Using 2 D5 pumps in line with each other?

[newls1]

Would using 2 D5 pumps in line with each other (1 right after the other) and each pump feeding from its own res, would this be a good setup for 24/7 use reliability? Just incase 1 pump fails, the other one is there to back it up?

[jmilcher]

Would using 2 D5 pumps in line with each other (1 right after the other) and each pump feeding from its own res, would this be a good setup for 24/7 use reliability? Just incase 1 pump fails, the other one is there to back it up?

Not useful at all really. You wouldnt gain anymore pressure, nor would you gain head ft, the only thing it might do besides cause headaches and a loss of $75 more dollars is yes, in the event of a tragedy (if plumbed correctly) it would keep the system running..

thats why flow indicators are only $7 :woot:

[serialk11r]

No, it doubles the maximum pressure, which means that restriction won't hurt the pump's performance as much (but its not a huge gain). Some people buy a second pump for safety, but to me it looks like a waste of money.

[ziddey]

So if I understand you correctly, you're going to have res->pump->res->pump?
Just can the second res. In that event, especially if the second pump fails, you'll be positive pressuring the second res.

[mcoffey]

Not useful at all really. You wouldnt gain anymore pressure, nor would you gain head ft, the only thing it might do besides cause headaches and a loss of $75 more dollars is yes, in the event of a tragedy (if plumbed correctly) it would keep the system running..

thats why flow indicators are only $7 :woot:

what? no added head or flow. where did you come up with that?

andyc

[CedricFP]

Using two pumps in series will help your flow. But I don't really see that anything needs 2 D5's unless you've got 2 storms in your loop.

[Senater_Cache]

from what I have heard, in series you will gain about 30% head pressure and in parallel you will gain about 30% flowrate. In both cases you will gain ~100% of the heat dump

[serialk11r]

from what I have heard, in series you will gain about 30% head pressure and in parallel you will gain about 30% flowrate. In both cases you will gain ~100% of the heat dump

Well actually with series you should gain 100% max pressure but that's only MAX pressure. So restriction just has a smaller effect on the pump, so flow increase is small. In parallel, you should gain 100% max flow but restriction goes up by a LOT so you'll only get a little more flow as well.
With pumps that aren't too big heat dump shouldn't be a big issue unless you really have inadequate rad needs.
Sometime I'll have to do a test to see how much heat actually makes it into the loop... it'll require lots of math to make up for my lack of fancy equipment.

[MaxxxRacer]

Two D5's in series just isnt a good idea due to the heat dump unless you have a really powerful radiator(s) like nikhsub1 does.

If you want to go with dual pumps I suggest you used the AQZ50Z. It has a 7 or 8 watt heat dump IIRC so it is MUCH more suited to the task. its performance is only slightly less than the D5. Dual 50Z's equates to 1 D5 in heat dump but is nearly equal to two D5's in the performance sector.

To plumb it you do the following
Res -> Pump1 -> Pump2 -> Rad -> Waterblocks.

You can swap the rad and waterblocks if u want. It doesnt make much difference, especially at higher flowrates, for reasons I wont get into.

From what I have seen the average flow increase is between 30-40% depending on the pumps and how restrictive the system is.

[mcoffey]

Two D5's in series just isnt a good idea due to the heat dump unless you have a really powerful radiator(s) like nikhsub1 does.

If you want to go with dual pumps I suggest you used the AQZ50Z. It has a 7 or 8 watt heat dump IIRC so it is MUCH more suited to the task. its performance is only slightly less than the D5. Dual 50Z's equates to 1 D5 in heat dump but is nearly equal to two D5's in the performance sector.

To plumb it you do the following
Res -> Pump1 -> Pump2 -> Rad -> Waterblocks.

You can swap the rad and waterblocks if u want. It doesnt make much difference, especially at higher flowrates, for reasons I wont get into.

From what I have seen the average flow increase is between 30-40% depending on the pumps and how restrictive the system is.


I soon as I saw the post that said there was no gain, I went and read your sticky on that very topic. Thx for the clarification.

andyc

[newls1]

So if I understand you correctly, you're going to have res->pump->res->pump?
Just can the second res. In that event, especially if the second pump fails, you'll be positive pressuring the second res.

Basically what im looking for is MAX stability in case of 1 pump failure. Im not looking for added performance by any means. The reason for the second RES is so when I first start the loop up, the second pump wont cavitate (or run dry) as it will suck fluid from the RES (Im sure the pump would be thankful) Here is a crappy MS Paint diagram that I just did so you can see what im talking about. http://img249.imageshack.us/my.php?image=duald5pumpss0.jpg Either way, I that this was a good idea

[ziddey]

that's an awful lot of extra work to go through and not a good idea.

as with filling the system anyway, you shouldn't be running the full system anyway. just don't plug in the second pump until after you've finished filling. run only the first pump while filling. then you're good to go.

[serialk11r]

Wouldn't getting an ebayed iwaki be much more effective? I can imagine 2 D5s taking up a good deal of space. Iwakis are very very reliable I hear, and they are powerful (possibly more efficient?).

[jmilcher]

No, it doubles the maximum pressure, which means that restriction won't hurt the pump's performance as much (but its not a huge gain). Some people buy a second pump for safety, but to me it looks like a waste of money.


that was pretty much the point of my post... :)

Like batteries in series

[mcoffey]

you might want to use the EK multi res for the res between the pumps, that way you wouldn't need the tee fitting. They come in 3 different sizes, some of the best built and looking res's around IMO.

http://www.petrastechshop.com/ekmure.html

Just a thought,

andyc

[IanY]

Thankfully I brought my notebook on vacation :)

What are you guys? Crazy? Two, three pumps in series works incredibly well. Have you guys tried it? I'm going to do four pumps soon..

[mcoffey]

that was pretty much the point of my post... :)

Like batteries in series

you're losing me. in your previous post you said you wouldn't gain anymore presure. in this post you seem to be agreeing, and something about batteries in series. So are you saying that two batteries in series doesn't create more power like your saying two pump in series doesn't increase presure, or what?

Like Maxx and Ian says, of course it increases flow and presure, how in the world could you think it doesn't. Just like hooking 2 batteries in series increases voltage.

andyc

[mcoffey]

Wouldn't getting an ebayed iwaki be much more effective? I can imagine 2 D5s taking up a good deal of space. Iwakis are very very reliable I hear, and they are powerful (possibly more efficient?).

2 D-5's taking up less space than a Iwaki. Do you realize how big those pumps are? Certainly aren't know as space savers, especialy considering the seperate PS if you run the RD-30. Incredibly good pumps, but space savers they aren't.

you're confusing me again,

andyc

[IanY]

Like I have said before.. if you cannot afford to buy a second pump, and if it does not make monetary sense to you, don't go accuse it as not being effective, or accuse it of not working. It may not be cost effective to you, but so what? 0.1 degree C lower is 0.1 degree lower. Its just like saying a car is not faster if it only gains 0.1 second to 60 mph.

Forget that, daisy chain Iwakis...

[serialk11r]

Like I have said before.. if you cannot afford to buy a second pump, and if it does not make monetary sense to you, don't go accuse it as not being effective, or accuse it of not working. It may not be cost effective to you, but so what? 0.1 degree C lower is 0.1 degree lower. Its just like saying a car is not faster if it only gains 0.1 second to 60 mph.

Forget that, daisy chain Iwakis...

Its not monetarily effective ;) Everyone we gotta correct our phrasing.

[IanY]

Ohh.. a second pump is expensive, and therefore is not cost effective... AND YET... if I were to look inside all of your computers, I bet I'll be able to dump out $200 to $300 of absolute nonsense... like metal barbs, case windows, cold cathodes, LCDs, custom grills, UV lights, bling plexi water block tops, "good looking" delrin tops, zip ties, sleeving, heatshrink, UV dye, neon fluid additive ..... stuff that do absolutely nothing to aid system performance. Dump out that garbage and you will buy yourself two more pumps and two more radiators.

Ya.. bling is *absolutely* necessary... that garbage is monetarily effective!!

[[cTx] Nooc]

Garbage? I would rather have all that "garbage" in my system than buy a second pump

[serialk11r]

Ohh.. a second pump is expensive, and therefore is not cost effective... AND YET... if I were to look inside all of your computers, I bet I'll be able to dump out $200 to $300 of absolute nonsense... like metal barbs, case windows, cold cathodes, LCDs, custom grills, UV lights, bling plexi water block tops, "good looking" delrin tops, zip ties, sleeving, heatshrink, UV dye, neon fluid additive ..... stuff that do absolutely nothing to aid system performance. Dump out that garbage and you will buy yourself two more pumps and two more radiators.

Ya.. bling is *absolutely* necessary... that garbage is monetarily effective!!

Really? Lets count up the money I spent on bling:
LEDs and CCFLs $10
Zip ties $1
Dye (used when blocks are done) $5
blingy fans $25 (this one doesn't count lol, I needed something other than thermaltakes!)
Blingy controller (rheobus lol) $10 (yup, 10 bucks!)
Fancy UV green IDE cable for my drive $2
Can $53 buy a new pump?

[Duh]

Not useful at all really. You wouldnt gain anymore pressure, nor would you gain head ft, the only thing it might do besides cause headaches and a loss of $75 more dollars is yes, in the event of a tragedy (if plumbed correctly) it would keep the system running..

thats why flow indicators are only $7 :woot:


that was pretty much the point of my post... :)

Like batteries in series

dude did you remember to have the pill this mornin´? jk :D

[Omastar]

Ohh.. a second pump is expensive, and therefore is not cost effective... AND YET... if I were to look inside all of your computers, I bet I'll be able to dump out $200 to $300 of absolute nonsense... like metal barbs, case windows, cold cathodes, LCDs, custom grills, UV lights, bling plexi water block tops, "good looking" delrin tops, zip ties, sleeving, heatshrink, UV dye, neon fluid additive ..... stuff that do absolutely nothing to aid system performance. Dump out that garbage and you will buy yourself two more pumps and two more radiators.

Ya.. bling is *absolutely* necessary... that garbage is monetarily effective!!

I'm gonna agree here, but only to an extent. The difference between a 'plain' case with no lighting, stupid dyes that do nothing, UV reactive cable sleeving, etc. and a case that's so gaudy, you'd think Freddie Mercury designed it, is vast. I agree that performance is of utmost importance, far beyond aesthetics that usually look like hell because they're poorly implemented or over implemented. Sure, people have the right to have a Koolance kit and 12 CCFLs and whatever other BS they want, but their free will is limiting their performance, by way of their poor prioritizing.

However, and this is a BIG exception, when tastefully done, lighting and all the other 'gimmicks' can actually look really nice. I know everyone has their own tastes, but when you see a really impeccable build, there's no denying the effort that went into it, regardless of your personal feelings.

[Cronos]

Two pumps in series can be especially useful when undervolted.
You'll get lower noise and higher reliability at the same level of performance.

[newls1]

Two pumps in series can be especially useful when undervolted.
You'll get lower noise and higher reliability at the same level of performance.

Thank you. People, can we please stay on topic in my post;) My question was not about "BLING" factor, or how it is a waist of money, Im simply looking for a way to be safe if 1 pump fails as this PC will be on for 24/7 use, 365 days a year. Possible will be a folding PC:confused:

[Circaflex]

Im in the same situation, Im going to be running Res>Pump>Pump>PA120.3>CPU>Chipset>PA160.1>GPU

Thinking between 2 d5s or 2 AQZ50Zs right now im leaning more towards the AQZ50Z from what max said on the first page of this thread.

[Cronos]

It is desirable to separate pumps in the loop as much as possible for better performance.

[IanY]

Thank you. People, can we please stay on topic in my post;) My question was not about "BLING" factor, or how it is a waist of money, Im simply looking for a way to be safe if 1 pump fails as this PC will be on for 24/7 use, 365 days a year. Possible will be a folding PC:confused:


The answer is yes. I have three loops, each with two pumps for redundancy. It provides a sensible safety net. The machine has been folding non-stop (well, to the extent possible) for about three months now. See the stats below.. I've climbed from over 280th place to where I am now. I will cross 100K points after only three months.

[mcoffey]

Thank you. People, can we please stay on topic in my post;) My question was not about "BLING" factor, or how it is a waist of money, Im simply looking for a way to be safe if 1 pump fails as this PC will be on for 24/7 use, 365 days a year. Possible will be a folding PC:confused:

ok,

well you have your answer then. I'll will help with redundancy and better flow.

gl,

andyc

[sbinh]

Would it be efficient if having 2 (DD-Black-Ice) rads:

Res -> Pump -> Rad1 -> Pump -> Rad2 -> CPU -> Chipset -> back to Res ?

[Cronos]

Pump1->Rad1->Rad2->Res->Pump2->CPU->Chipset->Pump1

This has two advantages
1) Better pump separation
2) Your res will always hold the coldest water in the loop.

[Lekko]

Why not split the tubing with a Y after the reservoir, send it to the two pumps, then feed them both back into the loop with a second Y afterwards. Then just set it up so that if your flowmeter starts reading too low, it switches power to the second pump.

_______-pump-
Res -<_________>-CPU/whatever...
_______-pump-

the underscores should be blank, but it deleted the spaces.

[fscussel]

I don't think D5 will fail in the first 3-4 years (it's MTBF is 100.000 hours), so it's not necessary to have 2 of them. You are already buying a great pump so you can rely in it. Just keep you processor Temperature monitor turned on (it's CPU TM function in C2D / Speedstep) so the system will shut down if the pump stops.