IntroductionI have been testing out a lot of watercooling parts lately, and most of them happen to be those that the general public wants to know more about. As such, sometimes products slip through the cracks which may have benefited a lot of people and this may just be one of those. XSPC somehow manages to have watercooling kits at prices that make me question my math at times, and that's with the more popular Laing pumps. Today we get to take a look at their newest option for those constrained on space, and also budget- the X20 420 pump. This comes currently as part as a single bay res option, and also in a twin pump format with the pumps in series (in the same bay res). But XSPC has announced plans to have an even more compact res/pump option called the Ion. A big thanks to Paul from XSPC for providing a sample of the X20 420 bay reservoir model for testing.
Let's begin by taking a look at the specifications from
the product page:
Reservoir
Dimensions:149 x 42.5 x 79mm
Capacity: 105ml
Material: Clear Acrylic
Pump
4 Pin Molex
Maximum Flow: 420 lph (12V)
Maximum Head Delivery: 2m
Voltage: 12V
Startup Voltage: 9V
Power Consumption: 6W
Noise level: 35dBA
Maximum Temperature: 60C
Dimensions: 66.5 x 40 x 40 mm
Supplied with black faceplate, 4x M3 x 4mm screws, G1/4″ Plug, and blue LED
This is a single slot bay res/pump combo unit occupying a standard 5.25" bay in a chassis. The reservoir is clear acrylic and comes with an aluminum faceplate in the front as we will see soon. The pump is plug and play here, with a single full size MOLEX connector for power and nothing else. You don't get to monitor the pump speed, or control it any which way as-is. There are some devices that provide voltage controlled MOLEX outputs but I would not recommend them here. With a max pressure head of 2 meters (~2.85 PSI) and a max flow rate of 420 liters/hour (~1.54 GPM), you want to get as much out of the pump as possible. XSPC claims a power consumption of 6 W, and a noise level of 35 dBA but we will see soon how it holds up during testing. As with most pumps, it is rated to operate up to 60 ºC and I dare say if your loop hits that temperature then there is something wrong as it is. In terms of dimensions, it is longer than a Laing DDC is shorter in all other dimensions- especially once you have a heatsink on it. It also occupies a smaller volume compared to the D5, and this is one of the main reasons the upcoming Ion is so much smaller than most cylinder res/pump units. But enough of this, let's take a look at the product now.
Unboxing, Overview and Disassembly
The box arrives plastic wrapped, and has a clean look to it. The black/grey color scheme is similar to that of XSPC's other products, with the product name and XSPC written on the front and sides, and the back telling you which model of their single bay, single pump reservoirs this happens to be.
The contents inside include the bayres/pump unit in bubble wrap for protection, 4 screws to help secure this in a 5.25" bay and a blue LED that is powered by a single, full size MOLEX connector. The contents are pretty snug in the box with no space for moving around, and as such are packaged well enough to withstand standard delivery service handling.
Here we can see the unit outside of the bubble wrap. As mentioned earlier, the reservoir itself is composed of clear acrylic which is quite thick and the adhesive was very well applied here with no signs of it anywhere. From the front, we also see the thin aluminum faceplate that comes stock in a black anodized, brushed version. The reservoir part itself is on the left (as seen from the front) with the pump inlet in the middle, and the pump itself on the right. There is no top or bottom cover in the right compartment so people can use the accompanied power cable either way to suit their cable management needs. On the top is a large
M20 threaded black plastic fillcap but you can buy metal versions in a
chrome and
black chrome finish if you so desire. On the back, we see the single LED hole at the top right, and 3 G/14" threaded ports, two of which are inlets to the reservoir and one is the outlet from the pump:
There is a G1/4" stop plug included (metal fitting here) to cover the extra inlet that is not in use normally. On the side we see the pressed in threads that the screws go into and help secure the unit in place.
There is cutout in the top to route the power cable in between as seen above, but most may find it easier to navigate the cable from the bottom if you have this at the top of your case.
To remove the pump, you have to unscrew 4 screws (2 on each side):
The pump has a relatively simple design on the exterior with 5 blades on the impeller that essentially rotates fast and pushes the water coming in perpendicularly outwards like a water wheel. The impeller shaft is held in place by a small C-clip and I did not take it out- mostly because I was not able to with the tools I have here! We can also see the "top" in the reservoir itself where the water enters the gaps in the blades of the pump impeller and get pushed out the outlet port.
The window in the faceplate allows for a look at the coolant level in the reservoir. XSPC rates the reservoir to hold a coolant volume of 105 mL and mine held closer to 108 mL (close enough). The provided blue LED makes it easier to take a look at the coolant level. A potential issue here is that the inlet port is in the middle (height wise) and air bubbles will be sucked in by the pump. As such, bleeding this will take longer than dual bay reservoirs and definitely more than cylinder style reservoirs.
But if black isn't your desired color, there are
white,
red and
silver faceplate options available for all their single and dual bay reservoirs. Here's the white replacement faceplate kit for this single bay design:
The kit comes with 4 replacement screws but since they are black anyway, I don't see the point as one could just use the stock screws. I would have rather seen color matched screws provided, but also recognize that white anodized aluminum screws are not easy to come by. An L-shaped hex key (2.5mm) is provided to help with installation, and the faceplate itself comes wrapped in soft foam. It is fundamentally identical to the stock black faceplate aside from the color, of course.
The white is powdercoated over the black faceplate, and is slightly creamier of a white than the new white Raystorm Pro mounting bracket as seen above. XSPC is aware of this and will be working on QC to get the colors the same henceforth. Installing it is fairly simple with 4 screws to unscrews, lining up the faceplate and screwing it on. Now let's see how the bay res/pump unit performs.
Performance Testing
Due to the nature of this product, I have to test it as a bay res/pump combo unit. There is no dedicated top to the pump alone. As such, the P-Q curve below to show how the pump handles liquid flow restriction and manages loop flow rates has this specific reservoir built in. The pump was powered directly from the PSU and fed 12 V constantly. A Dwyer 490-1 wet-wet manometer with the high pressure and low pressure ports between the inlet and outlet of the res/pump unit measured the pressure drop from the loop (which had a valve to help increase restriction easily) and a King Instruments rotameter helped measure the average flow rate. In order to provide context, I did the same with my other bay res/pump product here- the XSPC dual bay D5 with a Laing D5 Vario pump. I set the D5 at setting 5, and 3 (you will soon see why) for a total of 3 P-Q curves including the X2O 420.
Generally speaking, you would like to see a pump have a good balance of pressure head and flow rate. The higher both are, the better of course. The Laing D5 has great max flowrate but at the expense of slightly lower head compared to the Laing DDC and Swiftech MCP50X pumps. As such the D5 is recommended for a lot of low-medium restriction loops, and being quieter and cooler than the DDC makes it one of the most popular pumps in this industry.
The XSPC X20 420 is in a similar position in that while it has lower max head and flowrate than the D5 on Setting 5 (4800 RPM here), it can be compared with the D5 Vario at Setting 3 (3500 RPM here). An average loop with a CPU block, a GPU block, some fittings and a radiator will have enough restriction to the point where both these pumps will give you a similar flow rate. Compared to the older X2O 750, the X2O 450 has a higher pressure head but lower max flow rate (and is quieter) and this worked out better with most loops.
But enough of guessing, let's try it out in an average loop and see. I used a Raystorm Pro CPU block, a Swiftech Komodo R9-LE GPU block, a HWLabs SR2 360 radiator and 2 Koolance QD3 quick disconnects to go with 6 Bitspower 1/2" x 3/4" fittings and Primochill Advanced LRT tubing in the same size. The blocks and radiator are one of the least restrictive in each category and the addition of 2 QDCs brings up the average restriction to what an average mITX loop would have. Aside from the D5 Vario at the two settings mentioned above, I also added in a Laing DDC PWM (At 100, and 50% PWM signal) and a Swiftech MCP30 (a slower version of the MCP50X, again at 100 and 50% PWM signal). Note that the DDC does not have a big difference in RPM response from 50% PWM to 100% PWM, and as such the respective numbers will be similar.
Finally, noise measurements were done as always in an anechoic chamber at 19 dBA ambient noise levels with the probe held 6" away in the front from the pump. No thermal measurements were done because that's not really the point here, and as such no fans were installed either to eliminate any other source of noise. All the loops were allowed to bleed air for 24 hours before measurements were taken.
As I mentioned above, the X20 420 comes very close in performance to a D5 Vario on Setting 3 in an example loop where the higher max flow of the D5 doesn't really come into play anymore due to the pressure head limitations. It is naturally outperformed by the full speed, full power D5 and DDC in general, and also the DDC at 50%, but it outperforms the MCP30 at 50% PWM signal. For those wondering about the custom 6 W DDC that EK has in their Predator units, it would perform similar to the MCP30 at 50% PWM signal as well. As such, the X2O 420 does a decent job- especially once noise levels are taken into consideration. The DDC is louder each time, as is the MCP30 and D5 at full speed. So there is the usual trade-off between performance and noise levels. Once more, the closest competition is the D5 Vario at Setting 3 which is slightly quieter but also performs slightly weaker than the X20 420. The noise level measured here is close to the reported number which was measured at 12", as opposed to 6" here so it makes sense also.
ConclusionThe XSPC X2O 420 Single Bayres/pump costs $52.99 in the USA and £49.99 (inc. VAT) in the UK as of the date of this article. This price point is where most dual bay reservoirs for pumps such as the Laing DDC and D5 come in at, except without the pump included. So you can get an idea of how much is being saved with this unit. As such, expecting high end performance from it is also not a great idea. As seen in the previous page, the X2O 420 performs similarly to a D5 Vario on Setting 3 in a typical loop comprising of an average flow CPU block, a GPU block and a radiator. As such, I see this working great for mITX builds where space is limited for the res/pump and also number of components that can fit in but can also work with a larger system if choose your components carefully. High flow blocks and radiators, no quick disconnects and this may well be enough for your needs. But if you are targeting higher flow (1 GPM, say) then consider another solution. There is a twin pump version of this that occupies the same footprint but has an extra power cable for the second pump which may work out, or you may have to go with a DDC based solution if you are limited to a single 5.25" bay. For those without available front bays, there's a cylinder res/pump version of this coming out soon which should perform similarly but have the advantage of easier bleeding and more installation options.
I had tried out the older X2O 750 that a friend had purchased and was not very impressed by it. It was quite loud compared to the flow rate it generated in a very similar loop and XSPC seems to have learnt well from it. As such, for those who have overlooked this so far (pleading guilty myself!), I would recommend a second look.