Swiftech/Laing D5

[chorner]

EDIT: see below for a couple D5 mods. Thread also talking about D5 versus DDC in terms of performance versus silence versus PWM etc. At least a bit of interesting info in here.

I just got a new pump as I can't decide between my MCP35x and the D5 vario for a new build.

However, I've noticed no less than 2 different performance ratings on this D5. Confused, and wonder if anyone else has seen this/if it's anything to be concerned about.

On the box it says:

2Amps
50PSI
10 feet head
~317GPH

On the motor the sticker says:
1.9amps
29PSI

50 and 29 PSI is quite the difference between the two ratings. Is there some kind of pump revision difference, or? Swiftech can't seem to even agree on their own ratings ;)

[rge]

PSI is just maximum pressure/pounds per square inch that the pump housing is rated for use, ie it is a product of material/construction, and has nothing to do with pump performance. mcp 355 are plastic/noryl whatever, ie not metal and they are rated at 22 psi. Laing makes D5 in Noryl that is rated at 21 psi according to this (http://www.dangerden.com/store/files/images/pumps/d5/D5_Vario.pdf), and brass housing rated much higher.

Neither d5 nor mcp 355 has history of problems handling psi in our loops, ie no housing exploding with leaks.

I have seen D5 with 50 rating psi at different sites, and 21 rating at laing for Noryl, and now 29 you see on pump. Maybe laing stamps them 29 psi rating, and someone else retested and found they are ok to 50 psi. Either way, wouldnt worry, but dont know why difference, other than either is ok and neither is relevant for performance.

[chorner]

PSI is just maximum pressure/pounds per square inch that the pump housing is rated for use, ie it is a product of material/construction, and has nothing to do with pump performance. mcp 355 are plastic/noryl whatever, ie not metal and they are rated at 22 psi. Laing makes D5 in Noryl that is rated at 21 psi according to this (http://www.dangerden.com/store/files/images/pumps/d5/D5_Vario.pdf), and brass housing rated much higher.

Neither d5 nor mcp 355 has history of problems handling psi in our loops, ie no housing exploding with leaks.

I have seen D5 with 50 rating psi at different sites, and 21 rating at laing for Noryl, and now 29 you see on pump. Maybe laing stamps them 29 psi rating, and someone else retested and found they are ok to 50 psi. Either way, wouldnt worry, but dont know why difference, other than either is ok and neither is relevant for performance.

Very interesting, thank you very much for that information. I was always curious what the PSI rating was and I thought it was directly related to performance.

Good to know :up:

[Utnorris]

Personally, I would stick with the MCP35x and skip the D5. The MCP35x is PMW capable so you could set it up to run on high when you load and slower when the load is lighter making it quieter and more efficient.

[chorner]

I've thought of that Utnorris, but ideally you should just set your loop to run at an optimal GPM and leave it there idle or not. If you drop your flow way down at idle, your loop won't cool back down nearly as fast after load, no? Dropping the MCP35x to 'silent' for me means 1,800rpm or so at very most and for my loop, that's not going to get me the proper flow I need anyhow. From all reviews (and especially Martins pump noise review) the D5 takes the crown squarely with a good after-market top. The Koolance 450 with the their mod kit is very silent from settings 1 - 5 never reaching over ~32dba even at full, while the MCP35x at 2.5gpm spikes up between 37 and 38dba... no thank you. The DDC can keep their added 6+dba difference even with properly decoupled.

For me, as long as I can get the quietest pump possible while maintaining that target 'minimum' then I really don't care if one pump has better maximum output over another. To me, I'm leaning towards the D5 especially after modding it!

Pictures attached below of my new Swiftech MCP655 vario. I threw on my own Tach wire as I don't know why the hell it doesn't come with one stock, put on the EK D5 xTop Rev 2 and... I've used noised dampening material inside the motor can and under the lid. This should prove to be an excellent noise/performance combo much more than you can get out of the tiny little DDC pumps with the motor basically exposed... all it has covering it is the tiny bottom box, which is both thin and hard/brittle which doesn't do anything to reduce motor noise.

http://www.efluxdesign.com/parts/d5/d51.jpg

http://www.efluxdesign.com/parts/d5/d52.jpg

[Martinm210]

I'm throttling my 35X, and LOVE it that way. I also have my fans on a sunbeam rheosmart doing the same speed throttling.

I can not express enough how awesome it is...try it..you won't regret it.

I USED to think along those same lines of adequate pump and fan speeds, but in the end it's a middle ground compromise between performance and noise. My office ambient with my old fixed pump and fixed fan speed setup was around 42dbA.

Now using PWM to dial down both the fans and pumps, I can set the fans at a higher speed and pump full speed as well. I get full out performance when my CPU hits 42C or higher which never happens in the desktop environment short of benching and gaming.

More importantly, my noise level is WAY down 99% of the time, from 42dbA to 35dbA...that's huge.

If you haven't already tried throttling pumps and fans via PWM or other "Smart" thermal throttling, you are missing out.

Only thing I see is about 1C higher idle temps, but who cares about 29 vs 30C on idle. My pump also runs very cool in comparison when it's drawing a few watts vs 20 something..... Only time I hear load speeds is when rendering video or gaming in which case I either walk away for the render or I'm wearing headphones to game with.

Pure bliss..

D5 is a great pump too, but PWM is soooooo nice!

[chorner]

IMO, a revised D5 with a larger impeller like the DDC3.25 and PWM would be the ultimate. Would be sweet take the internals of the DDC3.25 and put it in a larger silenced can like the D5 with a larger denser top :up:

However... Martin, if I can keep the noise as low as the MCP35x is between 10-20% PWM (if not lower) and keep this D5 pinned up for optimal flow at all times, does that not negate the benefit of PWM at this point? *wink* It definitely does.

PWM is only a benefit if your pump is too noisy for you at its optimal setting for your loop. I have a 35x hooked up right now under my current build which is controlled by a curve I've setup on my Aquaero 5. Previous to that... I was using EasyTune 6 to control the PWM of the 35x. Well, of course... there's the fact that it uses less power when throttled down at idle. So I guess you can't deny that bonus as well as the DDC's being slightly more efficient.

I bought this 980x one because I love hardware and having the best, but also because I use it for work - as well as a little play and to me, the whole point of water cooling outside of the added benefit of giving more headroom is to be quiet at all times. Making it louder while gaming isn't an option either, though is more tolerable since a) I've got my pair of studio monitors blasting the sound or b) later at night I use headphones. But, it certainly isn't tolerable while rendering or doing the multitasking I do on a daily basis for working. I get to work from home as a small business owner very often - and my office including computer setup like you're mentioning, doesn't tip over the 35dba point. If anything, like you can find over on silentpcreview.com you can have a system that performs well at no more than 20-25dba - with air cooling, and granted not a high end video card however, but that's where water cooling takes over. My point about buying the 980x is that I use it as a workstation with a good deal of multitasking, so I can't have this thing throttling up all day while working so for me, I pretty much just set the PWM to stay at max of ~24% to keep noise down. Optimally, outside of saving a few watts I'm going to pin this D5 at full (or as high as I need to keep optimal flow rate) and leave it there and it should be much more quiet than a 35x at 50%+

Right now, my Enermax PSU is the absolute loudest part on my system and they are one of THE brands of PSU to have for a quiet setup. Unfortunately, I'd love a bit more headroom and in my new system one of my things is to use this D5 with my mods above, and replace my PSU fan with a 600rpm GT fan which will be the last two pieces of the puzzle.

I'm willing to bet I can beat your 35dba system no problem, and keep my flow rate up to boot ;)

[CedricFP]

I'm throttling my 35X, and LOVE it that way. I also have my fans on a sunbeam rheosmart doing the same speed throttling.

I can not express enough how awesome it is...try it..you won't regret it.

I USED to think along those same lines of adequate pump and fan speeds, but in the end it's a middle ground compromise between performance and noise. My office ambient with my old fixed pump and fixed fan speed setup was around 42dbA.

Now using PWM to dial down both the fans and pumps, I can set the fans at a higher speed and pump full speed as well. I get full out performance when my CPU hits 42C or higher which never happens in the desktop environment short of benching and gaming.

More importantly, my noise level is WAY down 99% of the time, from 42dbA to 35dbA...that's huge.

If you haven't already tried throttling pumps and fans via PWM or other "Smart" thermal throttling, you are missing out.

Only thing I see is about 1C higher idle temps, but who cares about 29 vs 30C on idle. My pump also runs very cool in comparison when it's drawing a few watts vs 20 something..... Only time I hear load speeds is when rendering video or gaming in which case I either walk away for the render or I'm wearing headphones to game with.

Pure bliss..

D5 is a great pump too, but PWM is soooooo nice!

I'm leaning toward 3 x 35x w/ PWM and a XSPC trip top or Dual + Single res from EK / XPSC. Why 3? Well I'm tricking out the loop with QDCs and making every part removeable.

However, my question would be, is there any way I can get all 3 pumps hooked up to the same PWM header on the motherboard?

Thanks.

[chorner]

Sure you can - easily. Buy a 3 way 4 pin splitter wire (FrozenCPU has one I'm sure) and cut the RPM sensor wire leading to two of the pumps and you're good to go.

[Martinm210]

IMO, a revised D5 with a larger impeller like the DDC3.25 and PWM would be the ultimate. Would be sweet take the internals of the DDC3.25 and put it in a larger silenced can like the D5 with a larger denser top :up:

However... Martin, if I can keep the noise as low as the MCP35x is between 10-20% PWM (if not lower) and keep this D5 pinned up for optimal flow at all times, does that not negate the benefit of PWM at this point? *wink* It definitely does.

PWM is only a benefit if your pump is too noisy for you at its optimal setting for your loop. I have a 35x hooked up right now under my current build which is controlled by a curve I've setup on my Aquaero 5. Previous to that... I was using EasyTune 6 to control the PWM of the 35x. Well, of course... there's the fact that it uses less power when throttled down at idle. So I guess you can't deny that bonus as well as the DDC's being slightly more efficient.

I bought this 980x one because I love hardware and having the best, but also because I use it for work - as well as a little play and to me, the whole point of water cooling outside of the added benefit of giving more headroom is to be quiet at all times. Making it louder while gaming isn't an option either, though is more tolerable since a) I've got my pair of studio monitors blasting the sound or b) later at night I use headphones. But, it certainly isn't tolerable while rendering or doing the multitasking I do on a daily basis for working. I get to work from home as a small business owner very often - and my office including computer setup like you're mentioning, doesn't tip over the 35dba point. If anything, like you can find over on silentpcreview.com you can have a system that performs well at no more than 20-25dba - with air cooling, and granted not a high end video card however, but that's where water cooling takes over. My point about buying the 980x is that I use it as a workstation with a good deal of multitasking, so I can't have this thing throttling up all day while working so for me, I pretty much just set the PWM to stay at max of ~24% to keep noise down. Optimally, outside of saving a few watts I'm going to pin this D5 at full (or as high as I need to keep optimal flow rate) and leave it there and it should be much more quiet than a 35x at 50%+

Right now, my Enermax PSU is the absolute loudest part on my system and they are one of THE brands of PSU to have for a quiet setup. Unfortunately, I'd love a bit more headroom and in my new system one of my things is to use this D5 with my mods above, and replace my PSU fan with a 600rpm GT fan which will be the last two pieces of the puzzle.

I'm willing to bet I can beat your 35dba system no problem, and keep my flow rate up to boot ;)

Don't forget the rather substantial power differences. While I may not be one for running dual pumps, the 35X is very much like running 1.5X D5s in series...and it does it with less power consumption. From a pure raw power and least heat dump perspective, the 35X is superior without a doubt.

The 35X at 20%PWM is quieter than a D5 at full speed, even quieter than one with a top. Sure D5 is quieter when both are at full speed, but I'm not running it at full speed except and only when it needs it. Mine runs at 28% idle, 70% load and rarely ever ramps up to load except for gaming or rendering HD video, etc.

35X minimum noise level =31.5dbA
D5 stock top max noise level = 35dbA
D5 with aftermarket top max noise level = 33dbA
http://martinsliquidlab.files.wordpress.com/2011/03/pnr1-b-comparisondecoupled.png?w=569&h=756

With PWM you can have the best of both worlds. 31.5dbA when you want silence doing office like tasks, and 40% more power than a D5 when I'm loading.

PWM is awesome that way, silence and power. I wish all pumps had PWM..:D

You can do the same thing with advanced fan controllers using voltage control, but PWM provides more RPM range and it's cheap to deploy.

[chorner]

Very interesting Martin, great stuff :)

Ok, so... can we make a comparison here then with an example system and figure out which pump will be more quiet at a given flow versus power consumption? I guess basically pulling your calculations from your spreadsheet. Of course, negating the advantages of PWM and the small power draw differences (I do agree, PWM is much better than voltage control). My focus is 100% noise reduction, and to run a loop at a minimum acceptable flow level - I'm aware too that at idle, a couple extra degrees dropping pump pressure/speed ... so what, right? I'd even gladly give up a few degrees under load no issue for silence which was my main objective above all else with water cooling :)

Can we assume basically that:

Maximum head pressure at ~1.0GPM

MCP35X Stock: 6.3 PSI
MCP655: 4.7 PSI (with my EK top, it'd probably be more like ~5 PSI I think I saw)

So lets map out a system with approximate PSI drop per component.

EK Supreme HF (Ni) : 0.7 PSI
EVGA GTX 580 HC2 : 1.0 PSI (wild guess, can't find anything)
Coolgate 360 rad: 0.2 PSI
Reservoir : 0.05 PSI
Tubing and fittings 1/2" : 0.3 PSI (seems about right?)

Total = 2.25 PSI, and lets be generous and say 2.5 PSI


So, knowing the required PSI to meet a 1.0 GPM flow rate, at setting 5 on the D5 vario there is ample total dynamic head pressure to feed this loop. With this data, you could figure out the required pump settings on each to match this roughly, but... from just your data above and nothing else the D5 vario is noticeably more quiet than the MCP35x. Anything above a couple dba's you're going to hear a difference over a very quiet system and I would suggest that you could definitely drop the D5 down in the scenario to somewhere between setting 3 - 5.

You wouldn't need anything over about 1.2 to 1.5GPM flow rate correct? This is where the D5 really takes off noise wise on your chart above, and the MCP35x takes a sharp spike upwards in dba while the diba hike on the D5 vario is relatively 'flat'. This is further accentuated when you take into account in your pump tests you're mounting the DDC pumps flat on your foam - where for optimal cooling on the DDC's you're going to want to lift it off the foam, and this means exposing the under side of the DDC pump where the thin brittle plastic is a great sound enhancer. I bet if you mounted your MCP35x on its side you would have had increased your dba measurements without a doubt.

Pure noise wise to keep your loop going reliably .... the D5 wins. Especially with a good top like the EK, and what I've done by further dampening noise inside the motor can.

I would never need to modify the speed because as we can see, it covers it beautifully. Outside of the extra few watts of power the D5 takes... there's a benefit to the D5 in which the motor can is water cooled so you don't have to worry about proper mounting of the DDC as much, and compared to the MCP35x tops (I have the EK Xtop 2 on my MCP35x) you can trade a little performance on the D5 for barb flexibility.

However, again there is no denying PWM is much better for regulating the pump motors as with PWM you reduce power losses (increased efficiency). When can we get a modified D5 with PWM and the impeller/RPM control curves of the DDC3.25?? :D

Overall.... I'll take the D5 :D

[Utnorris]

Well, just to toss this in the mix, I run a PMP400 and an MCP355 with a Heatkiller dual top and they make absolutely no noise sitting on a 1/16" of foam. Seriously, I can turn my fans off and still cannot hear them. I get what you are saying about wanting the best noise/performance ratio and for me the only way I found it was by trying various setups. I tried single D5's, not enough flow. Two D5's in series, just enough at 1.04Gpm, but could hear them slightly, a RD30 which gave me 1.34Gpm at 18v, but was loud compared to the rest and then finally the dual MCP355 setup I just stated which gives me around 1.2Gpm and I cannot hear it. Now, my system isn't uber quiet, I have Triberk fans, 11 of them, but with a fan controller and them dialed back to 1000-1100Rpm I can hear cars passing outside and the stupid crickets chirping away. Even at the RPM, I have a 3 degree delta between water temps and air intake, so not too bad.

[chorner]

Nice Ut :) I guess obviously my playing isn't finished yet, though I don't ever plan to build a loop that requires more than 1 pump (yet, anyhow) - but I will say, I won 3 free Triebwerks from Dazmode and tossed them out of the system after a week lol the bloody things even when under volted had a craptacular motor hum to them IMO. This was with the slow speed down at 700rpm. With 11 of them, I think I might go mad! haha so I think I'm probably a touch more picky with noise tolerance as I couldn't even bear 3 of them at 300rpm less never mind 11. If anyone can't hear cars at 12am and crickets over their computer in the first place.... I don't know but I personally wouldn't know what the hell to do with myself lol I don't find it acceptable, especially sitting on a computer for 8+ hours a day for work etc. and more time in your leisure.

[coolhandluke41]

I think you are on the right track here ,D5 is the quietest pump of them all and looking at your sig will provide more then enough flow for your set up
even at setting (3),i have similar set up ( two blocks ..CPU/GPU) and test multiple times @3 setting and 5 (P95 @4.7 for around 30 min).. i would be surprised if you see more than 1C difference ,
so if you don't have MB, RAM block stick with single D5 (most reliable and super quiet pump )

[chorner]

Thanks - I tend to agree :)

[Martinm210]

Good plan! Looks good and looking sharp with the top on there too.

Neat idea with the internal foam, have you noticed a subjective improvement with that?

[chorner]

I've only had the chance so far to test it on a small loop by my sink to make sure the pump is working properly; so that's to say that I don't know what it sounds like when loaded up with a few blocks worth of pressure but subjectively I definitely notice a difference. I tested it before I did the mod with the stock top, but unfortunately not just with the EK top on it before throwing the pieces of sound dampening mat inside I had kicking around.

I'd be interested to see just how much a difference it makes in numbers and not just subjectively; I'm willing to bet you'd notice a couple DB difference for sure in a best-case scenario. If anything, it muffles the pitch of the pump when running full out - especially compared to the 35x which has a much more pronounced high pitched note to it in comparison.

Overall, I'd definitely do it again and it was worth the extra 10-15 minutes of cutting to size and wrestling with the sticky side of the foam to get it placed in there nicely :D