Ever tested your power supply unit with a multimeter?

[Antipodal]

¿Are you interested on the subject?

Though Dewbie's message on last August 14 05:17 UTC @ http://www.murga-linux.com/puppy/viewtopic.php?t=80209 has provided me with a certain amount of hope, I haven't overcome yet the shocking effect of reading some of the posts of the thread that's located @ http://www.murga-linux.com/puppy/viewtopic.php?t=75262 .

Because of all this, until I don't discover the sort of responses this message receives (if it ever receives any) I don't think it's wise to add anything else to it.

Cheers

[Dewbie]

Though Dewbie's message on last August 14...has provided me with a certain amount of hope

Uh-oh!
For ATX, click here.

[8-bit]

A desktop power supply will not put out any voltage until it receives a "power good" signal from the computer.
There are ways around that.
But make sure you have a load of some sort connected to the power supply as I have read somewhere that trying to run one with no load can damage the power supply.

[nooby]

What about Laptops? The Battery died on one of my laptops
and it refuse to start despite me have ac/dc power connected.
It is an ASUS L3800
Other laptops works with just the power supply needing not battery.
Can some laptops have something similar that the mother board
never start up if the battery is not high enough to give a signal.
Maybe I need to connect a kind of fake battery that gives that signal?

[Bruce B]

I've tested some wires to verify they had the right voltage. But that was sort of quick and dirty, meaning nothing thorough.

I have a watt meter and by using it and watching it over a period of time, I can learn more than by using a volt meter alone.

But my reason for posting is to explain something that I have observed and not yet discussed. Suppose we buy a 300 watt power supply. We presume there is always about 300 watts available, but there are some exceptions.

A power supply might be semi functional. It can fluctuate in the amount of watts it can output. For an example, it can deliver enough watts to keep the computer running for minutes or even hours or days. Then it's wattage output fails and the machine acts up.

~~

Here is what I basically do with an old power supply.

1) smell it for burning. the coating material on the windings is, I suppose, some kind of shellac and when burned leaves a distinctive smell. If it smell burned, then condemn it.

2) open it up and clean it out. especially the cup sides of the fan. also oil the fan motor but you'll have to figure out how to do it.

~~

Summary: It is my theory that a power supply is not necessarily 'it works or it doesn't' kind of device. It can behave erratically, according to my theory, and this can make it harder to test some power supplies.

Don't drive yourself crazy. If you suspect it might be an erratic behaving power supply, set it aside and install another and see if the symptoms disappear.

~~

[sickgut]

Dont most BIOS show you the voltages that are being generated by the power supply?

ATX power supply can be made to switch on and off or stay on all the time without being connected to a motherboard simply by shorting 2 pins/ wires.

Power supplies degrade with age. Other than CPU fans, video card gpu fans and dust clogging things and making them over heat, power supply degradation is the next most common cause of a computer failing thats if you dont count the motherboard rechargable battery wearing out and no longer able to keep accurate time and BIOS settings while switched off... as a fail.

Humidity is a major killer of everything else like ram and motherboards. Humidity can affect a power supply as well and if you see and leaks from the capacitors in the power supply then its most likely humidity.

I had collected 6 compaq desktop computers and created a lan with them all so i could invite people over and play games and that was fine. Then i didnt use them for 2 years and they just sat there in my shed in the humidity and when i tried to turn them on, none of them worked. The motherboards where noticeably rusty and the capacitors on the motherboards where leaking.

Sorry ive dribbled on a bit, too much hyper drinks today.

[Antipodal]

Thank you guys
I wasn't expecting so many posts in this thread and I only entered the forum with the purpose of sending Dewbie this message:

Hi Dewbie
Thank you for your concern!

For ATX, click here.

I'll take a second look at that site after I end this message.
Sorry for not being very explicit in my previous message.

Six years ago, when I was selecting the components to assemble a desktop PC, I read quite a lot about PSUs. And from what I can recall, the advice was "never test a PSU with a multimeter unless know how to load the PSU properly because you are risking to ruin the PSU and other components of the system".
At that time I trusted the manufacturer, never tested it with a multimeter, and things worked OK. But lately, since I'm trouble shooting a system that won't boot, and I only have 2 options left (the failing device is the motherboard or the PSU)my interest has re-grown. Because I have only found one place where they frankly discourage you from testing a PSU with a multimeter, I have followed the suggestions of another article that encourages you to do it but the results are quite puzzling.

Do you have any personal experience with these sort of tests?

Antipodal

But most of the things I have written to him can also apply to you.

However, I will carefully read your messages, ponder over them and post my comments as soon as possible.

Thank you very much

Antipodal

[Dewbie]

Antipodal wrote:

Do you have any personal experience with these sort of tests?

I've never used a multitester on a power supply.
Only one PSU has failed on me; it didn't pass the green-to-black jumper test, so that was easy.

I only have 2 options left (the failing device is the motherboard or the PSU)

Can you test that PSU with another (healthy) motherboard?
(One that uses the same kind of power supply.)
This would be an easy way to rule it in or out.

But first, inspect the capacitors on the motherboard.
Failing "caps" can cause all kinds of voltage-stability problems.
Details here.

[Softwaregurl]

Just a couple of my own thoughts.

Yes on all the capacitor stuff. This predates most modern computers but if your ever inside a tube radio replace to waxpaper caps with electrolitics. The wax paper caps are so old now you can expect them to fail and take other components with them. I was inside someones Stromburg-Carlson shortwave reciever and a shorted cap had taken a resistor with it and was overvoltaging the filament in one tube. Also look for underrated caps. These are a failure. Putting 160 volt filter caps on rectified 120vac line will fail just about when the warrenty fails.

In small power stuff it is more common to loose a power transistor or ic then the xfmr. Burnt smell and or any transistors that have part of the epoxy case blown off or a whole burnt in them are a bad sign.

Most of my experience has been with older power supplies. Older ones had to have a load across 5v before any other voltage was produced. All to other voltages were slaved off of the 5v oscilator or some such noise. personaly I like 110v light bulbs as a dummy load. They are available in a variety of values, can be easily interchanged using a socket in the test circuit and wont burn up like a ceramic resistor if overloaded.

but my biggest piece of advice: If you dont know what your doing, don't. My experiences are not typical. I gave up on trying to count the number of times I accidently got hit by 110+ volts when I was 16 and I have only suffered severe brain damage so far.

[RetroTechGuy]

A desktop power supply will not put out any voltage until it receives a "power good" signal from the computer.
There are ways around that.
But make sure you have a load of some sort connected to the power supply as I have read somewhere that trying to run one with no load can damage the power supply.

Yup.

Even the old ones (AT -- pre-ATX) required a load to run. Otherwise they would just shut down.

I don't know if modern HDD draw enough to keep a power supply on, but in the old days they did. So you could test a power supply with just and old hard drive.

[Antipodal]

Ladies & Gentlemen:

Though your posts have been full of useful and interesting information and experiences I am now convinced that (because of my poor knowledge about lots of things) a multimeter is not the best tool to rule out if the problem in booting my computer is in the PSU, the motherboard or in both.

Again, thank you very much!!

Antipodal

[Bruce B]

Ladies & Gentlemen:

Though your posts have been full of useful and interesting information and experiences I am now convinced that (because of my poor knowledge about lots of things) a multimeter is not the best tool to rule out if the problem in booting my computer is in the PSU, the motherboard or in both.

Again, thank you very much!!

Antipodal

I think the best thing to do or to have done is let the helpers help you by using symptoms which are; a problem with booting.

[Antipodal]

Thank you very much for your last message, Bruce
I was about to begin a new thread on this forum when I discovered your post.

The following may seem a foolish question or a joke but it is not.
Since you live in "The Peoples Republic of California" and I live on "the other side of the world" I want to be sure I'm not falling intocultural misunderstandings.

I think the best thing to do or to have done is let the helpers help you by using symptoms which are; a problem with booting.

Are you suggesting I should post or should have posted in this forum the symptoms that are related with the problem I have with booting my computer in order to make things easier to those who want to help?

I look forward to hearing from you
Best regards

Antipodal

[Bruce B]

Whatever I wrote was sincere. First, about the People's Republic of California.

Some settlers declared independence from Mexico and named Califoria a Republic.

Almost concurrently the United States 'bought' California and many other states from Mexico.

After this 'purchase', California became a territory of the United States. Shortly thereafter it became a state of the US.

It kept the Republic of California flag. BTW - Texas was also a republic.

Massachusetts calls itself the Commonwealth of Massachusetts

We are states of the United States but not all states call themselves states in many situations because of history and tradition. For example, the Commonwealth of Massachusetts had a history and identity proceeding its joining the the Union, there was no reason to stop calling itself the Commonwealth of Massachusetts.

~

[Bruce B]

The following may seem a foolish question or a joke but it is not. Since you live in "The Peoples Republic of California" and I live on "the other side of the world" I want to be sure I'm not falling intocultural misunderstandings.

I think the best thing to do or to have done is let the helpers help you by using symptoms which are; a problem with booting.

Are you suggesting I should post or should have posted in this forum the symptoms that are related with the problem I have with booting my computer in order to make things easier to those who want to help?

Antipodal,

You seem to understand what I was saying. If the problem was a power supply we should have some fairly concrete answers by now I think, but it seems we are in a mystery. And it still might be a bad power supply.

Presently, I think, I understand that the computer won't boot, didn't boot before your post and you want to know how to check the power supply, because you suspect it as the culprit.

A problem I have now

1) my computer won't boot when its cold, I have to leave it running for several minutes before it will boot.

2) I also locks up tight, especially when the ambient temperature is low. I wrap it in a blanket and leave only an intake and outtake hole open.

3) I locks up more when playing multi-media files.

4) I've done a lot of work on it and have completely isolated this to a hardware problem and having nothing to do with software.

My guess is if I communicate all the symptoms as well as all the things of done it would help the helpers help me.

What would you say and ask in my shoes?

And the symptoms I listed could be a power supply problem. But I suspect a oddity in the hard disk.

Hope this makes sense.

PS Please don't try and help me with this, because I'll make a new topic for it, rather than tack it on to this topic.

~~~~~~~~~

[mini-jaguar]

What about Laptops? The Battery died on one of my laptops
and it refuse to start despite me have ac/dc power connected.
It is an ASUS L3800
Other laptops works with just the power supply needing not battery.
Can some laptops have something similar that the mother board
never start up if the battery is not high enough to give a signal.
Maybe I need to connect a kind of fake battery that gives that signal?

Yes, you are correct that certain laptops need the battery to be a little charged.

With my old Toshiba from 1998 I could take out the battery and it would still run, but on an Apple the battery needs to have a minimal charge otherwise it won't run.

I guess it might be possible to set up a fake battery. On the laptops where they don't power up without a battery, the power supply is just a charger for the battery, so most likely it would work by connecting the power supply directly to the battery connectors.

It might also not work as it might need a slightly different voltage or other characteristics, and you might burn out something in the computer, but if you have nothing to lose, then why not?

And yes, most laptop power supplies can be tested by multimeter and will give a reading.