One of the problems that is currently least seen in the world of custom PCs is that of home repairs and tests, since as a rule the quality level of the components is very high.
Of the components that most often fail we have motherboards , graphics cards and power supplies , where we will focus on the latter to learn how we can know and test its operation.
Power supplies are an extremely complex component
And they are from the electronic point of view, since their well-known function includes regulating the voltage in the most precise way possible, wasting the least energy in the form of heat that it allows and, above all, with the greatest stability in voltages and amps. that I can achieve.
Although as we have said before, they have improved the average quality a lot, reaching in the cases of extreme sources to last for decades in many cases (even covered by warranty)
they are not components that have a very low failure rate, since each time they are demands more from the rest of the components.
As you already know, we always recommend investing a significant amount of money from each budget in this component, which in most cases will avoid problems in the rest of the components.
But if a source fails, it is best to do a brief manual or automatic seft-test and thus be able to detect where the problem is coming from.
For this we will have to get a good multimeter or even in the most delicate cases with a complete station. In most cases, a medium multimeter will be
more than enough, since by detecting a minimum variation in voltage we will be able to satisfy the test.
Tester vs Polymeter
There are several ways to test the stability of the power supply. The most comfortable is undoubtedly an exclusive tester for sources,
where it acts as a multimeter, except that on many occasions it is not as precise and updates the values less frequently than its “rival” achieves.
Another problem with this type of tester is that it does not allow you to test 8-pin cables for graphics cards, that is, the current 8-pin PCIe, although it is possible with the 6-pin cable.
Remember that the 12-volt 8-pin EPS is not the same as the PCIe 8-pin. For this reason, the multimeter option is more complete,
so we are going to know the simple steps that we need to be able to measure the voltage.
Source outside the PC
In the first place, we must have the source outside the PC , for simple safety and to rule out energy derivations towards the tower or possible short circuits with any component.
Obviously we will have to place it on a surface that does not transmit current, not even static electricity. Once this is done and with the rear switch of the same in circulation and not in flow (“off” so that we understand each other), we proceed to bridge pins 15 and 16 such that:
With any cable or clip we can achieve it, as long as it transmits the current clearly. They also sell an anchor that comes prefabricated, so it’s just a matter of arriving and plugging it in.
Once bridged in either of the two ways, we will proceed to activate the source by means of its rear switch, so that the current flows towards the cables .
As we surely know, all PC power supplies play with three totally different voltages that vary depending on the cable we have in our hands: +12V, +5V and + 3.3V.
These voltages are only going to be present at one time on the 24-pin cable (20 if the source is very old), so this is the cable that will have the most work. To know what each pin corresponds to, we will place the complete diagram.
Starting to measure
Before starting to measure, we will set our multimeter in direct current (continuous line and points) in a measurement setting of approximately 20 ( tens ) so that the measurement is correct.
The black connector is placed in any of the ground connections , while the red one is the one that we will have to move between pins to check voltages. Normally everything will be fine as long as no value exceeds ± 5% on load (positive voltages).
The negative voltages can exceed ± 10% , but in both cases the ideal is that they not only have a final voltage closer to the original numbers, but also that they do not vary at all.
If all the values of all the cables are good and do not fluctuate, the next step is to check them with the source already mounted, measuring from the top of the connectors and in real use,
where we will have to do another follow-up to see if the problem it’s not so much the source itself but another component.