Voltage 3.7V from the computer's power supply - a cause for alarm?

Question

Do it yourself

I measured the voltage from the Patriot AA400-K power supply (ATX form factor), it turned out +3.7, instead of +3.3. Is this value critical for the equipment? How can I downgrade if critical?
Tester readings in the photo: postimg.org/image/yseher7m3/full
Measurement result with a multimeter without load (with the legs closed to start the PSU): https://s22.postimg.org/5jqn8ju1t/atx.png

Answer 1

Voltage 3.7V from the computer's power supply - a cause for alarm?

I join Artem - How was it measured? Under load, no load, on a running computer?
Again - is this most likely a Chinese tester lying?
Under load, the correct tester should not show so much, you have problems with the PSU

Answer 2

I will answer as a practitioner. The tolerance rate for Uout values ​​under load is ± 5%, i.e. for 5 volts it is ± 0.25, for 3.3 - ± 0.17 V. In practice, computer electronics normally tolerate even a tolerance of ± 10% if the voltage is "smooth", not pulsating (that is, if the smoothing capacitors have not dried up ). Further, all output voltages are connected to each other, the correct ratios are maintained between them. Those. if, for some reason, Uout of one channel increases, then, as a rule , the voltages of the remaining channels also increase proportionally. However, this proportionality is maintained only at nominal loads of these channels. If one of the channels is not loaded, then its Uout may well go beyond the standard limits.
In this case, the output of the 3-volt channel clearly went beyond even a 10 percent tolerance. The first step is to make sure that the current of this channel is at least 1 ampere (and preferably several). You can connect a light bulb from a car headlight to it and see how much Uout will sink and whether it will return to normal. And only then decide whether to throw out this PSU or repair it further. And of course (as other respondents correctly noted here), you don’t have to believe a single measuring device, you definitely need to duplicate the measurements with something else, better more precisely.