Two Liang D5 Strong (in series) or one Iwaki RD-30?

[avabaska]

Hello!

I have 3 x D5 Strong at 24V (in series). And now, after 1 month, one pump as I assume - dead. I have not analyzed yet what it was but one pump sometimes started to make a very loud crackle. Than even at 15V this hack stopped to work (Aquaero started to buzzing because flow rate dropped below 200 l/h). Than when I restarted all pumps - this thing immediately started to make those sounds).
So now I'm thinking what to do, use only two D5S at 24V (210 l/h) or buy another D5S and use it at 24V (260 l/h as it was) or use three of them at 15V (210 l/h).

Or buy Iwaki RD-30 and use it alone at 24V (I think it will be approximately like two D5S @ 24V). I'm sure that Iwaki RD-30, it's not a mass consumption product like Laing D5(S).

These three pumps are now installed as they were, I just disconnected the middle pump from the PSU and have not disassemble it (don't want to spend a time and just sick of it ). So only when a new pump will arrive, I will do something.

Also I want to write, that pumps are located in the kitchen, in the corner and under the table:



And packed into some towels (not on the photo) so it's a really silent thing even for a room (but better to install that thing in another place ). I'm not sure what will be with RD-30, but I guess equal (?).

Thanks!

But most sad that I played in BF3 and it was a hard round and as I heard a buzz from Aquaero, I had to exit the game and switch my PC off

[earthwormjim]

Do you have a 24V power supply large enough for an RD-30?

An RD-30 puts out almost the same pressure as two D5 strongs do at 24V.

I think it's a bit louder, D5s even at 24V are fairly quiet. You also lose some redundancy with just one pump, but RD series pumps are ridiculously reliable.

Here's some graphs taken from Martin.

D5 Strong @ 24V.


RD-30 @ 24V.




Head pressure is the important thing for pumps in this hobby, not really flow rate capabilities.


Personally I'd keep the two pumps you have now. No real sense in buying a brand new expensive pump for no real gains to be had.

[Church]

.. and I'd remove 9 redundant angle fittings . I can understand use of them here or there in limited amount inside case for neater tubing routing, but in this pump combo? Which looks like as placed somewhere externally / out of sight / with no accent on looks? Simply no need for those angle fittings. And will almost compensate death of 3rd pump and thus lower head pressure

[CrazyNutz]

.. and I'd remove 9 redundant angle fittings . I can understand use of them here or there in limited amount inside case for neater tubing routing, but in this pump combo? Which looks like as placed somewhere externally / out of sight / with no accent on looks? Simply no need for those angle fittings. And will almost compensate death of 3rd pump and thus lower head pressure

^^I totally agree. BTW I see 11 unnecessary angle fittings.

As for pump replacement, I would just do a new D5 Strong.

Regarding being packed in towels, perhaps this would the cause the pumps to overheat? The water should keep the head part cool, however the circuit in the back may not get much contact. How warm are they with the towel over them?

[avabaska]

Hello

Today I guess died another pump. Similar crackle but not so loud. Don't know which one exactly, first or third, because I just disconnected one of two PSUs from the wall. So now I have only 140 l/h (one D5S @ 24 V).

I think these pumps are not designed to work that way (in series + high restriction + 6000 rpm + 24/7). Possible? But it was no more than ~700 hours

So now I'll definitely go RD-30 (ordered).

But what to do know, before I'll get RD-30? I'm afraid that the last pump will die too (who knows in what condition it is now). So I will buy tomorrow 2 x D5 Vario and install them into dual EK top (have it). Also another PSU with 12V, because I'm afraid to use this pumps at 15V (my two PSU are 15-24V).
I have one D5 Vario, but it "died" after one day in series with these three D5S @ 24 V - it just started to work like a D5S - at full RPM, according to sound in EK top (or it possible that the bearing was damaged, or somethin else went wrong, because this Vario has worked for several months in my first loop, only one pump, 12V and low restriction).

First it was like this:



When I tested one, two and three D5S @ 15-24V, from all of them I've heard strange noise. Even recorded it - link. In this video discharge valve is totally closed. But even in normal situation (my testing loop was on the floor) we can hear that kind of sound. I guess it was a sound of impeller that touched interior of EK and Alphpacool tops. So after that I bought an original tops but with G1/4, as you can see on photo.

earthwormjim
>Do you have a 24V power supply large enough for an RD-30?
I have 150W PSU. But Iwaki wrote in manual that RD-30 has starting current - 10A. I think it's too much. So I'm not sure about it.

>I think it's a bit louder, D5s even at 24V are fairly quiet.
Well... Towels

>Here's some graphs taken from Martin.
Thanks. I looked through it many times. Even his Exel configurator (it is an amazing thing).

churchy
>and I'd remove 9 redundant angle fittings
Space is limited Btw, I have about 7 QDC from Koolance (female and male counted together) and total about 25 pcs of 90 dg. fittings. And a filter with mesh width 0,125 mm.



When all three pumps worked and I had a CPU water block Kryos HF (245 l/h), I removed this filter - nothing has changed, maybe + 2-3 l/h. So I think restriction in this loop so high, so we can forget about few 90 dg. fittings.

CrazyNutz
>Regarding being packed in towels, perhaps this would the cause the pumps to overheat?
I am afraid of that. I know only that most heated part of D5S was a place near the retaining ring. In that area temp about + 10-13C than water temp. When I unwound the towels and pointed to this area an IR-GUN, it showed 45C. So I think max temp in that area was up to 50C during all the time.

I don't know what will be with RD-30 but know only that ambient temperature must be not higher than 40C (like in manual). My water temp will be maximum 35C (at normal high loads). I can't use this things without a towels or something. Maybe it is possible to install two DDC in dual top without covering them (I've heard that they not so loud), but I'll do this only after RD-30 death


When I heard an unpleasant sound from another pump, at that moment I have not been played in BF3. So now I know exactly - my first pump was destroyed not by the enemy

[Erklat]

I have one D5 Vario, but it "died" after one day in series with these three D5S @ 24 V - it just started to work like a D5S

You're not supposed to do that. That pump linked in series with three other stronger ones was constantly running out of it's specs.

so we can forget about few 90 dg. fittings.

Your loop is poorly designed and that is most likely the reason they die. You're not supposed to have angled fitting immediately on the pump and sure
as hell you're not supposed to have 25 of them as they increase restriction quite a bit.
Think of the costs of the dead pumps at this point. Surely it will make you reconsider redesigning your loop.

[earthwormjim]

I've heard that running a D5 even for an extremely short amount of time unprimed will damage them permanently or reduce their life.

You have three pumps in series, did you make sure to only have the first pump powered when priming?

The two pumps that died, are they the second and third pumps in your set up?

[spitter3]

Put a volt meter on that power supply and see how may volts your actually putting into those pumps!!

[AmuseMe]

I use an Iwaki RD-30 with a dedicated MeanWell PSU. Also, don't gimp the pump by giving the inlet a 1/2" ID feed, that starves the bastard.
I'm feeding mine with a 3/4" feed. Also, you'll need a BIG res, not a little 250. Get a 400+ and thread a 3/4" NPT into it, so you you attach it to the feed of the pump.





You'll probably replace six D5's before this RD30 dies

[avabaska]

Hello

I've recorded a video:

Pump that died first - www.youtube.com/watch?v=rfkyEZMCBPE
Second - www.youtube.com/watch?v=_ar-EPT3_N0
Survived pump - www.youtube.com/watch?v=zl1UCejZEqs

Also I've made photos:

A filter (installed after the pumps):


One of the died pumps (to the left) and a new Vario. You can see that part with an impeller fell slightly down.


A died D5S:




Survived D5S



I've also installed two new D5V in dual TOP from EK. And of course can hear an unpleasant sound under high restriction (sure even when both valves are open) - www.youtube.com/watch?v=6EAqp4w2HJE

Than I've installed these D5V into ugly tops and certainly these sounds are disappeared. But flow rate dropped from 170-175 l/h to about 150-155 l/h.

So after that day I assembled all three D5V, even my old D5V (it's works as normal ) in TOPs from Alphacool and again we can hear that sound - www.youtube.com/watch?v=-9hTBIUawBw
But was 200 l/h

After that, was a final stage - an assembly of all three pumps in the original (ugly) TOPs. So now it's nearly silent and after adding a few towels - silent. But only 180 l/h. D5S @ 24V gave up to 260 l/h with these TOPs, but...

And I will definitely use a coolant with propylene glycol, I want a Koolance LIQ-702. But I don't think that pumps are died because of some reactions in my loop - now it's only distilled water, without any additives. Also I've changed this water three times within one month.
To change water every few months in a hard loop (high restriction + more than 5 litres or water) it's not a good idea. Even for 3 x D5S @ 24V it's not so easy, to start run this loop normally, the pumps work harder (start/stop anyway) So better to start it one time and forget about it for 1-2 years.


Erkla
>You're not supposed to do that. That pump linked in series with three other stronger ones was constantly running out of it's specs.
But mechanically they are identical. Just RPM "a little" lower.

>Your loop is poorly designed and that is most likely the reason they die.
It was the main requirement, not to think about any problems and limitations except high pressure, that will destroy pumps, water blocks, tubes and connections. But I think it was quite small (well, about 15 PSI, I guess).

>You're not supposed to have angled fitting immediately on the pump and sure as hell you're not supposed to have 25 of them as they increase restriction quite a bit.
Maybe that pumps are not designed to work in series and at high restriction? If so, I need another pump. Hope RD-30 will handle this.

>Think of the costs of the dead pumps at this point. Surely it will make you reconsider redesigning your loop.
No, even if RD-30 will die, I will just think about another pump. I know that in Vaillant gas boilers installed some good pumps. But now I don't want to think about it But I will never change my loop, because it's powerful, silent and very comfortable in maintenance.


earthwormjim
>I've heard that running a D5 even for an extremely short amount of time unprimed will damage them permanently or reduce their life.
Well... There was water all the time. I guess.

>You have three pumps in series, did you make sure to only have the first pump powered when priming?
I think yes. Maybe some time (during priming) there was not enough water to fill a whole pump, but I'm sure it was just seconds (tens of seconds), but that pumps are never worked without water at all and there was a water all the time near the bearing (thanks to clever construction).

>The two pumps that died, are they the second and third pumps in your set up?
Yes But can it be because first pump was in soft conditions (lower pressure and was no effect of the other pump)?

But... It all happened after I've changed water again. But I'm sure that never gave these pumps to operate without water


spitter3
>Put a volt meter on that power supply and see how may volts your actually putting into those pumps!!
Both PSUs - about 24V (up to +1%).


AmuseMe
>I use an Iwaki RD-30 with a dedicated MeanWell PSU. Also, don't gimp the pump by giving the inlet a 1/2" ID feed, that starves the bastard.
It is OK that I will use "PPCS 5/8" Barb to 1/2" Barb Pump Adapters"? My tubes are 13/16 mm.

>Also, you'll need a BIG res, not a little 250. Get a 400+
I've got Phobya Balancer 450. Hope it will be enough. But 3 x D5S @ 24V drank out of it very quickly

>You'll probably replace six D5's before this RD30 dies
I hope.

But as I'm understand, RD-x0 and Laing D5 and DDC, it's the pumps of various designs? Laing it's a magnetic centrifugal pump and RD-x0 it's a direct drive centrifugal pump. If I'm understand it correct, I'm afraid that Rd-30 will overheat if I will cover it by towels. Or it's OK?



Thank you all for your advices and time!

[n00b 0f l337]

Either version will overheat if its going to under towels.
That's partially to do with the load you place on the pump.
I can't see a direct drive being more likely, but the RD is a big beast, lots of power being used there, so I'd really not recommend covering it with towels. If it's "too noisey", then build a small box around it with noise deadening foam, and give it a tiny air hole or so.

[spitter3]

http://www.anonforums.com/builds/rd30/rd30parts.jpg

Not direct drive, a magnetic centrifugal pump.

[stren]

Either version will overheat if its going to under towels.
That's partially to do with the load you place on the pump.
I can't see a direct drive being more likely, but the RD is a big beast, lots of power being used there, so I'd really not recommend covering it with towels. If it's "too noisey", then build a small box around it with noise deadening foam, and give it a tiny air hole or so.

I thought both D5 and RD30s dumped their heat to the water, unlike the DDC type which didn't and required a heatsink (at least on the higher end ones)?

Something to bear in mind - there was some discussion as to whether Martin's RD30 results were not as good as they should have been due to the "gimped" feed. I have 2 D5strongs and a RD30 with a meanwell 24v supply, so one of these days this summer I'll try and see if I can recreate Martin's tests with an "amuseme" style feed and see how the two compare. Might be too late for you though

[avabaska]

I remember, when I've drain a water from my loop last time, it was a little dark (black) and from one of the rads (MO-RA3) it flowed very dark (almost black). I think it's not a rad's fault, it just big and got many tubes, so catched all of it. So all this dark water and small pieces, it was a pump's parts I guess.

n00b 0f l337
>Either version will overheat if its going to under towels.
We'll see

>If it's "too noisey", then build a small box around it with noise deadening foam, and give it a tiny air hole or so.
Tried to do this in first time with 4 x Laing but without a foam:



It was very loud. Resonance I guess.

But I think that D5S died not because of overheat, but everything can be.

spitter3
Nice, It's the first photo of this pump in disassembled condition that I've seen. So it will be cooled by water I think.

stren
>I thought both D5 and RD30s dumped their heat to the water, unlike the DDC type which didn't and required a heatsink (at least on the higher end ones)?
As now I'm understand, for DDC very necessary air cooling if this pump will work at high flow rates. I'm not sure, but maybe because of this in Martin's pump and rad optimizer we can see inscription "EXTREME DDC Cooling Suggested" when flow rate go up to 2 GPM and more.

>Something to bear in mind - there was some discussion as to whether Martin's RD30 results were not as good as they should have been due to the "gimped" feed.
Well, it's ok for me. Now I know that in my loop I need flow rate not below 170 l/h (0.74 GPM) or my 2-nd and 3-d cars will be hotter (up to 1-2 degree ). Now with 3 x Vario in original tops, I've got up to 185 l/h (0.8 GPM). When I will use a coolant it will be a little lower but still OK.

Anyway, when I will get my RD-30 I will install it. I'm not sure that all three D5 in my conditions, even at lower RPM, will feel OK during long period.

There is another way - to install a serious pump - circulating pump. For example Grundfos, Wilo or other. But they all AC and need grounding, and got a higher price. It will be last stage but I'm almost sure that RD-30 will work good even in my conditions.

[avabaska]

Hey, I've installed RD-30 and it works good.



With ~5 towels and sponge this thing is silent. But pretty warm even w/o the towels (all sides and back got +13C to water temp).
First it was 190 l/h, but after about three days ~195 l/h and I've got Koolance LIQ-702BU-B (27% of propylene glycol). It's about like 2 x D5S @ 24 V in original TOPs (was ~210 l/h on water).

PSU takes from the wall about 67 watts.

I think now I can relax. This pump got a good spindel+impeller fixation, because it's double, not like in D5, this part is OK.

Of course with 3 x D5S @ 24 V I had more fun Wanna second RD-30 but don't have enough space under the table... and for what. But I will have more than 260 l/h for sure

[Liam_G]

You are starving your pump of water. The inlet needs to be a "Big Slappy" kind of set up like all these other guys have used. I think that has been your problem all along, your D5's were not getting constant flow of water into them and there was air and bubbles flowing through and the bearings/impeller was operating at high rpm without enough water, especially using those 90* fittings. You are just starving your pumps of water.... Get a reservoir and place it right in the corner of your room just above your pump and run a 3/4" or 1 inch tube from your res to your pumps inlet with a nice slow curve of the tubing and no 90* fittings. place all those ball valves before the res. to not do this is just stupid. do exactly as AmuseMe has posted pictures http://www.xtremesystems.org/forums/...hmentid=127627 http://www.xtremesystems.org/forums/...hmentid=127628 of or you will just pointlessly kill another pump in time, even if you don't kill it you are wasting the pumps capability in your current set up.

[AmuseMe]

Ugh... I feel bad for that Iwaki! You're starving it! There is no point even using the damn Iwaki if your going to use some small inlet tube.
Scroll up... look at mine. Liam even posted a link. It has a 3/4" feed. My reservoir has a 3/4" NPT thread and it screws directly onto the inlet.

[Martinm210]

I thought both D5 and RD30s dumped their heat to the water, unlike the DDC type
which didn't and required a heatsink (at least on the higher end ones)?

Something to bear in mind - there was some discussion as to whether Martin's RD30 results were not as good as they should have been due to the "gimped" feed. I have 2 D5strongs and a RD30 with a meanwell 24v supply, so one of these days this summer I'll try and see if I can recreate Martin's tests with an "amuseme" style feed and see how the two compare. Might be too late for you though

That's only for people that haven't got a clue what total dynamic head pressure means. When I test pumps I use a manometer to measure pressure differential of the inlet vs outlet so any restriction at the inlet end ends up being captured as negative pressure. But I don't even have to wait for results, Iwaki has published pump curves and guess what...same result as I got. If anything my dynamic pressure numbers were slightly higher than Iwaki's published curve.

Thanks for supporting people that like to bash people without even looking at other available data. It is all right their on iwaki's site.. I guess Iwaki must have "gimped" their test also..

Is a small inlet good for operation..maybe not but it won't change a "pressure differential" measurement.

[stren]

That's only for people that haven't got a clue what total dynamic head pressure means. When I test pumps I use a manometer to measure pressure differential of the inlet vs outlet so any restriction at the inlet end ends up being captured as negative pressure. But I don't even have to wait for results, Iwaki has published pump curves and guess what...same result as I got. If anything my dynamic pressure numbers were slightly higher than Iwaki's published curve.

Thanks for supporting people that like to bash people without even looking at other available data. It is all right their on iwaki's site.. I guess Iwaki must have "gimped" their test also..

Is a small inlet good for operation..maybe not but it won't change a "pressure differential" measurement.

Lol very good points Martin thanks for that. I'd read rumors about it and was merely repeating them in case it was useful. Maybe that's a bad thing though. Besides I'm not a big fan of rumors vs hard data. The Iwaki fans seem to love the pump more than your data would indicate they should so I'm looking forward to doing some real world tests on it at some point to see if it really is justified at the high end.

[Martinm210]

I don't see the results as showing anything bad or weak. What other pump out there is capable of 24psi shutoff head pressure with a single pump that fits in a pc reasonably well with industrial build qualities? None that I know of, but an RD30 is also not perfect. Its big..it requires non standard voltage to operate at full speed, it's not silent, it consumes and dumps high heat levels at 24v, and by the time you buy the pump plus a good 24v power supply you will have invested several hundred dollars. I personally find a few DDCs or D5s in series to be more practical fitting solution to most extrme systems. Anything more than about 40watts of pumping power will typically result in diminish returns..no wonder RD30s were suggested to be operated at 18v back in the days..24v is typically beyond optimal.

Nothing synthetic or unusal about a pump curve, it is THE standard convetion to report hydraulic performance on pumps in all areas of engineering because it test all possible conditions..not just one point that may or may not represent the condition others will see..it shows the full spectrum.

[avabaska]

Hello. RD-30 works good, but I've got a small problem - leakage around the pump connections and only at the outlet (cause of pressure). I think it's because of stupid screw clamps and weak nozzle over the pump (I'm afraid to twist screw very strongly).



Also it's becasue of 27% of propylene glycol - it's oily and very slick, and got a higher fluidity. So only when I've changed screw clamps to plastic ties (three on each) and one on 16 mm (but I've got there a spring clamp!) the problem of leakage is gone (I hope).

But it's only ~1 atm. What it would be if I will install second RD-30 and it was about 1,85 atm? I don't want to do that but just interesting.

Liam_G
>You are starving your pump of water. The inlet needs to be a "Big Slappy" kind of set up like all these other guys have used.
I don't have any choice (pump too ). I'v got all hoses 13/16 and pump is in another room, so the hoses pass through wall. 16 mm it's pretty thick for it. But any way, I'm sure for 100%, that even if I had thicker hoses, RD-30 will give the same ~195 l/h in my case. And D5S are died anyway (and they got smaller diameter, just like for 13 mm or about).

Example - I've got three EK-FC570/580-GTX+ in series and 195 l/h. Series-parallel connection gives only 215 l/h. So I don't see any point. Only the pressure will be not so high (I guess).

Regarding to 90 dgs fittings, I knew it from the beginning. I'v got a big loop (radiators on the shelves, pump in another room and so on). Not enough flow rate? Well, we will get another pump. 195 l/h it's not great but good enough. RD-30 is still alive (even in towels and when it's case is heated up to 50C).

Only leakage was a problem... Hm...

AmuseMe
>There is no point even using the damn Iwaki if your going to use some small inlet tube.
It's about like two D5S @ 24V in series. So it's normal.

Scroll up... look at mine. Liam even posted a link.
Is your pump located in another room?

[Kayin]

From what I see in your sig, a single standard D5 or two would keep a decent flow rate and temps if you removed some of the 90 degree fittings. Those rads aren't restrictive at all, nor are the GPU blocks. Board block is a little, there's some from the CPU block, but most of your restriction is coming in from fittings-QDCs and 90 degree bends. Because water is incompressible, any pump at all is going to generate loop movement, but you're pitching extreme overkill at the loop without actually addressing the issues that cause pump death in the first place. THe same concept of water being incompressible is why the pumps tear up-something has to give at the enormous backpressures generated by your loop. It's almost designed to kill pumps.

[Liam_G]

Liam_G
>You are starving your pump of water. The inlet needs to be a "Big Slappy" kind of set up like all these other guys have used.
I don't have any choice (pump too ). I'v got all hoses 13/16 and pump is in another room, so the hoses pass through wall. 16 mm it's pretty thick for it. But any way, I'm sure for 100%, that even if I had thicker hoses, RD-30 will give the same ~195 l/h in my case.


Regardless of the tubing size you run on your whole loop, if you want to take some stress off the RD-30, install a reservoir right above your pump with about 300mm of 3/4 inch inner diameter tubing or pipe between the reservoir and the pump, that way the pump is always getting enough water. there is nothing stopping you doing this so i can only guess that your not quite understanding what we are describing. remove a couple of 90 degree fittings and you will have a very nicely performing loop.

[Vinas]

It can be hard to keep these pumps from leaking. I would suggest using worm screw type clamps or compression's. I've been using an iwaki clone (http://www.marinedepot.com/pumps_pan..._px_ps-ap.html) for about 5 years and quickly learned that you can't be *sure* your loop isn't leaking unless you use very tight connections.

[Martinm210]

This is an example of where running two smaller pumps in series can be a pressure benefit. Modifying the inlet to reduce negative pressure isn't going to lower the outlet posotive pressure at all..actually it will make it higher. But...two pumps spread out around the loop will lower the maximum system pressure since pressure drop by parts between the two will mitigate the problem. Regardless, 1/2" x 3/4" tubing and worm drive clamps have always been my most trusted. You need enough wall thickness to both support the pressure and distribute irregularities in the clamp