Potentiometers to adjust the PSU output?

O

Ozone77

Limp Gawd
Joined
Oct 14, 2004
Messages
326
I have a 1U 300W (Skyhawk?) power supply. I noticed that it has two small holes with mini-screws inside the case, I suppose these screws are just potentiometers and you can adjust the output voltage/current. But This is just a guess, does anybody know?

Can someone tell me how to adjust these propperly? I already messed with them and they are not at the factory default settings and I don't wan to fry my yet-to-be-bought mobo.. ok, ok, so I am exagerating, but I am sure my memory will not like to be overclocked as much if the +5V line is not supplying enough juice or some little detail like that.
 
It'll probably adjust voltage, but not current. Anyways, if you're going to mess with these pots further, do it while the power supply is not connected to a computer. Do you know how to turn on the power supply without a motherboard? Look on the ATX power connector, and short the green wire to any black wire. When you do this, connect a fan or floppy drive or something to act as a load.

Okay, now, connect a multimeter or voltmeter to the voltage line that corresponds with the potentiometer (it's labeled, right?). Slowly adjust the potentiometer so that the voltages are back in line.

That's the general idea. Just make sure to ask questions, preferably before problems come up.
 
The motherboard regulates voltage to your cpu, memory, pci, agp, etc...

If you change the output voltage of your PSU, there will be no real change there. The only difference is that the output lines of the PSU lines are closer to spec. This makes the job easier for the motherboard. When you adjust the pot there, you are really just helping the PSu account for the unique load of your system.
 
xonik said:
It'll probably adjust voltage, but not current. Anyways, if you're going to mess with these pots further, do it while the power supply is not connected to a computer. Do you know how to turn on the power supply without a motherboard? Look on the ATX power connector, and short the green wire to any black wire. When you do this, connect a fan or floppy drive or something to act as a load.

Okay, now, connect a multimeter or voltmeter to the voltage line that corresponds with the potentiometer (it's labeled, right?). Slowly adjust the potentiometer so that the voltages are back in line.

That's the general idea. Just make sure to ask questions, preferably before problems come up.
Click to expand...

Gotcha, seems simple enough. They are not labled, thats why I thought they could be keeping the PSU together (I was taking it apart at the time to look at it) I suppose one is for 5v and the other for 12v. I will connect a 12v fan and a 5v fan to get both sides of the circuit loaded and then use the mulitmeter to adjust voltages...

wait.. Voltmeter has to be in paralel, should I measure the voltage across the fan or across two empty pins? ...Two empty pins, right?

Thanks
 
All the wires for a given voltage are connected together inside the PSU, so to measure 12v, you could connect one lead to 0v, and the other to an empty 12v wire, or the 12v wire you have connected to a fan, it makes no difference.
 
A parallel configuration means that the voltage across the fan must be the same as that of two empty pins (as long as the two pins are +12 V and ground).
 
If the psu can be modeled as a thevenin equivelant.. the resistance of the fan would affect the power output, wouldn't it?
 
A switchmode powersupply is more complicated then just a Thevenin style model. The PSU is actively adjusting its output voltage. In particular some PSU won't output proper voltage without a minimum load being connected.

I would suggest you connect the fan and use your multimeter to adjust the voltage to slightly above the nominal voltage. For example set the 5v line to maybe 5.1volts. This will hopefully account for any voltage positioning that the PSU circuitry uses. Once there is a real load on the PSU the voltage should drop down closer to 5volts again.
 
aL Mac said:
If the psu can be modeled as a thevenin equivelant.. the resistance of the fan would affect the power output, wouldn't it?
Click to expand...
No. The resistor that accompanies the ideal voltage source in the Thevenin model represents the voltage source's output impedence, whereas the fan could more accurately be considered a load. However, it's both a resistive and an inductive load, so there would be an inductor also involved.

The load is a necessary function for the voltage regulation circuit. The regulator is a feedback system, therefore it reacts to changes in the load, but without the load, it will probably not function.. Think of it this way:

If you want to ride your bike at a certain speed, you have to vary your pedalling speed based on the terrain you encounter. To maintain a certain speed, you may have to pedal faster or slower to do that. Still, you have to be aware of your speed in order to maintain a given speed, obviously. You have to receive feedback from the environment to judge your speed and adjust accordingly.

The switching regulator in your power supply works in a similar way. Inside, there are capacitors and inductors which attempt to hold charge at a certain voltage. Now since they are non-ideal, these devices lose charge over time if they are not fed more charge. The control ICs open and close large power MOSFETs which act as valves to let charge in when opened, and block charge when closed. This keeps the voltage level at a constant parameter.

But still, a feedback system is needed to adjust to the varying load levels. This is what separates a voltage regulator from an ideal voltage source and the Thevenin equivalent model. The ideal source produces a perfect voltage regardless of load, and requires no feedback. A real power supply, however, makes use of feedback to produce an output that adjusts to the decidedly non-ideal nature of real electrical components and varying loads.
 
hehe, seems like I opened a can of worms..

Ok, I shorted the green wire to trick the PSU in to the on position, I connected a Floppy (but I never did see its light turn on) and I connected a fan.

One of the "screws" modified both power outputs, so I left one at 11.78 and the other at 5.35. I left it like that, becasue to get 12v I had to get 5.5v on the other line. The other screw did not seem to adjust anything, I suppose it could be the 3v circuit, but I dont know how/where to measure that one. Any Ideas?
 
someone tell them their real values via a hardware monitor.. I can't get mine working..


I know everyone's is different but that way he can tell someone's elses values and just make his close.
 
11.75V and 5.35V are pretty close to what it should be. If anything I would bring the 5V down to say 5.1V. There usually isn't anything too critical running of the 12V line so having it a little low isn't a problem. The other pot probably does control the 3.3V, a suitable load for this is a car tail light bulb around 20W or what ever you have lying around.
 
Sorry to nag you guys, but it is so much faster just to ask you which one is the 3/3.5v cable than having to hunt down the 20-pin ATX cable specs.
 
Ok, so I adjusted the 3.5V rail to 3.45V. Since I had to choose, I made the 5V rail 5.15 and the 12V rail 11.35V.

I cut and covered (to prevent any shorts) the cables for a molex set (one yellow, one red, two blacks) and for the sever plug (several orange and several black).

Finally, I removed the power switch within the PSU (on-off switch for one of the wires from the power socket into the PCB)

What I noticed the other day is that the PSU was quite warm even though the computer was off (had been off for over a day).

Could I have a shorted wire somewhere? did I cross the "Hot" and the "neutral" from the power socket to the PCB? what else could do this?
 
hot? the PSU is constatly creating a 5vsb, which is alwase powering the motherboard, even when the power is "off" (unless its unpluged, in which case if its getting hot you have discoverd free energy, PM me now!!!). but it shouldent be getting above, say, 10 degrees above ambient
 
The +5VSB supply in computer supplies can run quite hot. After all, it's running in a closed metal box with no air convection, when your computer's off.

Don't worry about it.
 
Ozone77 said:
Sorry to nag you guys, but it is so much faster just to ask you which one is the 3/3.5v cable than having to hunt down the 20-pin ATX cable specs.
Click to expand...

Googling for "ATX pinout" will return plenty of results. In fact, this works for most major pinouts.
 
It must be the "1U" PSU style (everything crammed inside), all standard PSUs I have had before were not warm to the touch when they were turned off.. Thats a relief, tnx
 
You must log in or register to reply here.
Back
Top