I use a SunMoon SM-8800 load tester to put artificial loads on power supplies.
Different cables plug into different connectors on the front of the load tester
and each connector is assigned a load.
With a single +12V rail power supply,
I usually don't pay too much mind to what plugs in where. I typically just
plug in as many connectors as possible into the front of my load tester and
fire it up. But for
this
power
supply, I couldn't do a full load and was wondering if the power supply was
really a 1250W or if it was tripping off because of the high temperatures.
I even ran my tests at room temperature and the power supply still shut down.
XFX ensured me that the power supply should do 1250W without issue and sent
me a second unit. I plugged the second unit into the load tester and experienced
the same problem. At high loads, the power supply kept shutting down.
So what's the deal? Well, it turns out that EasyRail isn't EasyRail after
all. This power supply actually has four +12V rails. The way I hooked up the
cables
to
the
power
supply, and the way the +12V is distributed was causing the OCP to trip. I
was plugging the 4+4-pin into my load tester and the 4+4-pin shares a +12V
rail with the 24-pin connector; so putting a load on both of these connectors
was putting the total load on one +12V rail.
After some emails and trial and error (like, a whole day of trial and error),
it turns out the DC output table should look something like this:
XFX
Pro 1250W
+3.3V
+5V
+12V1
+12V2
+12V3
+12V4
12V
+5VSB
25A
25A
30A
30A
45A
45A
0.5A
3A
Max Power
150W
1248W
6W
15W
1250W
Now, will this be a problem for someone installing this power supply in a
real computer? Well, the most you'd probably ever load up the CPU power connector
is around 20A. The most you're going to send through the 24-pin connector's
+12V wires is around 10A. The likelihood that you're going to send that much
power through both of the connectors all at once is next to nil.
Below is our cable listing again, this time showing what +12V rail goes to
what connector:
5.25" Peripheral Power Connector (1 cables w/ 2 connectors) (360mm+170mm)
2
+12V4
3.5" Drive power adapter (Splitter that adds two +170mm cables
to the end of 5.25" Peripheral Power Connector)
2
+12V4
Unit Dimensions(L x W x H)
190mm x 86mm x 150mm
Now that we got that all figured out, I unplugged the 4+4-pin power connector
and used the 8-pin CPU power connector and plugged it into the +12V4 connector
that can be used for either an 8-pin CPU power connector or PCIe cable.
The +12V load on my load tester is split up across six different connectors. Each connector gives me a different voltage, though. Because the differences are largely because of resistance in the connectors and/or wires, I log the highest voltage reading in my table. That said, if you "do the math", you might find total wattage numbers lower than "they should be" because ALL voltage readings are taken into account when total wattage is calculated.
Finally, these tests are performed in a 19°C room, but the hot air from the load tester is pumped into the intake of the power supply in an effort to emulate actual operating temperatures, so as the load increases the temperature of the power supply increases just as it would in your PC.
XFX Pro 1250W HOT load tests
Test #
+3.3V
+5V
+12V
+5VSB
DC Watts/ AC Watts
Eff.
Power Factor
Intake/ Exhaust
Simulated system load tests
Test 0
0A
0A
0A
1.5A
7.6W/ 9.8W
77.6%
.54
19°C/ 19°C
5.07V
Test 1
1.8A
1.8A
8.5A
1.5A
124.4W/ 150.1W
82.9%
.98
23°C/ 25°C
3.33V
5.04V
12.02V
5.04V
Test 2
3.6A
3.6A
17.7A
1.5A
249.8W/ 282.1W
88.6%
.98
27°C/ 30°C
3.33V
5.04V
12.03V
5.05V
Test 3
5.4A
5.4A
26.9A
1.5A
375.2W/ 417.8W
89.8%
.99
30°C/ 34°C
3.33V
5.04V
12.03V
5.04V
Test 4
7.2A
7.2A
36.0A
1.5A
498.7W/ 553.3W
90.1%
.99
34°C/ 40°C
3.32V
5.04V
12.03V
5.02V
Test 5
9.0A
9.0A
45.2A
1.5A
623.5W/ 689W
90.5%
.99
36°C/ 42°C
3.32V
5.01V
12.03V
5.00V
Test 6
10.8A
10.8A
54.4A
1.5A
747.9W/ 836W
89.5%
.99
41°C/ 45°C
3.32V
5.03V
12.03V
4.99V
Test 7
14.5A
14.5A
72.7A
1.5A
995.5W/ 1132W
87.9%
.99
44°C/ 50°C
3.31V
5.03V
12.03V
4.92V
Test 8
18.1A
18.1A
91A
1.5A
1238.5W/ 1472W
84.1%
.99
46°C/ 58°C
3.29V
5.01V
12.02V
4.88V
Test CL1
0A
0A
104A
0A
1239W/ 1445W
85.7%
.99
47°C/ 59°C
3.31V
5.03V
12.07V
4.95V
Test CL2
18.1A
18.1A
2A
0A
175.1W/ 216.1W
81%
.99
42°C/ 45°C
3.32V
5.04V
12.01V
5.09V
Low and behold, the power supply stopped shutting down, even when exhaust
temperatures hit 59°C.
The first observation I take away from these results is how well the voltage
sense circuitry works on this unit. +3.3V only dropped .04V. +5V dropped a
mere .03V and the +12V actually went up .01V before coming back down to +12.02
where we started.
Voltages also stayed well within specification when crossloaded.
Efficiency looks very good on this unit. Test 2, 5 and 8 are 20%, 50% and
100% loads respectively. 88.6%, 90.5% and 84.1% efficiencies at these loads
means this unit comes very close to 80 Plus Silver status. There's an 80 Plus
Gold logo on the box, but with full load efficiency of 84.1% this unit is pretty
far off from the 87% efficiency required to get an 80 Plus Gold certification.
I would re-test the unit using the 80 Plus Gold loads from the test report
on their website, but there
is no 1250W power supply on the list of XFX power supplies certified by 80
Plus. Of course, we already know that this power supply is based off the
Seasonic
X-Series 1250W, so we can use the test settings for that unit to see
if we can hit 80 Plus Gold.
XFX
Pro 1250W 80 Plus load tests
Test #
+3.3V
+5V
-12V
+12V
+5VSB
DC Watts/ AC Watts
Eff.
Simulated system load tests
20%
3.2A
3.2A
0.1A
18.4A
0.5A
251.1W/ 282.3W
88.9%
3.33V
5.04V
-11.73V
12.03V
5.08V
50%
7.9A
7.9A
0.2A
45.9A
1.3A
624.3W/ 692W
90.2%
3.32V
5.04V
-11.88V
12.02V
5.01V
100%
15.8A
15.8A
0.4A
91.7A
2.6A
1241.7W/ 1449W
85.7%
3.31V
5.03V
-12.24V
12.02V
4.90V
Using the loads the 80 Plus folks used to test the Seasonic X-Series 1250W,
we did a little better at 100% load, but we're still not at 80 Plus Gold efficiency.
Now let's have a look at the ripple:
Oscilloscope Measurements for XFX Pro
1250W
Test #
+3.3V
+5V
+12V
+5VSB
Test 1
Test 2
Test 3
Test 4
Test 5
Test 6
Test 7
Test 8
Test 9 (CL 1)
Test 10 (CL2)
The ripple is a little wavy on the +12V rail, and downright choppy at full
load, but it's well within specification.
Now we're going to look for overshoot transients and rise time on the +5VSB
by going from full off to full load on the +5VSB, and on the +12V by going
from standby to full load on the +12V and full off to full load on the +12V.
Overshoot
Transient Testing for XFX
Pro 1250W
Off to VSB On
VSB to Full, 12V
Off to Full, 12V
So we can see here that rise time is about 10ms on both the +5VSB and +12V
and there's virtually no overshoot whatsoever.
Now let's bust out the screwdriver and take this power supply apart.
JonnyGURU Feeds
Add our RSS feeds to your favorite RSS Reader or homepage.