BFG EX 1200 Power Supply review | test - Preliminary test

The Test

Unless you have seriously professional equipment worth thousands of EURO's, it's very hard to test, and most of all stress a PSU. Firstly we do a quick test with this nifty new gadget I have been testing. With a power supply tester I'll quickly check if the rails are all working properly and what amount of voltage they are delivering without a PC connected to it.

This is just a test to see if the PSU is functional and working properly before we connect it to equipment worth a stack load of money.

Okay, so we can instantaneously see that all primary voltage rails are working properly. Voltages seem to be precisely what they need to be, and fall well within ATX specifications. So this is looking good. This however is not an objective enough tool to measure voltage at all.

For more precise measurements we'll test with a multi-meter. For those that are wondering what the PG delay being displayed on the screen is, here we go: Power Good (PG) Delay and Off Delay: It takes a power supply at least 100ms to completely power up before it can offer the proper DC voltages to the computer. The computer will not attempt to turn on until the power supply is finished powering up and sends a Power Good (PG) signal to the motherboard via a PG terminal. On power up the Power Good (PG) terminal is polled for a brief period before the PC turns on to check the status of the power supply. Only after the +5V and +3.3V rails are above threshold does the PG signal go "high." The PG signal stays on until the PC is ready to power off. The period of delay between switching on the PS and the actual application of power is the PG delay. Typical values are 100-350ms. The PG delay of the power supply should roughly match the PG delay specification of the motherboard or power on problems may occur.

So this is within specifications. Let's do a real-world test.

Load testing the PSU

As you can see, we had a bit of a challenge at hand. I mean without professional Chroma load testers it's pretty hard to stress a 1200 Watt power supply. Now then, that demonic plan I mentioned earlier ?:

So here's what we did. I decided to take a power hungry nForce 790i SLI motherboard. Then took 2x GTX 295 (four GPUs) for maximum power consumption. Then I added another GeForce GTX 280 purely for PhysX. Then we took a power hungry QX9770 processor and overclocked it towards 3600 MHz. Then we overclocked our memory as well.

These are the components used:

• nForce 790i motherboard (has high power consumption
• Core 2 Quad QX 9770 Processor (overclocked to 3.6 GHz)
• 2x GeForce GTX 295 primary (4 GPUs)
• 1x GeForce GTX 195 secondary
• GeForce GTX 280 card for Physx
• 3 GB Memory DDR3 @ 1600 MHz
• Optical Drive x1
• HDD x1

As such we'd figured / hoped to reach 800-850 Watt power consumption. But with power supplied getting more and more efficient these days .. that proved to be a hard task to accomplish. We gave it our best though. Let's have a look at the test results.

Organized chaos people ... organized chaos.

So here's where we start a selection of gaming tests well balanced for optimal CPU (multi-core) and GPU performance. We take a multi-meter and tap critical points during system under load and when in idle. We then observe the voltage fluctuation. A lot of de/increases would indicate instability.

Example: the graphics cards are consuming the most power. Therefore we need to focus on using software that can utilize that to the maximum.