So my motherboard died over the night.Pc entered hibernation and well it didnt wake up :( I`m sure its the motherboard as I tested it with 2 other psus and it didnt start, no post code no errors, fans didnt move etc.Moved cpu to other motherboard and it works . But what could be the reason ?Its asrock server board socket 2011-3 used with xeon so I didnt OC .The main question is, should I trust the psu (seasonic platinum) as I dont know the cause of failure.Faulty psu ? power surge?I could test voltages of psu but that wont say anything and now I am afraid to use it with other stuff.Any advice is welcome.
Testing the voltage will say something if it is over the limit, then you've got your possible culprit right there. If the voltages are fine, however, according to This guy it's not enough. He uses an oscilloscope to test for noise, i.e voltage ripples (I'm not really sure of the correct terminology here). As for which oscilloscope to use I'm not sure. I've heard you can get some cheap ones and hook them up to your computer, or, you can get some really pricy ones that is a stand alone device. Or, find someone who already own one and know how to operate it.
Possibly the noise could get worse when it's under load, But regarding the voltages you can measure with a multi-meter or with HWmonitor and such, they will normally only drop when you start a game as the graphics card will draw more amps all of a suddden. The voltages should go back to normal within a second or so. This will not damage your equipment, only maybe shut the device down if the low spike is too large. But you can of course test for noise under load too.
Edit: I too have a psu I'd like to test, will follow this thread. Got some knowledgeable folks to ask, this thread poked me to go ahead and ask them soon.
To truely test a PSU you need a lot of equipment; you have to check the base voltage, than the ripple (one can do that with a multimeter though; switch it to DC to check base voltage; switch it to AC to check the ripple) if the ripple is strong enough to picked up by a normal multimeters mV range... its likely to high.
But now comes the trouble.. that is without load, you need to check that with various loads on all the voltages, the same time, than you also have to mess with the input voltage.. if a spike comes through the PSU, or a surge, or ... or ... or... its hard.. could write a whole essay.. but without the equipmet needed... its pointless
1 Like
Psu tester is to costly so I`ll assume that psu not safe to use ;/ Time to order new stuff ,and check if the rest of the stuff is alive ;/ argh UPS I ordered didnt get here on time, it seems.
One can not say if it was caused by external issue; but if everything else in your house is fine, its fairly save to assume it was nothing that came from the line.
A PSU tester isn't much, $20, however it won't tell you anything you can't know from software.
1 Like
I mean a full ,standalone psu testing unit where one can create artificial load and measure a ripple :) I have a psu voltage tester, I`m making some custom sleeved cables so its a must.
1 Like
I was shure that many people, much more knowledgable than I am, did that before.
I just wrote that as excuse to stop writing XD
Those test setups are incredible expensive
1 Like
I want to do this. Already talked with a friend who's got a couple oscilloscopes. Now, only thing missing is how to test under load. Theoretically we can hook it up to a computer, though that is not very safe for the equipment or the fire hazard.
Does your friend have high wat resistors? like those huge cheramic packages?
P (W) = U(V) * I(A)
So if you want to test e.g. the 12V rail.. and assume 120W you need a resistor that can handle that in thermals O.o so many many of 5W resistors could do that.
I don't know, but I think he can pick some up at work or know someone who's got some. That way we can test maybe 2-3 loads. It's certainly better than just idle.
yeah. Its mostly important to have at least some load... as a well designed PSU will turn down the switching if it does not have to deliver power. even some designs feature differen transistor stages for different power levels...
Usually on todays PSUs the +12V and -12V rail is generated and the other voltages are derived from that
Okay, thanks. I'm thinking cant the psu create ripples on the 5v and 3,3v separately from the 12v, even though they are derived from it?
Depends on the actual design.. there are seperate switching transistors, filter caps and chokes.. and so on...
1 Like