Corsair HX620 versus HX850 on OCCT PSU Test

[Apoptosis]

I've never used the OCCT power supply test before, but just tried it out on my AMD test system... I'm not sure what to think of the results!

The test system:

AMD Phenom II X4 965 BE
XIGMATEK HDT-S1283 CPU Cooler
Gigabyte GA-890GPA-UD3H motherboard w/ BIOS F7B
4GB Kingston HyperX T1 DDR3 1600Mhz modules at 1.6V
ATI Radeon HD 5870 Video Card
Corsair P64 SSD
1x DVD-ROM

First up is the Corsair HX620W Power Supply from over 3 years ago:

hx620_3v.jpg (95.8 KiB) Viewed 1238 times

hx620_5v.jpg (85.35 KiB) Viewed 1238 times

hx620_12v.jpg (83.1 KiB) Viewed 1238 times

Now is a brand new Corsair HX850W Power Supply that was just opened up and taken out of the box this afternoon:

hx850_3v.jpg (94.84 KiB) Viewed 1238 times

hx850_5v.jpg (89.16 KiB) Viewed 1238 times

hx850_12v.jpg (80.9 KiB) Viewed 1238 times

Corsair HX620W Power Consumption:
76W Idle
196W Load w/ POV-RAY

Corsair HX620W Ripples:
3.3V - 59.33%
5V - 54.05%
12V - 8.90%

Corsair HX850W Power Consumption:
77W Idle
189W Load w/ POV-RAY

Corsair HX850W Ripples:
3.3V - 58.85%
5V - 1.05%
12V - 8.42%

How should I interpret these results? A ~60% ripple on the 3.3V rail would appear way too high... Is OCCT reliable for PSU testing?

[spitter]

Diodes are notorious for bad readings, if the power had fallen by as much as those graphs say, it would have been a "click" and a restart.

[Apoptosis]

yeah... That worries me as I see a number of review sites using OCCT for their PSU analysis.

[bubba]

Ran it on my box.

ABD 1000watt
ASUS P7P55D-E
i7-860 @ 2.8ghz
8gb Corsair ram
GTX-260
3x Seagate 1TB

[bubba]

ran it again in full screen mode, oh, for some reason its not picking up the 5v rail??

There is a difference in the 12v rail, but don't see how this would be better then an oscilloscope and a multi-meter

[DJ Tucker]

OCCT is very good for PSU testing as it Stresses everything in your system to max but unfortunatly a can't run it without system click and restarting so oh well!