Temperatures in relation to flow-rates. Data on several CPU-Blocks

[HESmelaugh]

I am confused, isn't the 1T the MCP350.
The Laing DDC-1 plus is the MCP355 innit?

Or is this a regional thing.

I thought it was:
MCP350 = DDC-1T
MCP355 = DDC-1T+

But I'm not sure. Memorizing cryptic product descriptions isn't one of my strengths. :-)



I got an Update for you guys:

As promised, I wanted to see if I could reach those 2 GPM everyone goes on about here. ;-)
So, I installed one of these:


I also tightened up the loop as far as I could, what with the ball valve, temperature-sensor and flowmeter needing to be in there. Result:



And what did this effort and a dual-pump top with two MCP355 achieve?



* drumroll *



... about 1.2 GPM.

Still far from 2 GPM.
I'll post the data from the test-run anyway. Tomorrow. I'll have to make a pretty graph or something, first.

[orclev]

Who makes that, and where did you get it? Every double pump top I've seen has always had one DDC on the bottom and one on the top, not side by side like that. I'm currently trying to figure out how to cram my dual pumps into my case (which is very quickly running out of usable room), connect them to each other, and not kink my tubing all at the same time. One of those would go a long way towards fixing that problem (although I'd want it made out of Delrin or similar, I don't want any acrylic in my build).

[HESmelaugh]

It's made by Watercool, a German company. I got it from Aquatuning, a German online store. No idea if or where it's available internationally. The acrylic version is the only one available, so far.

Link to product

[Hondacity]

wow does that mean two 355s with reservoir tops in series won't reach 2gpm?

[Giannis86]

adding a second ddc increases head pressure far more than in it increases the flow rate

[Hondacity]

Giannis

does it really increase head pressure? where do you base your statement?

[yngndrw]

Just out of interest, try them in parallel.

[skinnee]

wow does that mean two 355s with reservoir tops in series won't reach 2gpm?

That dual top might not be the best candidate for the testing.

[Hondacity]

That dual top might not be the best candidate for the testing.

its not? what is a good combination for 2gpm other than the iwaki? the iwakis are just too expensive...

[skinnee]

take a look at Martin's handy dandy Flowrate Estimator...only a select few tops on 2 mcp355's in series don't reach the 2+GPM mark.

Disclaimer: "handy dandy" is not a term endorsed by Martin. Purely the opinion of the poster.

[Hondacity]

have you tried it with two res tops on the 355 + ek supreme + 2 rads?

its says 2.09gpm

but i dunno if thats tested or not..its theoretical..but not proven...

thats why i'm suggesting to Hes to try it

what do you suggest other than the iwaki?

[HESmelaugh]

First of all, here's the data:

I measured temperatures with the same methods described in above posts. The setup reached a maximum of 1.19 GPM and I also measured temperatures in this same setup at 0.92 GPM and 0.66 GPM.



Clearly, the downward-trend in temperatures continues with increased flow-rates. Between 0.66 GPM and 1.19 GPM, temperatures dropped by about half a degree. Also very interesting to see: The added benefit of higher flowrates is greater than the drawback of an extra pump dumping it's heat into the loop - the temperatures at 1.19 GPM were the best I'd seen in all of the test-runs.


Considering further testing:
The loop pictured above will be disassembled today, since there is other testing, that needs to be done. I'm testing some pump tops, among them one or two that even you at XS might have not seen yet. :-)
After those tests, I might find some time to chase 2 GPM again.

Question: If I use two pumps with res tops in a loop, won't one of the reservoirs overflow?

[Hondacity]

yes the second will overflow...i used teflon tape to seal my second pump..i bought the non res top so it won't overflow for the third pump...if you look at the picture i posted

[HESmelaugh]

Thanks for the info. I might just try to use one res top and one non res top, then.

[Hondacity]

Just put the pump with res top before the regular top pump, it will work fine...

Pump (res top) > Pump (regular top)

have you tried this at all? even with martins flow calculator your suggestion fails 1.8gpm

[Hondacity]

regular top = stock top

lol

[jcool]

It's true, we Germans don't give a rat's ass about flow rates. And why should we, spending a lot of money for like 2-3C less with lots of additional noise is totally not worth it.
I run 2 EK Supremes, 2 big Rads, the HDDs and raidcon with around 60cm flow height and ~3m tubing length on a single Aquastream XT pump in one loop, that's probably not even 0,2gpm. Still I don't get more than 50C Coretemp on either CPU

You want better temps? Run phase.. that's what I do, anyway. Water is for good yet silent cooling, it's not about 2GPM+ flowrates.

You people will realize that soon enough

Oh yeah, really nice testing there.

[IanY]

The same people who insist on 600/700/800 lb. ft. of torque and 500/600 hp in a family passenger car, or even station wagon/estate, prefer 0.2 GPM pumps lol How strange this world is.

[IanY]

@HondaCity: dude, that second reservoir is robbing your flow rate.

[jcool]

The same people who insist on 600/700/800 lb. ft. of torque and 500/600 hp in a family passenger car, or even station wagon/estate, prefer 0.2 GPM pumps lol How strange this world is.

I gotta follow what my customers demand.. and they demand absolutely quiet rigs that eat close to 1KW. They want them to perform better than anything else on the market, while being quieter than your average office PC.

They don't care whether the CPU temp is 56C or 61C, so neither do I for water

[yngndrw]

jcool: I've got to agree with you. Look at the last graph that HESmelaugh posted:


Nearly double the flow rate lost 0.5C. That's only 1/50th, or 0.02%.

[G-UNIT91]

I thought it was:
MCP350 = DDC-1T
MCP355 = DDC-1T+

But I'm not sure. Memorizing cryptic product descriptions isn't one of my strengths. :-)



I got an Update for you guys:

As promised, I wanted to see if I could reach those 2 GPM everyone goes on about here. ;-)
So, I installed one of these:


I also tightened up the loop as far as I could, what with the ball valve, temperature-sensor and flowmeter needing to be in there. Result:



And what did this effort and a dual-pump top with two MCP355 achieve?



* drumroll *



... about 1.2 GPM.

Still far from 2 GPM.
I'll post the data from the test-run anyway. Tomorrow. I'll have to make a pretty graph or something, first.

You must try the SANSO PDH-054...

[Hondacity]

@HondaCity: dude, that second reservoir is robbing your flow rate.

it is? explain please

..sorry i don't have a flow meter stuck to it right now..and i have three pumps in series...lol..and its freaking cold outside...

[IanY]

it is? explain please

..sorry i don't have a flow meter stuck to it right now..and i have three pumps in series...lol..and its freaking cold outside...

Buddy,

Look I don't own XSPC reservoir tops, or for that matter, any XSPC tops. I'm not into scientific explanations or scientific data. I'm no engineer by training and honestly, all that lab data bores me to tears and couldn't disinterest me any greater. I'm the antithesis of Martin.

Generally, when I see graphs and charts, I go look at a different discussion. Is it because I don't understand graphs and charts? Not exactly. Years of applied math/maths took care of that. Its just that I have no interest in those.

What I do know is that I have tried multiple reservoirs in a loop before using small MCRESes. Why did I even do that? I don't know. Bored I guess.

The pertinent message here is that it does affect your flow rates, and quite considerably more than you can imagine. The turbulence inside a reservoir itself dampens flow rates.

I don't doubt Martin's recommendations and the XSPC reservoir top may be the very best top right now for the DDCs. However, I do know that Martin was referring to a single reservoir top in isolation. He wasn't envisioning the scenario where two are used in series.

Since money is not a great issue for you, I hgighly recommend that you purchase a regular XSPC top and use that in conjunction with the XSPC reservoir top.

Just a note. Just because I don't offer data or scientific explanations doesn't mean that I don't know what I am talking about, after all these years of observing others and trying things out.

As for me, I would never buy a XSPC top of any nature because its acrylic, but that's just me. It would give me a terminal heart attack. Shouldn't influence you in the least bit.

[HESmelaugh]

jcool: I've got to agree with you. Look at the last graph that HESmelaugh posted:


Nearly double the flow rate lost 0.5C. That's only 1/50th, or 0.02%.

I was wondering if it would be neatly exponential. Something like 0.5 deg less every time you double the flow-rates. But while you gain lots of cooling performance doubling a very low flowrate, it looks like it's not even going to be 0.5 for each doubling in higher flowrates... Can't tell for sure, though.

The same people who insist on 600/700/800 lb. ft. of torque and 500/600 hp in a family passenger car, or even station wagon/estate, prefer 0.2 GPM pumps lol How strange this world is.

Yeah, that is kinda weird. But the whole thing doesn't make sense anyway. I always find it amusing, when someone on a (usually German) forum goes: "what's the point in getting a better pump? you hardly get better temperatures at all!". We're talking about liquid cooling PCs, it's not like spendign piles of money to get a nice, shiny liquid-cooling setup in your PC is the most significant thing you can do in your life. It's just for fun. So what, if someone wants flowrates that put the Niagra Falls to shame?

You must try the SANSO PDH-054...

Really? In the shop where I found it, it's rated for 360l/h. The MCP355 is rated for 600l/h... I haven't ever tried the Sanso, though.