal ( active high, 5v) Mobo starts booting computer. If you want to have a keyswitch to keep out prying siblings then put it in series with the power switch. If you want the reset switch to require the key to be present then put the keyswitch in the ground line after the normal momentary switch then take the ground from the normal reset switch to the same pole on the keyswitch. I dont know of any pumps that are powered from a 3pin mobo fan header, most require more power than the mobo can provide so they run off a molex. If your pump has RPM monitoring then it will probably have a separate3 pin plug with only the RPM line that you can plug into the mobo. Can you link to your pump so I can see what it is you're dealing with? Moriquendi
Start with the last question first: I haven't bought it yet, but I'm looking at the Danger Den CPX-Pro which uses a 3 pin motherboard header. It says it pulls 1.8A, but there's nothing in my motherboard book about what kind of amperage I can get out of the CPU fan header. Even if I don't end up going with this whole two-switch scheme, I was definitely planning on using the motherboard header, so if that won't work this is good to know. Active low makes sense... I thought that maybe the power and reset had to complete a circuit or something. I didn't realize that that's how the power up sequence worked. I guess that other thread I linked was just plain wrong (or I really misunderstood what was going on there). I'm not going to do this, but if I understand correctly, I could skip the motherboard header entirely and just wire a momentary switch between the green wire and ground to use as a power on switch, right? I'm guessing it wouldn't work to power down the motherboard, and that's why there's a separate header. If you are still confused about what I want to do, the actually words in my picture explain it pretty well. I'm probably just going about it all wrong and making it confusing for you, since what I say I want and what my schematic does are two different things! If that idea won't work, or can't be done simply, then I'll just scrap it entirely. What I'll do in that case is essentially take out the top and bottom pole of my 3PDT switch above, leaving a SPST in series with the power/reset. Does that make sense? Or should I draw more pictures? Everything is better with pictures
Haven't seen that pump before but they seem pretty confident that it'll work off a mobo header. The green wire for the PSU needs to be held low to keep the PSU on, a momentary switch wouldn't work. I don't think they ever got that thread sorted out TBH. I don't know what will happen to your motherboard if you feed it all the voltages without telling it to turn on. I'm also not sure why you would want to run the pump without running the computer, in normal operation anyway. When you're filling the system it's useful to be able to run the pump without the computer to get air bubbles out and leak test, other than that the pump just starts with the computer. With the power and reset, yes, a SPST switch in series with the normal switches will work, you will have to have the SPST switch on in order for the power or reset switch to have any effect. Moriquendi
I dunno, it was just a crazy idea I had. I was envisioning a sort of airplane or rocket takeoff: flip a switch to turn on the PSU, then flip the switches for each of the lights and fans, then flip the switch to turn on the computer. The only reason I would want to run the pump without the computer is exactly as you said. I figured if I was going to put in a switch anyway, I might as well make it do more useful stuff, since at the time I thought it was just a simple operation. Oh well, it was worth a shot, and now at least I learned something new about PSUs! Thanks for your help.
No problem and to be honest, I guarantee that within a couple of weeks you'd be bored with having to flip 6 switches to turn on your computer. Moriquendi
Well I wasn't going to do that *every* time, just when I'm showing it off I have the 6 switches in my current computer, and unless I need to adjust something I just leave them where they are all the time anyway.
its a dream just like mine of being able to pilot the Apollo command module with all of those switches and primitive read-outs it truly is amazing what they were able to accomplish with the now cave-man tech
Why dont you alter the design slightly? So that the pump, fans and light can be switched on and off so long as the Master on/off switch (power switch) is in the on position? It would still fit in with your idea for a rocket/airplane takeoff.
I'm not sure what you mean. If you mean I should design it so that I can turn that stuff on before the computer comes on, well, that's what I was trying to do; I just don't know how At least, not without a bunch of relays.
Hey, I just realized it was my thread that was linked in here. Although, I'm not sure the information in my thread is quite applicable in your situation. I've not started the project yet, but have headed in a slightly different direction. I will be powering two systems off of the same PSU using the method linked below... http://www.overclock.net/faqs/112866-how-use-one-power-supply-power.html
See, I'm having a hard time understanding why in that case the PSU doesn't turn on the other motherboard, but in my case it does. The only difference between what I want to do and what that thread is doing is when I split the 24-pin I'm not plugging one end of the Y into another motherboard. But the principal should be the same: I'm shorting the green and ground exactly the same as a motherboard would do it, right? This should be pretty easy to test... all I need to do is short the green wire of my PSU while it's plugged in to the mobo and see if the computer turns on.
I didn't say that it would turn on the motherboard, only that I didn't know what would happen if you gave a mobo all the signals it needs to turn on other than the actual power switch. By the sounds of it Craigbru doesn't know either. It is also conceivable that grounding the ps-on signal (green wire) will damage the mobo if excessive current flows, it's unlikely i think but it could happen. Either way please post your results, I would like to know if it's possible. Moriquendi
There simply isn't enough information out there regarding what is trying to be done here, or what I was trying to do for that matter. Google searches prove to be a little unproductive. As CyberDruid points out in the thread I linked, the motherboard that is not on, will still be powered. I'll admit I don't understand what the negatives might be in a situation like this. I'll find out soon enough myself, as I've got to have a workable solution for my own project. I've got some old hardware I could experiment on...
Ok, I shorted the green wire while it was plugged into the mobo. I learned 3 things: 1) The motherboard does not turn on 2) The CPU fan and GFX fan turn on, at maximum voltage. On second thought, it probably would have been wise to unplug the power from the GFX card... but then again it might still receive power through the PCI slot, so who knows. 3) That noise I hear every time I move my computer around is a motherboard screw. It's been hiding somewhere in my case find for over a year now. Next experiment: trying to find the hole that it came out of. So I don't know whether or not it's bad for the motherboard, but seeing as how the CPU fan and GFX fan both got power, it's probably a big unnecessary risk.
