Load Tests (Cont’d )

As usual we pushed this power supply over its official limits to see what happens.

First we tried to see if over current protection was active and at what level. To test this we installed only the cables from the +12V1 virtual rail to the +12V1 input from our load tester. This included the main motherboard power cable, the two peripheral power cables and one of the ATX12V cables. Then we started increasing current at +12V1 until the power supply would shut down. This happened when we pulled more than 20 A, so over current protection (OCP) was active and set to shut down the power supply if we pulled more than 20 A from any +12V rail. This is great, because according to the power supply label each +12 V has a limit of 18 A, so OCP was configured really close to what was printed on the label. Several power supplies on the market have the OCP circuit configured with a value that is so high that it probably will never enter in action, so the power supply isn’t really protected.

Our next move was to discover what was the maximum amount of power this unit can deliver still working inside its specs.

Starting from pattern number six (see previous page) we increased current on both +12 V inputs from our load tester to 19 A. Above that the power supply would shut down.

We know that this power supply does not feature over load protection (OPP or OLP; both acronyms mean the same thing) because the monitor chip used on this unit doesn’t have this protection.

So basically overload protection for this power supply is being made a well adjusted OCP circuit. We knew that we would burn the +5 V or +3.3 V rectifiers if we tried to overload them. This is an unrealistic scenario, as on an overloaded computer the overloading will occur on the +12 V lines, due to power-hungry video cards and CPUs.

We were conservative on our overload tests this time, since we had already figured out that we could pull a maximum of 38 A (456 W) from the +12 V outputs. So we increased current on +5 V and +3.3 V to 15 A each. We think that this already represents a good overloading. Under this scenario we had the results presented on the table below.

At this scenario noise and ripple were at the same noise level presented on previous page.

We could pull even more current from +5 V and +3.3 V outputs (20 A each, for a total 641.4 W and pulling 823 W from the wall, 78% efficiency) but at this scenario the power supply silently died. After opening the power supply we tested all the main semiconductors and none of them were burned and we couldn't find out which component burned.

Short circuit protection (SCP) worked fine for both +5 V and +12 V lines.

The fan used on this power supply is really quiet, even when the power supply was hot and delivering 500 W.

Total or partial reproduction of the contents of this site, as well as that of the texts available for downloading, be this in the electronic media, in print, or any other form of distribution, is expressly forbidden. Those who do not comply with these copyright laws will be indicted and punished according to the International Copyrights Law.

We do not take responsibility for material damage of any kind caused by the use of information contained in Hardware Secrets.