MCR-QP Stackable Radiator Series Released

[DeathWalking]

you can't install a closed corner fan in a stackable config

I'm fairly sure that's what Naekuh is objecting to.

[Utnorris]

Dual-row radiators also run in parallel Utnorris. The only difference is the additional pressure drop caused by the presence of the fittings.

I get that, however in this situation the water is going to flow in the direction of least resistance and therefore my concern would be that it would bypass the first rad and just flow through the second rad. I am not saying that no water would flow through the first rad, but rather that the amount would be siginificantly less than the second rad and so forth. Now I have never ran two rads, video cards or anything else in parallel, so this is more of a question than statement and I am just trying to understand how this will work. I am thinking about running my 2 gpu's in parallel, but from what I have read is that you lose performance because the water by-passes the first block for the most part and goes to the second block. I guess it would help if someone could link me a picture or diagram showing how the water should flow.

[Alex Suraci]

Would this would fit at the top of an ATCS 840 and be out of the way of the motherboard? The total thickness is 93mm. (I'd measure the case myself but I don't have one yet.)

[gabe]

I am thinking about running my 2 gpu's in parallel, but from what I have read is that you lose performance because the water by-passes the first block for the most part and goes to the second block. I guess it would help if someone could link me a picture or diagram showing how the water should flow.

that's inaccurate. as long as the pressure drop in both devices is similar, coolant will flow in both.

There is a loss in coolant velocity in a parallel setup, causing a local loss in heat removal capacity, but since pumps having a finite ability to push fluid, parallel setups may become beneficial in many circumstances. GPU cooling is among the most typical candidates for such configurations. By minimizing the overall pressure drop in the loop, you maintain high heat removal where it counts most, the CPU, while keeping adequate heat removal capacity at the GPU's.

[Zehnsucht]

Have you done any testing gabe on a setup with one 120/240/360 rad compared to a stacked scenario? I think this would be extremely interesting results.

[Utnorris]

that's inaccurate. as long as the pressure drop in both devices is similar, coolant will flow in both.

There is a loss in coolant velocity in a parallel setup, causing a local loss in heat removal capacity, but since pumps having a finite ability to push fluid, parallel setups may become beneficial in many circumstances. GPU cooling is among the most typical candidates for such configurations. By minimizing the overall pressure drop in the loop, you maintain high heat removal where it counts most, the CPU, while keeping adequate heat removal capacity at the GPU's.

Thank you for the explanation.

[Martinm210]

Very clever, I especially like the compatibility with existing MCR rads so you only have to buy one stackable. I'm curious to see what the general performance gain is also. If double thicknesses are showing around 8-15%, and someone's earlier test was showing stacking in series worth nearly 30%, I'll guess at a 20% figure with preferences towards the higher speed or 38mm pressure fan scenario with just one set of fans more with more fans obviously....how much I couldn't even guess.

I agree some very well articulated test data to go along with this would be helpful for people. Something like a single MCR320 with fans in pull vs MR320 with the same fans in push & Pull vs the three fan scenarios on this stacking setup. Keep it simple like fans used and percent gained or degrees lost in water temperature of a simulated system heat load...something like that would be great.

This would provide the necessary fan power guidance people will be looking for as that's a bit unknown with this stacking, and likely fairly important with the air pressure drop doubling stacked like this.

My vote goes for the MCR420. Quad sizing is becoming common enough. Unfortunately right now, it's reserved for the high priced double thickness only segment. Time for a quad, please...

The stacking is a really really nice option though. I'd like to see a stacked MCR320 setup against a PA120.3 for example...I bet it's good!

[SNiiPE_DoGG]

if I had a tube of thermal paste and no homework I could do a test right now.... but at school its just not practical for me to remove radiators from the loop as my dorm room is small and not all of my tools are here :\

[gillbot]

you can't install a closed corner fan in a stackable config

Sure you can,

[Mark90]

3M Red tape!

Anyways any reviews for the stacked rads gabe?

[SaiNRuB]

Sure you can,

[Blacky]

If you have space, you are better off (always) installing a second rad by itself (in series), but if you do not have enough room, then this will give a very serious boost to your thermal performance at relatively low cost.

The problem we all face here is to strike a balance between space constraints, thermal performance, comfort of operations AND cost of solution. If space is the defining factor, then it forces compromises on the other 3. The stackable solution aims at minimizing the impact on cost of solution.

I see thanks

[SNiiPE_DoGG]

Plus if your cool like me then you can run the stacked rads in series with waterflow opposite to airflow and get even more performance

[ultimeus]

Is there a release date yet ?

[Quad-Damage]

I Want It!~

[eXa]

Plus if your cool like me then you can run the stacked rads in series with waterflow opposite to airflow and get even more performance

How?

[Chruschef]

How?

http://www.xtremesystems.org/forums/...0&postcount=27

[CyberDruid]

that's inaccurate. as long as the pressure drop in both devices is similar, coolant will flow in both.

There is a loss in coolant velocity in a parallel setup, causing a local loss in heat removal capacity, but since pumps having a finite ability to push fluid, parallel setups may become beneficial in many circumstances. GPU cooling is among the most typical candidates for such configurations. By minimizing the overall pressure drop in the loop, you maintain high heat removal where it counts most, the CPU, while keeping adequate heat removal capacity at the GPU's.

First off thanks a lot for involving the community in your product launch. It's a great idea and well implemented.

Second I am happy to hear a real Pro and Expert say what I've been telling people for years. There's more than one way to plumb a loop and it's crazy some of the notions people have about how water flows and closed loops operate. Even after setting up examples similar to your sandwiched rads and showing temps people still think it's less than ideal and point to linear setups as "normal" and "standard" so it's nice to see you promoting a New standard

[SNiiPE_DoGG]

First off thanks a lot for involving the community in your product launch. It's a great idea and well implemented.

Second I am happy to hear a real Pro and Expert say what I've been telling people for years. There's more than one way to plumb a loop and it's crazy some of the notions people have about how water flows and closed loops operate. Even after setting up examples similar to your sandwiched rads and showing temps people still think it's less than ideal and point to linear setups as "normal" and "standard" so it's nice to see you promoting a New standard

its like the old myth that used to perpetuate itself around the DD watercooling forums and maybe here too that running more than one rad in a single loop wasn't beneficial to temps because it created too much restriction

[CyberDruid]

Yeah like surface area doesn't matter to heat transfer.

I've always liked the MCR line...it's simply the most cooling for the least money. Now it'll be even easier to get the amount of cooling you need into a small area.

[eXa]

http://www.xtremesystems.org/forums/...0&postcount=27

Oh, i see. But that would require someting like 2x 38mm fans in thicknes between the rads. Well no biggy if you have room for it.

[eXa]

First off thanks a lot for involving the community in your product launch. It's a great idea and well implemented.

Second I am happy to hear a real Pro and Expert say what I've been telling people for years. There's more than one way to plumb a loop and it's crazy some of the notions people have about how water flows and closed loops operate. Even after setting up examples similar to your sandwiched rads and showing temps people still think it's less than ideal and point to linear setups as "normal" and "standard" so it's nice to see you promoting a New standard

well if you are able to feed both rad fresh air it is less ideal to stack them instead.

This idea is great because often you dont have any place to put a second rad, but you have quite alot of room behind the one rad you already have.

[SNiiPE_DoGG]

Oh, i see. But that would require someting like 2x 38mm fans in thicknes between the rads. Well no biggy if you have room for it.

see the premafrost worklog, I have pics of the sandwich and its ~150mm thick

[Alex Suraci]

What kind of fans would you want sandwiched between these?

Edit: Also, how would two of these stack (ha!) up to a thicker rad like the Feser?

[Martinm210]

First off thanks a lot for involving the community in your product launch. It's a great idea and well implemented.

Second I am happy to hear a real Pro and Expert say what I've been telling people for years. There's more than one way to plumb a loop and it's crazy some of the notions people have about how water flows and closed loops operate. Even after setting up examples similar to your sandwiched rads and showing temps people still think it's less than ideal and point to linear setups as "normal" and "standard" so it's nice to see you promoting a New standard

Yeah, running them sandwiched in parallel as this setup is doing is really no different than every single double thickness radiator out there with the exception of the very minor additional losses at the fittings. They all run two rows of tubes at the same time in parallel. Series has just become the standard because parallel requires more detailed planning around restriction of components and generally not necessary when pumps are strong enough to run in series anyhow.

The problem with parallel lies when someone might want to run a highly restrictive CPU block and something with very low restriction in parallel loops. There are combinations out there where you could potentially have 10X more restriction on one route than the other. In that case the CPU block could only see 10% of the system flow rate. Even at a high/strong system flow rate at something like 2GPM, the CPU block may only see .2GPM, that's entering flow rate loss territory (Below .7gpm)...and that's what you need to be careful with on parallel systems. Then there is the fact that we're generally spoiled now with overly powerful pressure oriented pumps. It's becoming pretty hard to run too much in series to a point where flow rates are below 1GPM with today's DDC/MCP type pumps. The old weak on pressure pond style pumps were much more critical in the pumping power and pressure drop specs. Who knows though, blocks continue to become more and more restrictive, so times may change again to where pumping power becomes more of an issue again..

I think the "Run it in series only" standard is just what's proven to work over the past couple of years, it requires no real planning or understanding of pressure drop difference between components or the net end resulting flow rate. Generally you can buy a Laing DDC or D5, string up your heart's desire in series and call it good.

In this case, it's stacked rads, equal pressure drop, 50% system flow rate provide to each radiator and the same as all double thickness radiators. Running them in series might make a laboratory test difference, but it's likely near immeasurable for us general use folk..