When we first opened up this package we were
quite happy to see that Plain Label PC had done a great
job of stopping the PSU box from moving around during shipping.
The actual PSU box isn't very pretty, it's just a regular
brown box. Well, what can you expect when it is generic?
We also thought the power supply unit itself was rather
dull, it looks just like any other: A plain grey case. But
hey, looks aren't everything right?
So we grabbed it out of the
box, and unraveled the cords. They appeared to be a little
on the short side. Now, this is one of two things. If you've
got a large server style case, it's bad. You may have trouble
reaching anything if its more than 18 inches away from the
psu (scroll down a little for all the lengths). However,
those of you with mid size towers may actually prefer this.
Why? Well, shorter cables avoid excess slack, and that excess
slack is not only unsightly but it also can restrict air-flow
if in the way.
Next, it was time to test
the voltages. We rigged the psu so it would be able to start-up
without being plugged into a motherboard. In order to obtain
accurate results (the bios can be wrong) we used our handy-dandy
Radioshack voltage meter. Once we powered the PSU on we
weren't even sure it was working. Let me tell you, this
thing is absolutely silent. I found myself plugging some
LED fans in just to make sure it was actually on. Anyways,
we measured the 3.3, 5, and 12 volt lines. We got good readings:
3.326, 5.19, and 12.12 respectively. The 3.3V line was pretty
much dead on, the 5.0V was only off by about 4%, and the
12V was off by 1%. After taking these readings, we were
quite impressed. However, it's not about a PSU giving the
right voltage without load. It's all about how close it
can stay to what it should under pressure.
Our results in a graph from the test described
above.
Installation was no different
than any other PSU: Get annoyed by stubborn molex plugs,
unscrew and pull out the old box, put the new box in, screw
a few screws, and plug in all the cables. We had no problem
with the shorter than normal cables, in fact they seemed
to fit the case quite nicely. Normally I have to try and
squish tons of extra slack here and there, but with this
PSU the cables did not have all that extra slack and therefore
didn't require as much hiding.
Finally, it was time to boot
up the system. Hearing stories of no name PSUs burning,
dying, or frying your system, we had to cross our fingers.
I guess we under estimated Codegen, everything was fine
once the power button was hit. Well, until we went into
the bios. We were greeted by great readings on the 12V (near
perfect) and also on the 5V (within 2% of perfect), but
the 3.3V was hovering at 3.12V. This reading meant that
neither the GeForce 4, nor the ram, nor the PCI cards were
getting enough power.
The recorded voltages while the computer was idle.
We booted into windows anyways,
and kept on doing normal things. Surprisingly we had no
stability problems for several days, even with overclocking
the card and system ram. I was very happy with this PSUs'
performance so far. We passed the low 3.3 reading off as
a hardware monitoring error. Well, later that same night
we did have stability problems. During a crucial moment
in UT2K3, the system rebooted out of nowhere (I know what
you're thinking, we said that in the intro, didn't we? Funny
thing is, this happened after we wrote that). Once we were
back into windows, we checked our 3.3V log in Asus Probe.
The last reading it got was 3.05, and obviously either the
ram or the GF4 just couldn't handle that low of a voltage.
However, we were quite happy to see that the 5V line had
not changed under pressure, and the 12V line jitter was
within 2.5%.
We applied a full load to the computer. These are the lowest
voltages we received..
Now we knew it wasn't just
a hardware monitoring problem like we had previously thought,
the PSU we had received seemed defective. So I contacted
the guy I talk to from Plain Label PC. Having used the same
model PSU himself, he was very surprised that I was having
any sort of problem. After laying out all the details, we
determined that I had received a defective unit that could
not keep the 3.3V line up (I even tried using a Geforce
2 to see if it would come up a little, but nope).
He apoligized to me, and was
more than happy to replace it. He also assured me that his
own had no problems with the 3.3 line (or any for that matter),
nor did any of the ones that they had in stock. And if somehow
they did, anyone who purchased one would be covered under
their 30-day warranty that allows a refund of any defective
products.
Cable
Length:
Strand 1 molex: 12 and 17 inches
Strand 2 molex: 12 and 17 inches
Motherboard and Pentium 4 plug: 11.5 inches
Floppy: 23 inches
6-pin Aux Power Connector: 12"
Test
Rig
Processor: Pentium 4 2.0A
Memory: 2x 256MB PC2100 DIMM
Motherboard: Asus P4S33 (bios rev. 1.008.001)
Motherboard chip set: SiS 645
NVIDIA reference drivers: 40.71
Hard-Drive: Western Digital 81GB 7200RPM ATA100
IDE Interface: Ultra ATA100
Sound: Creative Labs Sound Blaster Live!
CD/DVD: Elements 56x CD
Secondary CD/DVD: Mitsumi 32x Burner
Power Supply: Codegen 350 Watt
Operating System: Windows XP Pro
Two 80mm fans, 2 more tri-led fans
Two Cold Cathodes
