Danger Den MAG II LE (review & pics inside)

Alright, today I got my MAG II LE pump from Danger Den - first thoughts when i pulled it out of the box, GOD DAMN THIS THING IS SMALL!


its surprisingly light as well, and shipped with 2 peices of double sided sticky tape to mount it


here's pictures of the pump with a zippo (brushed chrome armor zippo if you need to know) as a size comparison:


http://www.fragtek.com/sneaky/illuminati/ill_31.jpg
http://www.fragtek.com/sneaky/illuminati/ill_32.jpg
http://www.fragtek.com/sneaky/illuminati/ill_33.jpg
http://www.fragtek.com/sneaky/illuminati/ill_34.jpg
http://www.fragtek.com/sneaky/illuminati/ill_35.jpg



and inside it:

http://www.fragtek.com/sneaky/illuminati/ill_36.jpg


inside, it looks like the impeller is made out of a white plastic, possibly acetal/delrin or nylon judging by the feel of it, but it is actually wrapped around the stator that drives it, and has the magnets build into the outside edge of the impeller



and with a pair of DD G1/4 High-Flow barbs installed:

http://www.fragtek.com/sneaky/illuminati/ill_37.jpg



when I get home from work today, I'll do some flow & pressure tests with it, and compare it with an 18v DDC w/ Alphacool top

oo im looking foreward to that!

Interesting design. Generally open-fin impellers are used for very viscous/fibrous fluids...

ANd whoa, thats tiny. Looks like a great pump for Shuttle/SFF watercooling.

yeah! looks like iwakis kan eat those for dinner!

I'm eager to hear how she performs... I ordered one to replace my LOUD D4 and it should be here tomorrow...

Give us performance! :)

This is a tiny pump!

can't see any pictures... anyone else have this problem?

No.

Cute :D Looking forward to some performance analysis!

I also hope to hear that the thing is dead silent, because that is what I am hoping....

hey you took my Zippo! ;)

Looking foreward to the review, please compare it noise wise to other pumps as well :p

At a glance at company specs (take it with a grain of salt):

MagII-LE:
•567l/hr
•2.5m head

It has more flow rate than the Pro DDc, but less than the Ultra, however it has less pressure head than either...

My prediction? Great little pump for low restriction loops...probably dead silent...

edit: Offtopic, am I the only person to ever somehow get the nut on the delrin fillport stuck on? I somehow managed to crossthread it (I think) even though it's screwed on the correct way..

Edit: Using a vice, an old shirt (to protect the thing) and a monkey wrench, I got it off...it's not damaged at all...it was just extremely stuck on..

Cant wait to hear what that pump does. It does look lie an awesome pump.

Last person who reviewed one said it vibrates like hell. And thats a nono in my loop. You see Im a hard lined bastard.

do they sell these somwhere else than on dds site? the shipping is so daam high...

should be a nice step up from hydor l20.... little less flow but twice the head and
less heatdump

Aren't most pumps put in the bottom of the case, between the harddrives and the motherboard PCI slots?
So wouldn't it have made more sense for the power cord to exit the other side of the housing?


I know its a stupid question, I just hate cable clutter, and stupid flaws like this that just show a lack of foward thinking.

i would put it on top of my psu so it suits me fine... :)

@jaguarking11


Originally Posted by introvert
In terms of noise, this little thing vibrates quite a bit. I had it setting on the counter running water through my rad and it sounded like a woodpecker. So i went and grabbed a piece of foam and i couldnt even tell it was running after that. Little to no motor hum what so ever.

Quote:

---

Originally Posted by Digz

@jaguarking11


Originally Posted by introvert
In terms of noise, this little thing vibrates quite a bit. I had it setting on the counter running water through my rad and it sounded like a woodpecker. So i went and grabbed a piece of foam and i couldnt even tell it was running after that. Little to no motor hum what so ever.

---

give that gel stuff from petras a try , the gel i clear and comes in 3"x3" blocks that are about half inch thick and clear they work amazingly well and you can't see'um

well poop... results were south of expected... extremely disappointed with the performance of this pump:

http://www.fragtek.com/sneaky/pumps/...umps%20010.jpg


and all it gave me was a meager dribble out of the hose
http://www.fragtek.com/sneaky/pumps/...umps%20017.jpg

compare that to a single DDC Rev. 2:

http://www.fragtek.com/sneaky/pumps/...umps%20005.jpg
http://www.fragtek.com/sneaky/pumps/...umps%20003.jpg


dual DDC Rev. 2's:

http://www.fragtek.com/sneaky/pumps/...umps%20018.jpg
http://www.fragtek.com/sneaky/pumps/...umps%20021.jpg

you should hold the output tube to 30 ~ 45 degree so we know how stronger and how far does it push, instead pointing it down

will do more stuff tomorrow night - tell me what you want to see and i'll do it

just did a quick 15 minute run through of it all tonight

Tests I'd like to see:
How high can the MAG II and a single and dual (11w?) DDCs push water in a tube?

I'm planing to build a projector (lumenlab.com rocks btw) and I'm thinking about watercooling it just because I can. My first thought when I saw the MAG II beside the zippo was that it would be perfect, but if it performs poorly I might just have to get a fifth DDC... :p

This does look disappointing... :(

But did we expect it to give the $100 DDC a run for its money?

Actually it does not look bad. That test shows nothing but unrestricted flow. My hydor l30 could push about 2feet from the tubing but it didn't stop it from being a bad performer.

Simply put different pumps handle the pressure differently from others.

A real test would be actually comparing the two pumps side by side with the same rad and blocks. Then actually measuring temps and flow through the entire loop.

I'm looking into buying one of these or maybe two of them.

If possible can you hook the pump up to a fan controller and tell me if the pump can be slowed down or not? And if possible can you give me a stall voltage?

And the noise of the pump?

nice work there buddy

If that was, in any way, a valid enough form of testing to constitute a review, then my name is mud. :rolleyes:

If you wanted to do some primitive testing (i.e. lacking any real instrumentation), then I would suggest the following:

- Simple flow tests: Timed filling of a graduated container (at least 2 gallon or so) at free flow. Not sucking water up hill, as you were doing, but by using a large-capacity (or continually fillable) reservoir plumbed directly to the pump's inlet. This test can also be expanded upon by adding common cooling loop components to the pump's outlet (before the graduated container).

- Simple head pressure tests: Using the same reservoir setup as before, connect a long, vertical run of tubing (at least 10' for the MAGII and at least 20'+ for the DDC+) to the pump's outlet and measure the height of the water column which results. Now, the reservoir (depending on how it's set up) will affect the max sustainable head pressure result... However, it would be reasonably easy to find and subtract from your measurement, if it's substantial. If you wanted to throw together a crude PQ curve, then you could cut a foot or so from the end of the vertical tubing run (allowing water to flow out of the top) and then set up some sort of collection pan to catch the water and drain it into the previously used large, graduated container. Do multiple timed runs, then cut another foot off and repeat.

lol petra's probably right...

He definately is.

basic head/pressure test:

filled a bucket with water, pump sucks water from bucket, and pumps it up a vertical length of 1/2" ID tubing, marked in 1 foot increments up to 6ft (at the ceiling)




Danger Den Mag II LE - just over 4ft of head:

http://www.fragtek.com/sneaky/pumps/mag_head.jpg




Laing DDC Rev. 2 w/ Alphacool Acrylic Top (inlet drilled out to 1/2") - pushes water up to 6ft and over the hook that the tubing was on, and flowed back down into the bucket without even thinking twice:

http://www.fragtek.com/sneaky/pumps/ddc_head.jpg




Laing DDC Rev. 2 w/ Alphacool Acrylic Top (inlet drilled out to 1/2") - pushes water up to 6ft and over the hook that the tubing was on, and flowed back down into the bucket again, but had alot more pressure at the end of the tubing when i tried to stop the flow by putting my thumb over the end of the tubing:

http://www.fragtek.com/sneaky/pumps/2ddc_head.jpg








***IMPORTANT*** During the head/pressure test with the Mag II LE, I turned the PSU on once, the pump sounded like it was on, but pumped no water (yes it was primed) - I had to hit the pump against my hand in order for it to start - I had to do this anytime I let the pump sit turned off for more than 30 seconds, so it seems that the Mag II LE has some serious reliability/start issues

I will be calling Danger Den tomorrow to explain the issues with the pump not starting, and to possibly return it, as I am not at all satisfied with its performance in comparison to other pumps that I've owned... but I guess you get what you pay for (not that I ever doubted this)

go to backyard and climb up to 20ft tree to hang your tube and try to test your DDC again :)

lol, i was gonna go outside, cause i knew the DDC would rape my 6ft head pressure setup, but it was dark out


tomorrow or sometime i might just droop some tubing out my second story window and see how the DDC's handle that ;)

I'm starting to dislike DangerDen. They claim 8ft head, results show 4. If thats not lying, I dont know what is. Oh wait, I know - its marketing!

Quote:

---

Originally Posted by Sneaky

lol, i was gonna go outside, cause i knew the DDC would rape my 6ft head pressure setup, but it was dark out


tomorrow or sometime i might just droop some tubing out my second story window and see how the DDC's handle that ;)

---

well you could use and y splitter an 2 tubes, then you get 12feet instead of 6. but thats probably not long enough anyways for an ddc....

Quote:

---

Originally Posted by eXa

well you could use and y splitter an 2 tubes, then you get 12feet instead of 6. but thats probably not long enough anyways for an ddc....

---

doesent quite work like that...

why not? with twice the wolume in the tube youll get half the height. should be the same as using a larger tube... it wont be any flow restriction when the water aint moving!

I think youll find head isnt dependant upon the ID of the tube used...

A larger ID tube will hold more weight of water per unit length .. Therefore it would be dependant on ID ?

Can you explain why it wouldn't be ?

Its not. If you put a large column of liquid above it itll longer to equalize because there's more space for the pump to fill, but the head of a pump is basically v^2/2g, where v is the velocity of the liquid at the periphery of the impeller, and g is simply the acceleration due to gravity. Use feets per second for both and you get the pump's head in feet.

Actual head is modified by impeller design, but dont become confused by inaccurate analogies with solids.

Take a tank of water, and put a pressure sensor at the bottom of it. The only water putting pressure on it will be that which is part of the column of water directly above it. A bigger tank will weigh more, but 2 tanks of different sizes with the same height column of water will have the same pressure for any one point on the bottom.

A pump with a hose is no different than a pump connected to a water tank.

So long as the ID of the tube isnt smaller than the smallest ID in the volute, head doesent change.

oh i see...

Nice work Sneaky! :)
Seems that this DD pump is worse than bigwater's one... :p:

I agree with sdkevin about the tree climbing though. :D

:eh: Okay... yeeaahhh..... something didn't look right about your results, for whatever reason... so I decided to make a few "modifications" to one end of my test-rig-in-progress and do it myself. Anyway, here goes..

The Data:

MAGII LE: (at 11.98V) ~3psig (6.92' -- 2.11m)
3psig ~= 83.03" water column at 4 deg. C (using the same conversion that Lee at SystemCooling does--1.0" = 0.03613psig)

CSP-MAG: (at 11.98V) ~2.8psig (6.46' -- 1.97m)
2.8psig ~= 77.5" water column at 4 deg. C

The Shi:banana::banana:y Photos:

MAGII LE:
http://site.petrastechshop.com/image...TESTMAG_II.jpg

CSP-MAG:
http://site.petrastechshop.com/image...ESTCSP_MAG.jpg

...and, just so that everyone knows, the MAGII LE turned out to be quieter than the CSP-MAG, the D5, and the 50Z... but, as far as the latter two go, that's no great accomplishment. :rolleyes:

None of this says anything about reliability, though... that takes a very large sample to determine (i.e. numerous pumps).

If anyone doesn't understand the setup, take a look at the shutoff valve position or ask me tomorrow.

Oh, one more thing, C-Systems usually does their testing at 12.5V... which is why their specs are a little fishy.

its water supply going to pump going to T-1 where 1 branch goes to that sensor above the mobo going to another branch on T-2 then back to Water supply. now the other T-1 branch goes to a smaller T-3 with a Pressure gauge on 1 branch then the other branch going to the T-2 branch before the Water Supply (a larger picture would be nice)

now using C clamp, clamping the smaller tube on T-3 branch so we could messure how much pressure is produced (close up on pressure gauge is required)...

how about testing a ddc ultra if possible request...

If the setup is still standing, please measure the power consumption.

deffiently seems like a nice little pump for the money. ive been considering picking two up to replace my mcp600 which is getting a bit worn out.

is that nearly 7 feet? thats pretty darn good for something thats smaller than my mouse...

So whats the consesus? Is this pump could for a heatercore, White water CPU block, and Maze 4 GPU block?

Its a little on the weak side (50Z and D5 have 10 feet, DDC-Pro 12-ish, Ultra 20-ish) for a double-rad, 2-block loop, imo. But its cheap as chips and very small, and quiet.

Depends on what you want from your loop really.

I want to get quality parts right now so I don't have to constantly upgrade, but my budget is also important.

How much of a difference do you guys think a Mag II le vs. D5 would make?

Quote:

---

Originally Posted by Rezag1000

I want to get quality parts right now so I don't have to constantly upgrade, but my budget is also important.

How much of a difference do you guys think a Mag II le vs. D5 would make?

---

a pretty significant one - especially if you're using a block with a jet-impengement based design

Quote:

---

Originally Posted by WeStSiDePLaYa

deffiently seems like a nice little pump for the money. ive been considering picking two up to replace my mcp600 which is getting a bit worn out.

---

What do you guys think of running with pump with a double 120 rad, mcw-30 and DDMaze 4? No CPU in the loop (going under phase) I like the fact that its small and quiet and also has a rpm sensor to plug into the mobo for safety. Will it be enough?

No way.

Course it'll be enough... stronger pumps would produce lower temps is all...
The question to be asked is, would you get any better temps than with the stock aircoolers... but certainly the MAGII-LE would be enough to adequately cool them to a point of equilibirum within the cards' operating specs. Probably would still beat air, just not allow as much of an OC as the same setup with a stronger pump.

probably with 0% OC i think...
used to run Tt Aquarius3, results almost same as aircooling, not even 5% OC yet...

woot, do you really belive that the difference between mag2 and ddc2 would be that much tempwise with only maze4 gpu and mcw30 in the loop! i dont belive that at all, i think it will beat aircoling with ease and i dont think you would be able to oc that much more with an ddc2...

But if you could proove me wrong id happily get down on my knees and say im sorry!

id personally just modify the Swiftech MCP350.

Click Here

Quote:

---

Originally Posted by eXa

woot, do you really belive that the difference between mag2 and ddc2 would be that much tempwise with only maze4 gpu and mcw30 in the loop! i dont belive that at all, i think it will beat aircoling with ease and i dont think you would be able to oc that much more with an ddc2...

But if you could proove me wrong id happily get down on my knees and say im sorry!

---

Although i can't be positive in my opinion, cause i haven't got a personal experience. I too believe that a DDC won't yield in significantly lower temps and higher oc's. My :2cents:

BTW i will also get down on my knees... :)

Quote:

---

Originally Posted by eXa

woot, do you really belive that the difference between mag2 and ddc2 would be that much tempwise with only maze4 gpu and mcw30 in the loop! i dont belive that at all, i think it will beat aircoling with ease and i dont think you would be able to oc that much more with an ddc2...

---

I couldn't agree more! My experience with watercooling (5 years) shows that a 1c difference in gpu temps only results in about 1-2 mhz increase in o/c (using atitool)

The difference between a mag2 and a ddc2 (or d5) should amount to less than 2c in most cases. Considering the noise, price, and size, I'd choose a mag2 over a d5 in a heartbeat.

i own a csp mag (the old square style running 24/7 have owned it about 9 months it cools a prescott 3.4 @ 4.1 with 1.55 volts and a 1900xtx @750/840
and has no problem keeping things cool and as far as noise my enermax liberty psu is the only part of my pc you can hear( except for drives when active). it is rare for my temp to get to 50c. i can post screens if needed.
i think the new style should do even better and wouldn't hesitate to get one.
by the way mine seems to have much more pressure than the one shown pumping into the sink??? maybe i just got a good one.

Hello.
Could someone please upload the pics again? :D PLEASE!!!

Regards Mads

Quote:

---

Originally Posted by thecoldanddark

can't see any pictures... anyone else have this problem?

---

I do.

I also don't see the pics.

Think i will try out this pump(if Quoc havent already packed my rather large order), helluwa lot better than my L20 atleast, Then i can switch to ddc-2 petra later...(with storm)

I have orderet it and looking forward to get my water running! So I can OC my Conroe more, im at 3100-3200mhz @ standart voltage and boxcooler.
But if the pics. could be uploaded again I would be very happy! Together with others im sure :)

Mads

Quote:

---

None of this says anything about reliability, though... that takes a very large sample to determine (i.e. numerous pumps).

---

Taking notes... so far from 12 sold out of last batch received, 4 have been returned for same fault - no pumping, loud whine being emitted.

Right away or dying after short usage?

Bit o' both. 2 were DoA. T'other 2 died within 48hrs.

Marci, are you talking about the Danger Den pump? Could it die very fast? And what are the odds for a working one?

I'm talking about the pump that shares it's name with this threads' title. Odds are shown above based on the small cross-section I've come across.

Okay.. I hope my one will live long time:D
It should be here tomorrow.

No DD pump i've seen of recent is lasting as long as it should!!! Shocking

Nothing new... s'just following the usual trend of high failures in pumps manufactured by C-Systems...

Quote:

---

Originally Posted by Sneaky

a pretty significant one - especially if you're using a block with a jet-impengement based design

---

I don't think so. Pump power is only important if you want the max. Performance from your system. A D5 has nearly double the flowrate in a normal loop, but that doesn't helpscooling that much. A old waterblock like a maze4 with a strong pump like a D5 performs worse that a modern waterblock like a nexxos xp with a small pump like the Mag 2 LE. If you have blocks without any turbulence increasing structure, jets or pins, you actually won't be able to spot the difference of the pumps. With better scaling waterblocks like Nexxos, Storm, etc. you will notice a few minor degrees.

http://www.procooling.com/reviews/as...cpu-air_sm.jpg Expect the Mag2 LE to lie right between the Eheim 1048 and 1250.
As the D5 is only slightly better than the aquaextreme 50Z. So with a Storm G5 and 71Watt of heat output, you could expect about 1°K difference between these Pumps. Well since cooling is linear, you could say that even with a modern overclocked 140Watts CPU and a extremely well scaling block, you won't get much more than 2°K difference.

And Marci, don't tell me that 2°K harms Overclocking the sightest bit, i don't believe that.

Biggest problem with this pump is in my opinion, that it isn't a proven design like DDC or eheim pumps. Eheims design is tens of years old and proved to be longliving while laing transfered their firing pump experiences into a well thought design. Compared to that these brushless dc controller pumps which adapt the electronics of common Fans are not proven in any way. I expect them to have a failure rate quite ahead of the other proven pumps. But when that has sorted out they are ceratinly a very good cheapo alternative.

Well, one thing you didn't take into consideration, when talking about the differences, is radiator performance. If you take a look at the tests that were done there, radiators do scale just like modern impingment blocks with higher flow rates, which then would double the performance difference between the very weak and the very strong pump. So then, if you continue to optimize every little bit of your cooling setup, you'd have more than a considerable performance difference :)

I don't think that Radiators scale as well with flowrates as jet impigment blocks, at least not in a normal setup with a heatload of 100-200watts. Can you give me a link or so with the numbers you mentioned?
of course, optimizing the loop can double or even triple the flowrate. 1/2" tubing instead of standard aquacomputer tubing is a helluva difference! I tink thats most important in a loop, more important than a high priced pump.
Always try to get the best out of the components for little money. Compared to Eheim Pumps, the Mag 2 LE ist a powerful pump. Just because we don't have anymore hydor L30s or Eheim 1250s for the highend but nearly double as powerful Laing pumps, doesn't mean that 2,1m head is not enough anymore. 40$ for a silent pump with 2,1m head is pretty good for entry level and mainstream watercooling.
Its just the range i would choose if i had to build a mainstream complete kit.
Its just that realiability thing that keeps me off from importing this pump.

Quote:

---

Originally Posted by Radical_53

Well, one thing you didn't take into consideration, when talking about the differences, is radiator performance. If you take a look at the tests that were done there, radiators do scale just like modern impingment blocks with higher flow rates, which then would double the performance difference between the very weak and the very strong pump. So then, if you continue to optimize every little bit of your cooling setup, you'd have more than a considerable performance difference :)

---

According to the most recent data from Bill Adams the opposite is true. For instance, using panaflo (medium) fans at 5 volts and a load of 150watts, the output temp of the CR-22T at .5gpm is 2.2C cooler than at 2gpm (10c vs 12.2c). Even at high airflow rates (panaflo@12volts) the output temp at .5gpm is still slightly lower than at 2 gpm(3.9c vs 4.1c).

@flb: Well, I took the numbers from Thermochill's HE series into consideration here.

http://thermal-management-testing.com/Thermo10.gif

Found here: http://www.over-clock.com/ivb/index.php?showtopic=14805

I took that as a baseline. If BillA now comes out with just the opposite, well... makes me curious to be honest! I would like to try that out myself somehow. Do you have a link of where you got that data from?

EDIT: Doesn't really make sense. Both HWLabs rads we had were much worse in the x-flow version. So if low flow was better for rads with low flow fans, that should have been seen just there... *normally*

@davidzo: Of course, you're right. No one needs the most powerful pump unless the rest of the system is as far optimized as possible. I'm sure there are quite many people who think that with a DDC type pump, they can do any cruelty to their loop and still end up with good performance... as that's definitely not true :)

Quote:

---

Originally Posted by Radical_53

Do you have a link of where you got that data from?

---

Here's the link.
http://translate.google.com/translat...language_tools

I don't think higher turbulence, due to increased flow rate, effects rads nearly as much as water blocks because the heat load is being spread over such a large area. In fact if we look at waterblocks like the mcw5000, we already see a major reduction in improvement from increased flowrate.

I think temperature differential, evenness of waterflow distribution between the parallel channels, airflow speed, and evenness of airflow are some of the variables that are effecting the performance of the rads in Bill's test.

It may be that increasing waterflow causes an imbalance between parallel channels. Perhaps, when this is paired with uneven airflow, we see a reduction in performance.

The bottom line is I think there are a few possiblities that no one has really tested for (that I know of) that may explain Bill's results.

Or, it may be that his data is all crap. Until any data is verified by independent sources, I always take it with a grain of salt.

Well, I'm sure that his data ain't crap. I think we can take that as granted, regarding his previous work.
I'll take a look at his setup, maybe there is a way that I can replicate some of it with my test stuff. I'm really curious how that came together, and if so: Is there a way to prevent it.
Would be quite stupid cranking up block performance with higher flow just to loose rad performance with this accordingly, as then the overall performance woulnd't improve at all, rendering it meaningless :(

there is no way that flowrate can have a negative impact on rad performance. Only increased pump heat dissipation could have something to do with these results, so in a loop one should considerate that as well.
I think that the HE numbers you showed here radical, aren't the whole truth too... From what i can estimate, there will be no turbulence in rad chambers due to high cross section and slowly water movement. Due to laminar boundary layer effect, there will be no measurable heat transfer improvement at all in a rad at those flowrates we have in a watercooling loop. Thats what i guess and since other predicates differ a small amount to both sides this must be near the truth.

Whoa, its noisy! gonna pack it in something, but maybe its a bit less noisy when it gets some more restriction than a bit of tubing....

Quote:

---

Originally Posted by davidzo

there is no way that flowrate can have a negative impact on rad performance. Only increased pump heat dissipation could have something to do with these results, so in a loop one should considerate that as well.
...

---

The Cooling-Masters review is excellent as usual. Looking at the negative impact of increased flowrate at lor airflow I am beginning to wonder how flowrate was regulated. I see a Iwaki next to the testsetup, did they undervolt it? If they used a adjustable valve to regulate, I wonder how much the pump power draw changes between low flow-max flow.

Oh no, got som rattling in it. sounds a bit like air bibles but not quite. it sounds more like impeller slightly touching something. If i point the inlet down it goes away...

I took one look at it, saw how cheaply built it is (Designed more for looks) and decided not to even waste time thinking about it. Now seing this thread I trust my instincts even more now!

I opened it up and smoothened the outer edges of the impeller slightly, not deforming them just removing som bits that stood out. and i think that did it!

Edit: hmm, still a little rattling, seemes like it comes back after running for a while. If i squeeze it a bit the rattle seems to dissapear, dont know if its from the preassure or that it vibrates less...

The mag2 is useless. Aint going to use it, its really loud and it rattles

guess ill order a ddc-2 in january then...

Quote:

---

Originally Posted by FLB

According to the most recent data from Bill Adams the opposite is true. For instance, using panaflo (medium) fans at 5 volts and a load of 150watts, the output temp of the CR-22T at .5gpm is 2.2C cooler than at 2gpm (10c vs 12.2c). Even at high airflow rates (panaflo@12volts) the output temp at .5gpm is still slightly lower than at 2 gpm(3.9c vs 4.1c).

---

I dont know where you are seeing that at higher airflow rates the lower flow outperforms the higher flow. I see at low airflow that this is the case and if you read that entire review it explains it quite well. With a set amount of airflow you can pump the water through the RAD fast enough that it does not have enough time to cool properly. Increasing the airflow aleviates this problem as can been seen in the graph you showed.

So yes that is something to consider if you are using low CFM fans for a silent computer. You can pump the water too fast...

Quote:

---

Originally Posted by eXa

The mag2 is useless. Aint going to use it, its really loud and it rattles

guess ill order a ddc-2 in january then...

---

same problem that my mag1 has recently developed, and I'm with you the ddc-2 is the way to go

Quote:

---

Originally Posted by Bun-Bun

I dont know where you are seeing that at higher airflow rates the lower flow outperforms the higher flow.

---

Bill's data is based on input temp while I'm using output temp. At .5gpm a 150watt load will drop 1.14c from the input temp, while at 2gpm it will only drop .28C

Quote:

---

Originally Posted by FLB

Bill's data is based on input temp while I'm using output temp. At .5gpm a 150watt load will drop 1.14c from the input temp, while at 2gpm it will only drop .28C

---

And how do you derive that from the graph you attached?

Quote:

---

Originally Posted by Bun-Bun

And how do you derive that from the graph you attached?

---

The temp drop is based on a proven scientific formula

Quote:

---

Originally Posted by FLB

The temp drop is based on a proven scientific formula

---

Which is?

EDIT: I went back and re-read the review again. Bill fixed the input temperature of the water at the inlet thus it makes perfect sense that the DeltaT is smaller the more flow you put through. The faster you move the fluid the less time it has to cool off, however it cools more water at the same time thus disapating more heat. Therefore in an actual loop the overall temperature of the water would be lower becasue more heat would be disapated and there is not a fixed temp going into the RAD.

EDIT: Flow is gph not gpm.

Quote:

---

Originally Posted by FLB

Flow is gpm.

---

lol you posted just as I edited my last post. Yeah I know that equation. Been a few months since I graduated I couldn't remember it off the top of my head...

Quote:

---

Originally Posted by davidzo

there is no way that flowrate can have a negative impact on rad performance.

From what i can estimate, there will be no turbulence in rad chambers due to high cross section and slowly water movement. Due to laminar boundary layer effect, there will be no measurable heat transfer improvement at all in a rad at those flowrates we have in a watercooling loop.

---

Well, I don’t agree with your first statement, but I believe your second statement is dead on. I’m coming up with Reynolds numbers around 930-935 thereabouts for 2GPM. It seems like the new generation of radiators are operating well within the laminar flow range.

There’s one thing that people never seem to pick up on when discussing the type of charts that were recently posted. When flow rate is increased, but the delta between inlet water temp and air temp remains the same, you have effectively increased the amount of energy being delivered to the radiator. It is exactly the same as taking your CPU from idle to full load. The radiator dissipates more heat at higher flows simply because it’s getting more heat!

As far as I am aware, no one tests radiators using a test rig that simulates a PC water-cooling loop. That is, a constant amount of energy transferred to the water regardless of flow rate and water temperature (within reason, of course.) THAT is what happens in our loops.

For years I’ve said that when flow increases, the radiator suffers in performance due to the lower deltaT. But this isn’t necessarily bad if the rad has excess cooling capacity to cover for the loss.