Question on Maze-2 and pumps

[Camelot One]

I currently have a Maze-2, and a VIA Aqua 306 (526gph 0ft.) Checking the pump pushing water only through 6ft. of 1/2"ID tubing, the pump output is about 180gph. When I run through the maze-2 and 6ft of tubing (all laying flat) flow drops to 49gph. I've run peroxide through the maze, but didn't get any gunk out of it. Its not very old, so I don't think its clogged, just a very restrictive block from what I've read. I think the pump is partly to blame for the massive drop in flow too.

With an 86W pelt on an AMD 2100 @ 13x145, vcore 1.775, CPU temp stays around 38C. Water temp in the reservoir stays about 35C. (the pump puts off a LOT of heat) Without the pelt, reservoir temps are still as high, but CPU temp goes to 46C. (those are both under full load)

Should I replace the block, or the pump? I'm thinking a higher output pump, not just the rating, but what it actualy pushes, would help. A cooler running pump would make a big difference as well. But am I way off base? I can get the VIA Aqua 1300 for $22, think it would be a worthwhile upgrade? Anyone have that pump and block that can verify their flow rate?

[gmat]

I have a maze 2 with a 1200 l/h pump. I think it's too much for such a restrictive block. When i have the money i'll try a 600 l/h pump to compare.

Given the efficiency of your current pump, i'd say, take a less powerful one !! Any more powerful pump will increase flow marginally while dumping even more heat !

[g.l.amour]

gmat is prolly right, pump is a tad too powerfull maybe. maze2 should be ok performance wise, not a top performer but ok, getting the best block on the block might, in extremis, get you 3-5°C temp advantage. might consider a bigger radiator?

i have about same specs (same WB, 1200l/h pump) and if i crank up the fans to 12V i can get 32°C full load, without pelts. so my guess is get some more surface area on the rad front.

[RoboTech]

Hey Camelot,

As a point of reference, I measured 141 GPH through my Maze 2.1 block (1/2" hose barbs) using an Eheim 1250 and two 18" pieces of 1/2" ID hose. With the Maze3, I get 153 GPH.

In a working system (E-1250, BIE radiator and 6' of 1/2" ID tubing) its more like 110 GPH for the Maze 2.1 and 125 GPH for the Maze3.

[BillA]

you fellows speak of pump "power", and I believe are confusing the issue

you have high volume, low head pumps,
for sure they will be crappy with the high flow resistance of the Maze2

you need, as has been stated over and over, high(er) head, low(er) volume pumps
-> then the pump is able to function as designed, and will also input less heat into the coolant

that such may be more expensive, so ?
perhaps then it is cheaper to change the wb ?

compare the P-Q curves of the Eheim 1250 that RoboTech describes to the VIA Aqua 306
I don't have these curves but can predict the difference

because a pump has high volume at low head does NOT mean that it is "powerful"

[g.l.amour]

so, if i get this right...

situation: it is real hard to find low volume, high head pumps; e.g. eheim doesn't give u alot of choice, with the jump from 1048 to 1250, head increases only slightly, volume doubles.
most commonly found pumps show the same characteristics (via aqua, maxi jet...)

with that variable in the total equation almost fixed (or goto iwaki or little giant, both really expensive and hard to come by in europe); it is almost always cheaper to find a high flow rad (reading bill's graphs right, hopefully) like the serc, or easy to find big momma. + a high flow block, with some performance to match, like the old swiftech with 1/2" barbs , or (not seen it on bill's graphs) new swifty 462 with lets say 1/2 barbs (looking at design, no maze, just a hollow chamber with some turbulance).

from a block's point of vue, it needs good c/w over the whole flow spectrum + a low pressure drop / flow rate. to have a max performing system, taking advantage of the low head, high volume pump.


am i getting this right, or wrong?

[BillA]

take it in a serial manner

if the choice of pumps is limited, pick that which you prefer
if such has a low head -> then ALL components must be selected with THAT parameter foremost
- low restriction wb
- low restriction radiator
- larger dia tubing
- NO fittings of any type

junk the Maze2 - but which of the Maze2 versions do you have ?
(it is a perfectly good wb, just not as good as some others for low head pumps)

do strive for 7lpm
avoid higher volume pumps with the same head capacity, just add heat with such

"from a block's point of view, it needs good c/w over the whole flow spectrum"
no, not really; a given system has only one flow rate

select a wb with good performance at THAT (design ?) flow rate
for example:
at 3lpm and below the Innovatek seems quite the best (of those I've tested)
at 7lpm and above the big Swiftechs seem to be superior (same qualifier)
- noting of course that the differences between the 'good' wbs are not so large

here's the latest on wb head losses (in the endless pursuit of greater accuracy):

should add the 462-UH later this week

[nuclear]

Hello BillA
Simple question about the graph
Where is the pink line ?
It would be a good idea (if possible) to make the little marker dot that makes the line a little bigger for the line below the other, like if there are 2 line one above the other, to at least see the main dot that make the line so we can see that there are multiple block sharing the same line
If it's not easy, maybe just add a little info on the bottom of the graph so we can know.
Thanks
/edit corrected BillA name /edit

[gmat]

Originally posted by unregistered

avoid higher volume pumps with the same head capacity, just add heat with such

That is exactly what i meant by "too powerful", thanx.
Like g.l.amour said high head + low volume pumps are hard to come by out there. I cannot find Via pumps in Europe - and i dont want to shell out $400 for an Iwaki. We have Eheim, Maxijet, Hydor, Project - those are easy to find.
By switching to a 1048 from a 1250 for example, can bring good results in a severely restricted circuit. The 1250 will not produce more flow - just more heat...

[g.l.amour]

well, so i found out the hard way. to make a very long story abit shorter...

knowing is everything.

bill a's work is great, but can't list everything.

i went from hayden,+ generic HC (series) , maze2, 3/8", 1048;
to dual large bigmomma style rads, rest staying the same. i was xpecting some nice temp drops, but didn't get any. i then switched to 1/2", didn't change too much either.

i know from BillA that big momma's don't restrict flow too much, about the hayden and the generic, i can't know nothing.

so i was thinking that i needed more flow, with the dual rads 'n all; got myself an eheim 1250. didn't change anything, xactly the same temps (prolly not, but i ain't using special equipment for testing, so large tolerance). so guessing some of it the hard way (lots of $'s spent), the maze2 was already at its relative peak performance with the 1048 pump; throwing in an 1250 just makes the 1250 run at shitty efficiency. the rads shouldn't be the ones to blame, judging the graphs they don't restrict flow too much.

i feel stoopid for spending all that money on 0 change in performance, don't make the same mistake as me camelot one. look in on your rad performance, if that doesn't restrict flow too much, i guess getting the swiftech block will serve u alot better than getting the larger pump.

anyway, step by step, i feel the mist clearing away, really really really too bad, i had to learn the hard way. sooo thx for making me(us?) understand the fundemantals BillA. sometimes it might seem it is all in vain, well, trust me, it isn't.

[BillA]

Originally posted by gmat
. . . . By switching to a 1048 from a 1250 for example, can bring good results in a severely restricted circuit. The 1250 will not produce more flow - just more heat...

errr, try that again (the first is in contradiction to the second)

nuclear
the graphs are made with Excel, with which I am learning that anything is possible
(except dual scales on both axis)
I have 21" Hitachi monitors, the graphs look fine here, I'll experiment

re the pink line ? - read the whole post (under the graph ?)

g.l.amour
yea, the expression here is "no free lessons"
(you cannot imagine the number of times I have rebuilt my test bench - I mean over, and over, and over; grrrr)

[gmat]

Originally posted by unregistered
errr, try that again (the first is in contradiction to the second)

Huh ? I meant, *from* a 1250, *to* a 1048 - lowering max flow but keeping similar head would help a lot in restrictive conditions.

[g.l.amour]

well BillA in my dutch mother tongue, we have about the same saying;

everybody needs to pay their learning fees one way or the other...

[BillA]

and what is "the egg of Columbus" ?

[g.l.amour]

don't know the historic truth to it all, but the story goes like this.

some rich king once set a quest for someone to find a way to place an egg upright. as u know it will always roll over so the length axis is parallel to the ground.

columbus was said to have found a way to do it, noone thought it was possible. well columbus just took the egg and slightly knocked it in position so the shell broke and made it stay upright.

so the expression goes for anything that seems really complicated and impossible, but can have a really dumbass simple solution.

explaining this, i can only hope i used it in the right context, because i can't recall too well. prolly i meant something in the line of: look a little further than which seems obvious.

anyway, speaking a couple of languages, it is cool to see how the same grandmother wisdoms exist in most different languages, and are applicable even in modern times.

edit: just did a google search on "ei van columbus"

WAT IS EEN EI VAN COLUMBUS?
Het oplossen van een moeilijkheid die heel eenvoudig is als men er maar eenmaal opgekomen is.

translation
WHAT IS AN EGG OF COLUMBUS?
finding the solution to a difficult question, where the solution is real simple once u know it.

there u have it

[BillA]

yea, all western languages seem to have about the same/similar 'set'

would be interesting to know if the 'universal verities' were also similar in Hindu, Chinese, Japanese, etc

anyone know ?

[g.l.amour]

well they should be the same, because the roman languages are as far away from the german languages as russian or chinese is from lets say german. there were alway trade routes from europe to azia from the early roman empire and before. look at the vikings (who were said to have really found the north american continent, that is not sure, but they were on greenland way before civilised europe ever thought it existed) they were harassing a couple of continents from sea, as early as 800 AC.

anyone know this one (laaaargely off topic)?

"don't sell the bear's fur , before having shot it."

haven't heard it out outside of dutch speaking belgium, but seems too universal for anyone not to understand.

[Camelot One]

So, if I understand you all correctly, going with the VIA Aqua 1300 would be the cheapest course of action? (the block is staying where it is for now, I am tired of messing with it this week)
Here is a flow rate chart I found:
http://www.wcaquatics.com/cgi-bin/it...aqua_pump.html
Like I said, I can get the 1300 prebarbed for $22+s/h.

[BillA]

Camelot One
I can't/don't address the cheap aspect
but the higher head pump will serve you best
though you may not see any improvement due to 'other factors'

g.l.amour
here one would say "don't count your chickens before they hatch"

[g.l.amour]

hot damn, thought i found a pretty exclusive one... shows ya

anyway, humans will always be the same, so i guess the same universal values won't change.

[Daleorama]

I think I should point out that the Via Aqua 1300 actually has a lower gph rating than the Via Aqua 306. I know I was unclear about this for a while when I was pump shopping. It just seems to be the way that the company labels their pumps.

Anyway, switching from the 306 to the 1300 would probably help your temps because you would be downgrading the spec. flow rate and downgrading the wattage of the pump as well. Your Via Aqua 306 wants to move the water pretty darn fast, but the backpressure from the restrictive water block hurts flow rate. So now, the energy the pump would have used for it's fast flow rate is wasted. It gets dumped into the water as heat.

I'll bet if you use a manifold and branch off to cool your north bridge, vid cooler, etc. you would actually get better temperatures. The manifold would open up flow a good amount. Then the pump can express more of its power as flow rate, rather than heat. Right now, I have a tiny little 105 gph Via Aqua 360 pump that powers a system with 1/2" tubing and a seperate branch for the NB and vid card. When I branched off the NB and Vid card, I saw an improvement in my cpu temps.