EVGA 650GM 650W SFX Power Supply

REVIEW INFORMATION
SUPPLIED BY: EVGA
PRODUCT: EVGA 650GM 650W SFX
PROD LINK: 650GM Product Page
PRICE: $129.99 @ EVGA.com
Price is at time of testing!


EVGA 650GM – Overshoot Transient Tests
VSB On VSB to Full, 12V Off to Full, 12V

Here’s another mere observation… these are some awesome looking scope shots right here. I see no power on spikes of any sort, just the smooth and continuous waveforms we want to see when it comes to figuring out the S&C scoring point for later.

Shall we heat things up now? No, no, put down the bong. You’ve had enough, sir. I meant it’s time to hot test this power supply now. I should be able to get this unit pretty close to forty degrees today.

EVGA 650GM – Hot Load Tests
Test # +3.3V +5V +12V DC Watts/
AC Watts
AC
Input
Eff. P.F. Intake/
Exhaust
Progressive Load Tests
1 1A 1A 4.6A 65.1W/
74.7W
118.8V 87.2% 0.950 26°C/
33°C*
3.294V 5.07V 12.06V
2 1.5A 1.5A 9.5A 129.5W/
142.8W
118.7V 90.7% 0.973 29°C/
44°C*
3.283V 5.06V 12.05V
3 3.5A 3.5A 24A 324.0W/
352.1W
117.6V 92.0% 0.993 33°C/
64°C*
3.258V 5.03V 12.03V
4 5A 5A 36A 483.5W/
535.3W
116.9V 90.3% 0.997 37°C/
72°C
3.236V 5.01V 12.01V
5 7A 7A 48.5A 650.2W/
741.0W
117.2V 87.7% 0.998 39°C/
79 °C
3.209V 4.98V 11.97V
Crossload Tests
CL1 12A 12A 0A 100.7W/
119.7W
119.1V 84.1% 0.970 28°C/
49°C*
3.255V 5.03V 12.06V
CL2 0A 0A 54.1A 648.3W/
735.4W
118.0V 88.2% 0.998 38°C/
80°C
3.237V 5.01V 11.96V

* Fanless operation.

This time, we did have the fan shut itself off in the first crossload test, with the rest of the tests showing about as much fan action as the cold tests. The high exhaust temperatures are due to me placing the probe right in front of the bridge rectifier area of the exhaust, so that’s nothing to worry about.

Sadly, this unit did not hit Platinum efficiency this time… merely the Gold performance we were promised in the first place. But that’s ok. We can’t always get extra performance like that, can we?

The unit must have heard me complaining about the 3.3V regulation, because it made an effort to improve things to 2.58% this time out. The other two rails are exactly the same as the cold tests, on the other hand. So, for the scoring page, we now have a 1.54% result. That’s almost but not quite good enough to get it into the excellent scoring bracket.

Time to look at some scope shots.

EVGA 650GM – Oscilloscope
Test # +3.3V +5V +12V
1
2
3
4
5
CL1
CL2

Oh. That’s unfortunate… looks like we have another one of those units that throws random spikes everywhere. Otherwise, these shots would look great.

As is, we have about 22.5mV on the 3.3V rail, 50mV on the 5V rail, and 50mV on the 12V rail. Only the 3.3V output can be considered excellent, while the other two are only “very good.” This wouldn’t have been too shabby a result for the form factor not too long ago, but now we have units like the Corsair SF600 to deal with, which gives you better regulation, better ripple control, and also can clear Platinum if you’re lucky.

Time to do some disassembly.