Question about PC PSU

[Cube777]

I'm planning to have all the normal voltages to pick from (12V, 5V, 3.3V and -12V), and one separate plug with an adjustable voltage which can handle up to 5 Amps. Thank you for all your help! I can't wait to get started on my PSU, especially after I saw your Blue Ice Machine +1 coming your and Eud's way :beer:

 

[Meatball]

Resistors are just fine as long as they are rated for the power the absorb. The fans inside the PSU will keep them cool - but really they shouldn't get hot. You can also put your minimum load on the 12V rail if its more convenient. The 12V rail is derived from the 5V rail - so when you load the 12V rail - the same switcher for the 5V rail is also loaded down properly.

Yes, all grounds are common, all similar voltage lines are common - it saves circuitry and surface area and cost for the manufacturer to do this way.

Oh yeah, and don't try the use two in series. That doesn't work.

Parallel works just fine. If the two (say... 12V) rail are slightly different voltages (+- .5V difference), the average of the two potentials should appear at the node when you connect them in parallel.

 

[honeyx]

I'm planning to have all the normal voltages to pick from (12V, 5V, 3.3V and -12V), and one separate plug with an adjustable voltage which can handle up to 5 Amps. Thank you for all your help! I can't wait to get started on my PSU, especially after I saw your Blue Ice Machine +1 coming your and Eud's way :beer:

If I were you I would buy a adjustable universal notebook psu with 100W output. They are even cheaper on the bay than the secound hand PSU you were talking about. All you need is to modify it a bit and of course an Amp- and a Voltmeter to turn it into a adjustable desk psu.

 

[Meatball]

Laptop PSUs have a single supply rail at 19V which is not all that useful.

Do you commonly modify them into adjustable PSUs? I have not seen that done before.

SMPS units by nature have strict load requirements when changing the duty cycle to the switcher.

 

[Cyparagon]

Some are "adjustable." There are more than two leads and the voltage is set by a divider on the DC jack that plugs into the laptop. In my experience, they only work between 12V and 25V. they get mad if you try to set them below 12V. I'm sure there's a way to change this by opening the thing up, but that would be the case for ATX PSUs as well.

 

[LaZeRz]

I can supply pics of my diy adjustable voltage pc psu if needed. Ive got all my circuits in a project box.

 

[Zeebit]

That would be great. Please post.

 

[Meatball]

I'd like to see a schematic or something if you've got. That would be great! Thanks!

 

[LaZeRz]

I have a whole lot of information on the build that I need to dig up. (resistors, pot values, current regulator etc.)

When I find them i'll post them up.

From left to right, I have 5v, 12v and adjustable voltage.

 

[Cube777]

You can also put your minimum load on the 12V rail if its more convenient. The 12V rail is derived from the 5V rail - so when you load the 12V rail - the same switcher for the 5V rail is also loaded down properly.

I've also read that the 3.3V rail and the 5V rail also shares a common power output, does this mean if I load either the 12V, 5V or 3.3V rail, all of them will be loaded properly?

@Lazerz Looks great, do you think there is a possibility that I would be able to fit all the regulating circuitry inside the PSU's box?

 

[honeyx]

Laptop PSUs have a single supply rail at 19V which is not all that useful.

Do you commonly modify them into adjustable PSUs? I have not seen that done before.

SMPS units by nature have strict load requirements when changing the duty cycle to the switcher.

As Cyparagon already said the adjustable can be set to certain voltages. Some are using dividers and others just a multi switch with preset resistors. Usually they can be adjusted between 12V and 25V. Most of them are also having a pot inside, so you can use this to adjust the voltage to your needs. Most common workbench PSUs also offer just one output, but it´s easier to realize such a PSU by using a notebook PSU than a ATX PSU. The PC PSU´s do have fixed voltages and it´s way harder to modify them to get more than 12V out of them. A lot of them are even using a simple 7912 inverter to realize the -12V. Last but not least they need some load to get stable voltages, so you are wasting a lot of power.

There is also the protection curcuit inside the ATX PSU´s that will make trouble if you will try to get higher currents by using other loads than PC parts.

When I was testing ATX PSUs for a hardware site, we looked for a way to get them working at their full load to measure the consumption, but it was almost impossible to use other loads than a ATX mainboards with some powerhungry videocards. We tried bulbs, resistors and a 12V to 220V DC/DC converter. The bulbs and resistors were a no-go. Just the converter worked up to 150W, but then the protection circuit kicked in. So the point is, ATX PSU´s are designed to only work correctly with switching loads like the step down converters used on mainboards and videocards. Everything else will be interpreted as a short circuit at higher loads and the PSU will shut down.

There are a few ATX PSU´s that can be modified for higher voltages and higher currents, but for a workbench PSU this is usually overkill.

A notebook PSU can be also used for small loads like turning them to a constant current drive for high power LEDs, so a 100W PSU should suit your needs best. All you will need is such a PSU, something like two LM338, two pots and to LCD panels.

 

[Cyparagon]

I sense a big fight coming on. This gentleman cannot admit when he's wrong, and will literally fight facts until he's banned. We'll see if things have changed I guess, because someone's gotta do it...

so you are wasting a lot of power.

Several watts of standby power for something you don't use all day every day is actually incredibly cheap. If we assume 20W at two hours a day five days a week, that's less than 10¢ per month.

There is also the protection curcuit inside the ATX PSU´s that will make trouble if you will try to get higher currents by using other loads than PC parts.

The protection circuit detects current. It does not magically know what you've hooked up to it and get mad about it. Laptop PSUs have over-current protection too.

Everything else will be interpreted as a short circuit at higher loads and the PSU will shut down.

I suppose next you're going to tell me the guy that had his outputting a regulated 23A into a short was practicing witchcraft?

Perhaps you chose the wrong value resistors. Bulbs might not work because high power bulbs are almost a dead short until they heat up. If you were to soft-start them, they'd work.

All you will need is such a PSU, something like two LM338, two pots and to LCD panels.

LEDs are ~3.6V and the laptop PSU usually cannot be set lower than 12V. For something simple like an XML-T6, That's 25W of waste heat that the regulator won't be very happy with.

Most of them are also having a pot inside, so you can use this to adjust the voltage to your needs.

This... I'm pretty sure you just made this up.

 

[honeyx]

I sense a big fight coming on. This gentleman cannot admit when he's wrong, and will literally fight facts until he's banned. We'll see if things have changed I guess, because someone's gotta do it...

Is saying who? The one who is catching each single word and starting the fight? Not again and not with me sir. Look for an other victim. But some things obviously didn´t change here....... one of them is obviously you......

 

[Cyparagon]

Well, if that's all you want to respond to, I think we're done here. Everything went better than planned

 

[Meatball]

I've also read that the 3.3V rail and the 5V rail also shares a common power output, does this mean if I load either the 12V, 5V or 3.3V rail, all of them will be loaded properly?

Could be the case with some, as some ATXs run all lines through the halfbridge flyback transformer for voltage conversion - but I would never guarantee it. I know for a fact that some have a separate 3.3V conversion sub-circuit. I would buy a couple extra resistors just in case. Start with whatever is easiest for you to load down - and load the rails until everything works like it should - without any excess heat. Putting no load on a switcher can be bad for it - but again it depends on the design.

If I'm only using the 12V rail, I snip all other lines and just load the 12V rail.

If I'm using all the rails, I load the 12V rail, the 5V rail and the -12V rail. Plan on doing this. Load them with something 10-20% of their max current rating. For instance, the 12V line should be able to regulate just fine with 500mA flowing at a minimum. The 5V at around 1A and so forth.

Assume the +12 and +5V rails are not interdependent and just load them each down. This will surely load the 3.3V rail as well. The -5V supply is derived from the -12V rail so just load down the -12V with ~20% of its max current.

If you get all the resistors to fit into the case, you won't even notice them anymore. It will work just like any other high current voltage regulated supply.

 

[honeyx]

Well, if that's all you want to respond to, I think we're done here. Everything went better than planned

Yes, we are done. I looked through your plans. They are very simple and obvious. Many things change with time but you will never change....

Just am suprised it took you so long to reply.

So why arguing with you? Even If I would show you a photo of a pot inside a notebook PSU, you would try to make me rediciolous and probably claim this is not a pot but a CO2 laser mirror. It´s not worth arguing with you.

I also told you in the past I´m not german, though I´m living there. So belive in what you want, even in witches.