The Kolink KL-SFX250 psu is a bare bone psu, it does not include a power cable.
The psu is pack with a bubble wrap inside the retail cartoon packaging and it only comes with 4 screws.
80mm fan, 2 molex for hdd/optical , 2 sata, 20+4pin for mobo, only 4pin
for cpu and no PCIe cable.
18.5A in the 12V rail and a decent 3A in the +5V Standby rail.
80 PLUS Bronze with Active Power Factor Correction.
Over voltage protection, Under Voltage, Short Circuit and Over Power
protection.
Were going to try to pull 249.9W from the unit in our short test.
Below is my little contraption of 5 DC Electronic Load board.
The 2 units on the left with the black heatsinks is rated at 110 Watts each but you can only pull about 109
Watts from the load ( the board protection limits it to < 110W ) , the remaining 3 is rated at 60 watts but
you can only pull about 59 Watts from the load.
The 2 black units is connected to the 12V rail.
The other 3 is connected to 5VSB, 5V rail and 3.3V rail.
1st unit black one is connected to the 12 Volts rail - Pulling 8.75Amps for 104.5 Watts
2nd unit black one is connected to the 12 Volts rail - Pulling 8.75Amps for 104.5 Watts
The 2 black Electronic Load board is pulling a total of 209 Watts from the 12 Volts rail.
3rd unit small one is connected to the 5 Volts Standby Rail pulling 3Amps for 15 Watts
4th unit small one is connected to the 5 Volts Rail pulling 3Amps for 15.4 Watts
5th unit small one is connected to the 3.3 Volts Rail pulling 3Amps for 9.9 Watts
The 3 smaller units is pulling a total of 40.3 watts.
All 5 of them together is pulling a total of about 249.3 watts from the psu.
Below is the Unloaded voltage reading from the 3 rails 12V, 5V and 3.3V
The 12V and 5V is tap at the molex collector which is not connected to any load.
The 3.3V is tap at the 24pin unused 3.3V connector.
You can see below that the 12V rail drop to about 11.7x Volts when loaded with 17.5A ( 209W ).
The psu uses AWG 20 for its cabling which causes the 0.2V drop, as this small voltage drop is lost in the
smaller AWG 20 cable, a bigger cable say AWG 16 would minimize this loss.
The good news is, all of the rails is still within the 5% voltage specs
Table below from http://www.formfactors.org
Below is the voltage reading from the 12V, 5V, 3.3V rails when loaded.
The 12V and 5V is tap at the molex collector which is not connected to any load which reflect the voltage at
the source inside the psu.
The voltage regulation of the 12V rail is actually very good ( within 1% ) even loaded with 17.5Amps, just
1Amp shy of its max 18.5Amps rating but is let down by the smaller cable use by the unit so there is about
0.2V voltage loss in the cable.
Below you can see the the Electonic load pulling 3A from the 5V standby rail and the 5V rail.
Below is the 3.3V rail loaded with 3Amps
Total power draw from the wall according to my cheap Energenie power meter.
I run the load for about 30 min with about 26°C ambient temp.
From this reading we can calculate the efficiency of the psu with the formula, efficiency % = Pout/Pin x 100.
249.3/289.2 x 100 = 86%
All of the stuff that was used here are all consumer grade and will vary a lot in specs and or reading/measurements especially when compared with industrial/lab grade intruments.
No psu teardown this time as this is just a quick check if the psu can deliver its rated power, which it did.
The psu is pack with a bubble wrap inside the retail cartoon packaging and it only comes with 4 screws.
80mm fan, 2 molex for hdd/optical , 2 sata, 20+4pin for mobo, only 4pin
for cpu and no PCIe cable.
18.5A in the 12V rail and a decent 3A in the +5V Standby rail.
80 PLUS Bronze with Active Power Factor Correction.
Over voltage protection, Under Voltage, Short Circuit and Over Power
protection.
Were going to try to pull 249.9W from the unit in our short test.
Below is my little contraption of 5 DC Electronic Load board.
The 2 units on the left with the black heatsinks is rated at 110 Watts each but you can only pull about 109
Watts from the load ( the board protection limits it to < 110W ) , the remaining 3 is rated at 60 watts but
you can only pull about 59 Watts from the load.
The 2 black units is connected to the 12V rail.
The other 3 is connected to 5VSB, 5V rail and 3.3V rail.
1st unit black one is connected to the 12 Volts rail - Pulling 8.75Amps for 104.5 Watts
2nd unit black one is connected to the 12 Volts rail - Pulling 8.75Amps for 104.5 Watts
The 2 black Electronic Load board is pulling a total of 209 Watts from the 12 Volts rail.
3rd unit small one is connected to the 5 Volts Standby Rail pulling 3Amps for 15 Watts
4th unit small one is connected to the 5 Volts Rail pulling 3Amps for 15.4 Watts
5th unit small one is connected to the 3.3 Volts Rail pulling 3Amps for 9.9 Watts
The 3 smaller units is pulling a total of 40.3 watts.
All 5 of them together is pulling a total of about 249.3 watts from the psu.
Below is the Unloaded voltage reading from the 3 rails 12V, 5V and 3.3V
The 12V and 5V is tap at the molex collector which is not connected to any load.
The 3.3V is tap at the 24pin unused 3.3V connector.
You can see below that the 12V rail drop to about 11.7x Volts when loaded with 17.5A ( 209W ).
The psu uses AWG 20 for its cabling which causes the 0.2V drop, as this small voltage drop is lost in the
smaller AWG 20 cable, a bigger cable say AWG 16 would minimize this loss.
The good news is, all of the rails is still within the 5% voltage specs
Table below from http://www.formfactors.org
Below is the voltage reading from the 12V, 5V, 3.3V rails when loaded.
The 12V and 5V is tap at the molex collector which is not connected to any load which reflect the voltage at
the source inside the psu.
The voltage regulation of the 12V rail is actually very good ( within 1% ) even loaded with 17.5Amps, just
1Amp shy of its max 18.5Amps rating but is let down by the smaller cable use by the unit so there is about
0.2V voltage loss in the cable.
Below you can see the the Electonic load pulling 3A from the 5V standby rail and the 5V rail.
Below is the 3.3V rail loaded with 3Amps
Total power draw from the wall according to my cheap Energenie power meter.
I run the load for about 30 min with about 26°C ambient temp.
From this reading we can calculate the efficiency of the psu with the formula, efficiency % = Pout/Pin x 100.
249.3/289.2 x 100 = 86%
All of the stuff that was used here are all consumer grade and will vary a lot in specs and or reading/measurements especially when compared with industrial/lab grade intruments.
No psu teardown this time as this is just a quick check if the psu can deliver its rated power, which it did.
Last edited:
