First I wanted to give a special thanks to those people who really made this test availailable.  The following folks are truely committed to providing the water cooling community better information and they provided me either loaner samples to test and return or product samples I can keep(I'm all for more water cooling goods!):

Coolmiester (XS member name) of Coolercases UK - Please visit their website and forums.  They provide a great resource to the community and wide range of products.  Coolmiester sent me the following loaner samples:  Petra's DDCT-01s Top, EK-DDC X-Top G1/4, EK-DDC X-Top G3/8, Alphacool DDC plexitop, and the Alphacool DDC Plexi Reservoir.  Five of the testing tops, thanks a bunch, very much appreciated!

Petra of Petra's Tech Shop - Provided me with a loaner sample of the Koolance DDC top.  Since I already had a a DDCT-01s top, they offered and sent me the Koolance top they had from thier own testing. Petra's Tech shop is my #1 pick for shopping in the USA, they are like the newegg of water cooling for us USA folks, only 10X the customer service.  I particularly encourage you to shop there, because they are also truely committed to testing and finding out the truth about water cooling products.  That means worlds to me and I order stuff from them all the time.  How many Petra pens do YOU have?...:)


Paul from XSPC provided me with both the XSPC Laing DDC top and their newly release XSPC Acrylic Reservoir DDC Top product samples.  A UK based company with some very refined performance products.

Jeremy from Danger Den provided me with a sample their Danger Den DDC acrylic top. Danger Den is another great manufacturer and reseller of all type of water cooling products, based here in my home state of Oregon in the USA.  As you'll see in many of my other reviews they are very committed to producing products with good flow rate capabilities.  I've really enjoyed reviewing many of Danger Den's products, they are great folks that manufacture great products!!

NaeKuh (XS member)

Another very active member in the water cooling comminuty always out to try new things.  I've been building him a TEC chiller and while I was at it he sent me his EK X-Res 140 . He also sent me an alpha cool dual top that I will test soon,  unfortunately it's not DDC 3.2 compatible (inlet nozzles) and I'll need a different manometer, but I will fix it and get this tested soon.

Thanks....you all made this possible!!

The Pump Used for Testing - Swiftech MPC 355 (18watt), or Laing DDC 3.2

The pump I used in all of this testing is one that I recently purchased at Petra's, it is the newest iteration of this popular little hard working pump.  Very possibly the most popular pump out there for water cooling.  So $75 bucks later along with Petra's amazing customer service, I had a happy little pump sitting on my test bench.  I happened to be a long time Swiftech MPC655 (Laing D5) vario user, so it was fun trying out this new little pump.  The older and more powerful DDC2 is no longer in production, while performing better it also had some proplems with pump failure.  So far this new DDC 3.2 pump appears to be doing much better, I havn't heard very many problems with it, and it appears to be a completely redesigned pump.  Anyhow, I'm not here to talk about the pump specs, there's plenty of that available on the internet, but I've linked you to some good articles down below here. It's a very good and popular pump.

Swiftech's website on the MPC355 - They rebrand this pump and include their own accesory package.
Laing's website for the Laing DDC3.2 - They manufacture this pump. DDC3.2 brochure.

One little bit I would like to add before I go on to the tops is a little testing I did with the stock pump to explore over/undervolting potential.  In the end I found that you get some decent gains from overvolting and you can really turn the pump down ALOT undervolting.   Unfortunately it's a little scary because exceeding some very defined limits will not allow the pump to start.  It appears there is some sort of protective circuitry built into the pump to prevent you from overvolting the pump very much.  I found that limit to be around 12.95 to 13.00 Volts and it was a limit on startup only.  I could start the pump at 12.00V and turn voltage up to 13.5V, but after turning off my PSU and trying to start it at 13.00 more, it simply would not start.  This could be a very bad situation if you were overvolting to 12.9V and just happend to have a surge of power that resulted in your pump not starting because your PSU surged and produced 13.0V!... Up to you if you want to overvolt, but it definately has some risk (possible failure to start and overheating).  I'm not very experienced in ATX power supply specifications, but I know my current PSU(OCZ GameExtreme 700) actually produces around 12.4V on average, so the 12.9V limit is probably designed to give just enough room for differences in your average power supply, very little extra.  Not sure it would be worth the extra .1GPM, and if you wanted to undervolt you could modify the DDC3.2 to be a DDC 3.1 by removing the soldered bridge.  DDC3.2 VS DDC3.1 Photo Difference

So you'll get about .1GPM gain with the extra overvolting, 10V runs pretty close to a DDC3.1, and 7V is really weak and nearly useless. In the end I don't think it's worth messing with, I'd suggest running it at 12V for maximum performance or mod it down to a 3.1 for complete silence, then you don't need to worry about any special power supply and exceeding the limits of voltage operation.

P/Q (Pressure Vs. Flow Rate Curves)

First and probably the most important is the ability to measure a pump’s PQ curve.  You’ve probably read in your own pump specs two general numbers that are specified.  One is typically the maximum head or lift.  The second is the maximum flow rate.  These two numbers are two points that are part of this curve.  PQ stands for P being pressure and Q being flow rate, so in the end it’s just a curve that represents the relationship of pressure and flow rate.

Unfortunately these two numbers or points on the curve only indicate points of value where our pumps never operate under a real water cooling system.  Maximum flow only occurs when there is absolutely no resistance.  And maximum pressure only occurs when you’ve completely stopped the flow (pinch and kink your tubing, now you have max pressure, but zero flow).  Furthermore they do not provide you with information about the curve or path in between these two points (The results will show you how important curve "shape" is as indicated by the DD top".  There are all types and shapes of this curve and it’s critical to plot the entire curve and particularly look at the range between 1 GPM and 2.5 GPM because this range is where the system pressure drop curves typically intersects the pump curve.  Where these two curves intersect is what determines your water cooling system flow rate.  So pressure vs. flow rate is “The” ideal measurement of pump performance that determines flow rate.  After collecting this data, I’ll evaluate these curves where it is important and explain what they mean in more detail.

Power Consumption vs. Flow Rate

In addition to measuring pressure and flow rate, I also collected the pump power draw in terms of voltage and current (amps).  Voltage was necessary to ensure a proper 12.0V being fed to the pump anyhow, and it’s very easy to add another multi-meter in to measure current.  This provided me with the ability to measure how many watts were being drawn by the pump at each test point.   This information is not particularly important in comparing tops using the same pump motor, but it is EXTREMELY important in pump selection because a large percentage 60-90% of the pump power consumption is dumped into the water cooling loop as HEAT!  The added benefit of collecting this information is explained later with efficiency comparisons.

Water Horsepower per Watt

In addition to providing the two above performance measures, I also wanted to provide some sort of means to measure efficiency.  I wanted to know how much “work am I getting per watt of power” and see where that peaks out.  I looked around the internet and found an equation to convert pressure and flow rate to “Water Horsepower”.  I already had how many watts it took, so I figured a measurement of HP per watt would be a good efficiency ratio to look at.  I also had this relative to flow rate, so plotting this ratio over flow rate there was a very distinctive parabolic curve that formed.  This isn’t particularly a standard format, but I think it’s valuable and if nothing else provides some means to see where the pump peaks in efficiency.  Considering most water cooling systems probably average in flow rates of 1.5 to 1.7 GPM, I would think a "perfect" pump would also seek to have an efficiency peak in this area.  You'll see later that the stock pump is lower than this and most pump tops are higher.  This tells me that there is still a little room for improvement and most of the tops are tuned a bit too much for higher flow rates and could use a little more emphasis on improving pressure in lower flow rates.

The DDC 3.2 Pump Top Contenders

The following alphabetical list of pump tops were included for testing:

Alphacool DDC plexitop

Alphacool DDC Plexi Reservoir

Danger Den DDC acrylic top

EK-DDC X-Top G1/4

EK-DDC X-Res 140 Reservoir Top

Koolance COV-PMP01P

Petra'sTech DDCT-01s Acetal Top

XSPC Laing DDC top

XSPC Acrylic Reservoir DDC Top

Stock DDC3.2 or MPC355 Pump with stock top

Coming Soon...I'll add these in as soon as I can...

EK X-Top G3/8 - Just need to order some G3/8 barbs, I have the top.
OClabs XP top - Out of stock at the moment, but OClabs said they would be sending me a sample to test.

The XSPC reservoir top is KING of the DDC 3.2 tops tested!

The following set of curves shows you how efficient each top is at different flow rates and shows you where the top is optimized as indicated by the high point on each curve.   The stock DDC 3.2 top is optimal at 1.1 GPM, so it's really not designed to operate at flow rates of 2.0 GPM.  The small inlet and outlets and particularly the sharp and small 90 degree bend at the inlet is reducing peack efficiency quite a bit.  The Alphacool reservoir top also must utilize the side barb which reintroduces the 90 bend at the inlet, so again peak efficiency is low compared to the other tops, but it is more efficient than some of the other tops up to about 1.2 GPM (high restriction).   So in general the more restrictive inlets seem to fair better in a high restrction (lower flow rate) system.  The EK G1/4 and Koolance top are peaking out around 2.25 GPM, and I belive this is due to the large inlet size.  Both tops are drilled completely through and tapped with G1/4 threading, this is beneficial for higher flows, but detrimental to pressure. The Danger Den and Alphacool tops are peaking around  1.7 GPM which is a good point to peak out as you're able to make the most of the curve produced for average to low restriction systems.

Overall I think most of the tops tend to be optimized for low restriction systems.  This leads me to believe there are some remaining tweaks available like smaller inlets diamters or different volute shapes that could improve the existing designs and provide more benefit to average and higher restriction systems.

Now lets look at the peak efficiency of the stock top vs. the XSPC reservoir top, that's .00017HP/Watt stock to .00030HP/Watt XSPC reservoir top, that's a 76% efficiency improvement!  All of the tops had quite an efficiency gain removing the small tight 90degre bend in the stock top.

It should be noted that in low restriction systems the tops do draw more power from the pump motor (brake horsepower) than the stock top would.  This would result in some added stress to the pump motor over stock.  But like any performance modification, nothing is for free.  These pump tops are alot like added an aftermarket air filter to your car or truck.  It's letting your motor breath a little better, but at the same time your motor can rev a little higher than normal.  This is the one case where more restriction is good...seems backwards, but a high restriction system will put less stress on your pump motor than one that is free flowing.

As noted all of the above curves were tested with the inlet barb located in the most efficient position which is the top (Pointing directly at the impeller Inlet) with the exception of the stock top and the Alphacool reservoir top which require a side inlet type of orientation.  I've had some folks ask if I could test the tops with the side inlet for those with paticular case installations that make it mandatory.  So out of the aftermarket tops, the Alphacool, Danger Den, and Koolance tops all provide that option in addition to the stock top.  I expected to see some significant reductions in performance and as expected here are the results.

I think it's pretty amazing that these pump tops can gain so much efficiency over the stock pump top, but they are.  Kudos for doing a great job there!!

The XSPC tops are performing the best in all system configurations, a true tribute to the design and tweaks that have occured in the current design...excellent!!  

The remaining tops all perform very well and have trade offs between each other depending on the system restriction.  Some are slightly more tuned lower flow rates and others for higher flow rates.

In the end, these tops can afford you as much as .8GPM gain on low restriction systems, .5GPM on an average restriction system, and .1GPM on extremely high restriction systems..... that's a good gain!!

Description	Detailed Test Report	High Restriction System Perform. Ranking	Average Restriction System Perform. Ranking	Low Restriction System Perform. Ranking	Inlet Dia. mm	Outlet Dia. mm	Volute Shape	Volute Depth mm	Reservoir Bleeding Performance	General Notes
Alphacool DDC plexitop	Test Data & Charts	#7	#7	#8	6.90	7.0	Circle With Relief Cut	7.9	N/A	Sharp Inlet Step. Used D-Tek Barbs.
Alphacool DDC Plexi Reservoir	Test Data & Charts	#4	#9	#9	6.90	7.0	Circle With Relief Cut	7.9	Some users have noted problems, but non apparent in testing.	Sharp Inlet Step 90 Degree Used D-Tek Barbs
Danger Den DDC acrylic top	Test Data & Charts	#3	#3	#7	12.00	5.75	Circle	8.5	N/A	Only tested Center Inlet. Used DD Fatboy G1/4 Barbs.
EK-DDC X-Top G1/4	Test Data & Charts	#9	#5	#3	11.69	8.85	Semi Spiral	7.6	N/A	Acetal/Delrin Very Nice Finish! Used D-Tek Barbs.
EK-DDC X-Res 140 Reservoir Top	Test Data & Charts	#8	#6	#6	11.60	8.85	Semi Spiral	7.6	See Youtube Video Here Recommend modifying and using anti-cyclone	Acetal/Delrin Very Nice Finish! Used EK Barbs.
Koolance COV-PMP01P	Test Data & Charts	#10	#7	#4	11.88	7.66	Semi Spiral	8.2	N/A	Only tested Center Inlet. Large step loss in dual outlet. Aluminum!!! Used D-Tek Barbs.
Petra'sTech DDCT-01s Acetal Top	Test Data & Charts	#5	#4	#5	9.50	7.00	Circle With Relief Cut	7.9	N/A	Acetail/Delrin Very Nice Finish! Used D-Tek Barbs
XSPC Laing DDC Top	Test Data & Charts	#2	#2	#2	7.18 & Taper	7.00	Spiral	7.8	N/A	Used XSPC Barbs.
XSPC Acrylic Reservoir DDC Top	Test Data & Charts	#1	#1	#1	7.18 & Taper	7.00	Spiral	7.8	See Youtube Video Here	Had trouble sealing reservoir cap. Used XSPC Barbs.
Stock DDC3.2 or MPC355 Pump stock top	Test Data & Charts	#6	#10	#10	6.85	5.8+-	Semi Spiral	7.8	N/A	It's stock.. 3/8" cast barbs!

Possible Design Tweaks

With that general trend for more restriction and nozzles, I would encourage any tweaks to these already suberb designs to give a little on the flow rate side and tune these pump tops for more pressure.  The engineer in me can't help but think there is still a way to make something already great even better..:) I think some of the tops could be improved by tweaks in the following:

• Utilize a true "Spiral" shaped volute - this is a little more complicated to design, but it only up front tool path design costs with CNC.  A circular volute is not the optimal shape to release flow around the perimeter of the impeller. Water is incompressible (for all practical purposes), so a spiral design with an ever changing radius to provide space for the impeller water coming off is the most efficient shape and allows the entire impeller to do work rather than primarily one side.  The semi spiral and relief cut to simulate a true spiral curve do not appear to be nearly as efficient as the full spiral.  It actually appears that a regular circular volute shape is doing better than those with relief cuts or two simple curves to simulate a spiral..
• Smaller Inlet Nozzle - I understand the manufacturing simplicity in using a 11.8mm drill bit to drill and then fully tap the inlet hole, but the size exceeds that of the impeller inlet hole, so all you do is introduce yet another sharp entrance loss and this is significant in the pump design.
• Smaller Outlet Nozzle - It appears the nozzle size really is optimally matching that of the impeller and going larger on the outlet port only help flow and going too big may hurt pressure.
• Experiment with tapered and protruding inlets - This is just a theory of mine, but I think something besides a straight tube inlet is beneficial as the water has to turn 90 degrees after entering the impeller.  If the jetstream of water is slightly moving towards the perimeter as it enters the impeller, I think that helps reduce that entrance loss.  A protruding inlet will also better seal the volute chamber from the inlet....you would just have to experiement and ensure impact is not possible which is difficult with the play possible in all but the stock top (which has cast alignment pins that accept the bolts and).
• Consider using a ball end endmill for the final perimeter cut of the volute.  Another theory of mine, but I think if the perimeter catching the water was radius and had a nice smooth spiral transition into the outlet port, you would gain even more efficiency.  The cross sectional shape of a circle or pipe is the most efficient shape (least amount of frictional surface area).  It appears that very small "details" are key in pump top design, so this may be another detail that returns with performance.
• Consider using recieving studs that extend from the pump top like the stock top.  This ensures the top is perfectly aligned with the impeller inlet and eliminates installation errors that can cause imperller to top impact and damage to the pump.
• Smooth finish - roughness coefficients can steal away alot of energy, and nice glass like smooth finish in the perimeter of the pump top and smooth transitions throughout can only be of benefit.

I hope you found this information useful.  If you're looking for a way to say thanks, you can send me a few bucks by paypal to ship all this stuff back to everyone..:)

If you have any additional tops to test or corrections in this collection of results please let me know.
Thanks!
Martin