Kolink Regulator ATX3.0/PCIE5 PSU review (ft. 15 year old Seasonic S12II)

Man of Honour
Joined
22 Jun 2006
Posts
14,207
OCUK were kind enough to send me Kolink's new ATX 3.0/PCIE5 PSU for testing and feedback. This is my review of the unit I received.

Since this isn't a professional review, I thought to make things more interesting, I'd compare it with the test PC's current PSU: Seasonic S12II-500.

Packaging and contents

406510019-675105674689571-6709943235115450714-n.jpg

The box is sturdy cardboard (flips open at the bottom) with a print finish, in shiny matt black. There's a plastic wrapper, which I assume is from the factory.

Some of the features:

- ATX 3.0 and PCIE5 compatible (AWG16 12VHPWR cable supports up to 600 watt).
- Modern LLC and DC/DC design, with a single 12v rail.
- The PSU's efficiency is Gold rated.
- Fanless mode (up to 40% load) and the fan uses a fluid dynamic bearing.
- 100% Japanese capacitors.
- A compact design (140mm depth), so it can fit in compact cases.
- Protections: overpower/overcurrent (OCP, OPP), undervoltage/overvoltage (OVP, UVP), shortcircuit (SCP), overtemperature (OTP) and SIP (surge and inrush).

Once the box is opened, there's some solid padding around the PSU.

407359593-675105831356222-1390225234118473258-n.jpg

I was pleased to see the cables and connectors were all fully intact and there are no dents or scrapes.

The box contents include:

- 1x PSU
- 2x (resealable) bags of cables.
- 1x user manual.
- 4x black PSU screws (for attaching to the case).
- 1x warranty booklet.

406508028-675105914689547-7551588717202006887-n.jpg

The cables:

Bag 1
- 12VHPWR cable (AWG16).
- 2x EPS12v (marked: CPU, 8-pin) cable (1 connector per cable).
- 24-pin motherboard cable.

Bag 2
- 2x SATA cables (3 connectors per cable).
- 1x molex cable (4 molex connectors).
- 3x PCIE 8-pin cables (2 connectors per cable).

406519978-675106174689521-7307283470582637122-n.jpg

My take:

Just 1x Molex cable is not unusual and makes sense, since they aren't used much anymore.
2x SATA cables is enough, with M.2 drives being more common for new PCs. An extra cable would have been handy for larger cases and heavier usage, especially since SATA power is sometimes used for fan controllers and RGB.
3x PCIE and a 12VHPWR cable is handy, since the PSU has enough cables for either a 7900 XTX (no daisy chain) or a 4070 Ti upwards.

Cable length: the Akasa Mirage is a big case, but I had no issues with any of the cables being long enough. Since there's some excess cable length, a few branded cable ties would have been a nice bonus.


The PSU

The PSU (apologies for the blurry pictures).

The front includes the switch for the fanless mode:

406523560-675105938022878-7082571164584592020-n.jpg

The side label:

406523562-675106044689534-1675326036201193987-n.jpg

Label is placed so that the fan is up (i.e. intended orientation: visible when placed at the bottom of the case).

3.3v - 20A
5v - 20A
12v - 70.5A (846 watts)

The PSU's rear:

406519930-675105984689540-526674741201751652-n.jpg

The connectors are numbered.

1 and 2: Motherboard.
3, 4, 5 and 6: SATA or PATA.
7: 12V HPWR.
8, 9, 10, 11: CPU or PCIE.

I suggest plugging in the SATA cables first (3-7). The motherboard and SATA ports have the clip on the top, the CPU/PCIE ports have the clip on the bottom.

On the other side of the PSU there's an embossed logo. You can find a picture on Kolink's website here.

Logo is placed so that the fan is down (i.e. intended orientation: visible when placed at the top of the case).


Testing

The (15 year old) S12II-500.

406508074-675106018022870-8019788017186845602-n.jpg

The test system:

i5-10600 (@ 3.3 Ghz)
Asus B560M-K
16GB Kingston 2133 (2x8)
Sapphire Pulse RX 6600
3x SATA SSDs

407359573-675105728022899-7399690588464941383-n.jpg

Results:

(powered with Seasonic S12II-500)

Idle @ desktop: 30-31 watts
Watching YouTube (The Batman trailer), Edge: 49-53 watts
Prime95 (Blend): 81-82 watts

Superposition Benchmark, Performance, 1080p Medium (DirectX): 166-170 watts
Score: 14090, min FPS 73.60, avg 105.39, max 131.85

3DMark06 Benchmark (1280 x 1024, AA: None): 62-73 watts
Score: 30505, SM2.0 10573, HDR/SM3.0 15435, CPU Score 9042

(powered with Kolink Regulator 850)

Idle @ desktop: 25-26 watts
Watching YouTube (The Batman trailer), Edge: 45-48 watts
Prime95 (Blend): 74-75 watts

Superposition Benchmark, Performance, 1080p Medium (DirectX): 151-158 watts
Score: 14093, min FPS 73.34, avg 105.41, max 133.12

3DMark06 Benchmark (1280 x 1024, AA: None): 53-64 watts
Score: 30378, SM2.0 10511, HDR/SM3.0 15414, CPU Score 8969


I did take screenshots of the sensor readings, but I don't think there's anything to see there:

406494155-675105578022914-8232671715162869675-n.jpg

406517519-675105574689581-7196922615573033861-n.jpg

My take:

The difference in power usage/efficiency was bigger than I expected, especially at idle with a much more powerful PSU. With a modest PC the saving isn't that big, but with a more demanding system, then it might be.

Some example figures, with assumed numbers (I added 10%, which aligns roughly with the Superposition results):
  • 350 watts while gaming, versus 385 (saving 35 watts)
  • 35 watts over 5 hours gaming is £0.05 (@ £0.27 per kWh).
  • Annual saving: £18.25.
  • Payback period based on PSU's list price of £110: 6 years.

Bonus pictures for cable spaghetti (aka: benefits of a modular PSU).

406507917-675105788022893-1788584692846702358-n.jpg

406506138-675106054689533-246445147435550800-n.jpg

Summary

(The PSU: packaging and build quality)
The box is robust, with plenty of padding. The PSU feels solidly put together, with a solid black metallic finish. The cables have tough black sleeving, while still being flexible enough to cable manage easily.

The connectors have a satisfying click when plugged in and they're marked for the PSU-end and for the device-end.

The only capacitors I could see through the PSU vents had "TK" on them, which I believe is Toshin Kogyo. The bulk caps and polymers were not identifiable.

I never heard the fan spinning and there was no audible buzzing or coil whine with my setup (at idle, or while benching).

(The PSU: features)
The PSU's features are competitive with other ATX 3.0/PCIE5 PSUs.

It has a modern design: LLC and DC-DC with a single 12v rail, so shouldn't have any of the problems associated with group regulated PSUs (i.e. issues with idle states, crossloads, or transients) and high efficiency.

There's a 'proper' 12VHPWR cable with sense pins present on both sides of the connector, which is reassuring and enough PCIE cables for any current graphics card.

The low idle power and fanless (passive) mode are useful if the PC left on a lot (since it minimises fan wear and excess noise), or for light server use, where there is infrequent load.

There's plenty of protections, which should protect the PSU and the attached devices and while I can't verify their operation, I can confirm inrush protection from the PSU's click when turned on/off.

(Final thoughts)
I've never had a PSU from Kolink before, so I didn't know what to expect, but from opening the box, plugging it all in and running the tests, I was very happy with the quality of the unit I received. There was good attention to detail in the packaging, the PSU (like how it works for both orientations) and the labels for the connectors. The number of cables supplied is appropriate for almost all gaming PCs and suitable for the PSU's wattage.

The feature list suggests a serious intent to compete and is better than some similarly priced units, e.g. Be quiet's 12 M, MSI's GL and Corsair's RMe don't have a switchable passive mode, or 100% Japanese capacitors.

Unfortunately I don't have a test PC powerful enough to draw the full 850 watts, but ATX 3.0 compatibility suggests that it has been designed to support the spikes/transients of modern GPUs without tripping and there are tighter efficiency targets for light/very light load, which is something my testing can confirm. There were no glitches or stability issues observed with the test system when powered by either PSU.

Feel free to ask any questions about the review.

Thank you to OCUK and Kolink for the sample.
 
Hi.

With the computer shut down, how much wall power was consumed with the PSU switched on? How much was consumed with the PSU switched off? I have a 1200W one that consumes more power (19W) when it is switched off, than when it is switched on (12W)!
 
With the computer shut down, how much wall power was consumed with the PSU switched on? How much was consumed with the PSU switched off? I have a 1200W one that consumes more power (19W) when it is switched off, than when it is switched on (12W)!
I'm afraid the test PC is not accessible right now, but I don't remember any weird behaviour like this. By switched off, do you mean the PC is powered down, or the PSU itself is turned off? If you mean the former, that could be due to the motherboard's behaviour rather than the PSU.
 
With the PC powered down (wall socket switch on, PSU switch on, computer in an off state), 12W is consumed by measuring at the wall socket. Then, with the PC still powered down, but with its PSU switch off, 19W is consumed. So, in both cases, when the PC is powered down and the wall socket on:
  • PSU switch on: 12W
  • PSU switch off: 19W
I've since discovered the 12W when PSU is on is because ErP was disabled. With ErP enabled, it consumes just over 1W with the PSU switch on.

However, it still consumes 19W when the PSU is switched off, and the only thing switched on is the wall socket. Some web searching suggests that the PSU switch may not be connected to the main AC input, which may be by design.
 
Yep, quite counter-intuitive, and for it to use more power!
Yeah, I don't know why that would be the case, maybe the alternative circuit is less efficient. If I manage to dig out the test PC I'll give it a go and update the thread.
 
Back
Top Bottom