Testing the HW Labs Micro 184 GTX Rad

[b@llz0r]

I set up my aquaero today and thought id use it to test out a little radiator ive had sitting there not doing anything much because i thought it was too small.

The HW Labs M184 GTX, a dual 91mm radiator. With a frontal area of just 165cm² its a bit smaller than a 140 (196cm²) but bigger than a 120 (144cm²).

Ive hooked up the aquaero as follows

Sensor 1 - Air in
Sensor 2 - Air Out
Sensor 3 - Coolant

Fan 1 and 2 are on the rad, with speeds controled by sensor 3 (coolant) min 25°C max 30°C



The Loop

The loop consists of EK Dual DDC>Dtec Fusion on a QX9650>S2E Fusion NB/SB > EK FC 295 > GTX184 > EK Multioption

ewwww


dont wory about the rough setup... this is just a test bed.

Fan Specs

The fans im using are *cough*thermaltake*cough* 91mm fans... I know I know... friends dont let friends bla bla bla, but the specs on these looked pretty good for what I wanted... and the shop only had these or "gentle typhoons"

Fan Dimension - 90 x 90 x 25 mm
Fan Speed - 1300 RPM @ 20C ~ 4000 RPM @ 55C
Bearing Type - 1 Ball 1 Sleeve
Noise Level - 17.0 dBA @ 1300 RPM 48.5 dBA @ 4800 RPM
Max. Air Flow - 24.6 ~ 78.7 CFM
Max. Air Pressure - 1.30 ~ 7.20 mm-H2O
Power Connector - 3-pin
Rated Voltage - 12V
Started Voltage - 6V
Rated Current - 0.15 ~ 0.48A



Idle temps...

well CPU crunching at 100% is idle for me since its going 24/7
Fans running at 5.9V, nice and quiet
Coolant is at 27.5°C with Air in at 20.8°C (6.7°C delta)

CPU was at 37°C
GPU at 31/38°C (damn my uneven 295s)

Ahh crap... I just realised the screencapture utility didnt get my everest osd



Full Load

As you can see the little bastard is barely breaking a sweat, Fans up to 77% (~10V), Air in 19.6°C, Coolant at 29.1°C (9.5°C Delta), GPU temps leveled off at 46°C.






overclocked and loaded



fans went to 12V trying to keep the coolant under 30°C.

All in all im very impressed by HW Labs little rads.


Testing the Micro 80 next

[alacheesu]

Thanks for testing. ~4000 rpm? o_O

[b@llz0r]

Thanks for testing. ~4000 rpm? o_O

yeah, but LinPack and Furmark are waaaay over the kinda load any real world app will place on your system.

With just boinc running it sits at 6V ~2000 RPM & ~20 dBA

With GTX rads its all about having fans with a good range and a good fan controler.

[Wadkiller]

Bloody hell, that little thing actually works! Thanks for the testing!

I've only started watercooling the beginning of this year & everywhere i've searched the conclusion were "get that big ass rad". this small little thing performs very well. I'm really glad i got 2 mcr220's, one for the cpu loop & the other for the gpu loop. I know bigger is better, but small might work good enough.

[Sadasius]

With GTX rads its all about having fans with a good range and a good fan controler.

Damn right! Should put an extra 2 behind the rad for a push/pull. I have two GTX 240's with San Ace fans at 12v in push/pull and I literally have almost everything in one loop with 2 pumps (no RAM and HDD cooling) and the water temp cannot break 27 degrees with a 24.4c ambient. 27 degree's would also be at my hottest sensor (have 2 in the loop before each pump) as the other sensor is usually .5c below.

[orangekiwii]

this is good for sff cases a push pull config on 2 of these stacked in drive bay would work well

[Waterlogged]

I've had an odd interest in these for a little while now, as I may have a use for them.

Very interesting results indeed, thanks for the test b@llz0r.

Is the flow pattern set up the same as their bigger brothers or is it like traditional rads?

[DarthBeavis]

I wonder if I would be better off using three of these in series or a Feser 480 with the Koolance manifold. The goal is to give the CPU fresh water then pass it to teh NB/SB/mofsets, the GPUs get fresh water, and the 360 get fresh water. I am not sure what would be the lowest restriction and most efficient.

[NaeKuh]

I wonder if I would be better off using three of these in series or a Feser 480 with the Koolance manifold. The goal is to give the CPU fresh water then pass it to teh NB/SB/mofsets, the GPUs get fresh water, and the 360 get fresh water. I am not sure what would be the lowest restriction and most efficient.

this would only be if water equalibrium wasnt playing a role DB.

[DarthBeavis]

this would only be if water equalibrium wasnt playing a role DB.

so you think multiple of the midget radiators would be better in series? I will probably use a single vario D5

[NaeKuh]

ballzor any chance you can connect an inline 300W heater and see how it does?

I think the majority of the testers use a 300W heater to get a delta.

Would love to see an exact load to see how it scales.

so you think multiple of the midget radiators would be better in series? I will probably use a single vario D5

I think you'd be better off with the single fesser 480. Less flow reduction, more passes in the same time, as well as a larger cooling area.

[DarthBeavis]

ballzor any chance you can connect an inline 300W heater and see how it does?

I think the majority of the testers use a 300W heater to get a delta.

Would love to see an exact load to see how it scales.



I think you'd be better off with the single fesser 480. Less flow reduction, more passes in the same time, as well as a larger cooling area.

I will setup the Feser loop in series and parallel (via manifold) and compare temps. Wish I had a way to test flow but temps at load should be good enough. Anyone want to guess what the delta will be and in which direction?

[b@llz0r]

Is the flow pattern set up the same as their bigger brothers or is it like traditional rads?

Im guessing its the same... is there a way to tell without cutting it open?

so you think multiple of the midget radiators would be better in series? I will probably use a single vario D5

Are you meaning Rad>Block>rad>block>rad...
remembering that fluid temps are pretty much even throughout the loop I think that a single rad would be preferable to multiple simply due to the reduced head loss.

One thing that would be cool is these rads are so small, you could tuck a few of them away into spare spots inside the case and add up to having the equivalent of a 480 without having a case that fits a 480 (not sure i explained that too well).

ballzor any chance you can connect an inline 300W heater and see how it does?

I dont have a heater.

[DarthBeavis]

One thing that would be cool is these rads are so small, you could tuck a few of them away into spare spots inside the case and add up to having the equivalent of a 480 without having a case that fits a 480

The case will have an MATX and Xbox 360 mobo mounted next to eachother. Plan is for case to be super skinny but as long and tall as needed (wall mount so thinness is the important factor). That means radiator size is not key as long as we are not talking a Monsta.

[orangekiwii]

b@llzor thats what I was *trying* to get at when I was talking about SFF cases...

just cram a ton of small radiators into a smaller case and you'll have a similar affect of having one large radiator

[Waterlogged]

Im guessing its the same... is there a way to tell without cutting it open?

Stick something like a pen down the fitting holes, it should go in deeper on one side if it's set up the same.

[ILikeCosmosS]

i think 2 of these + a swfitech micro res and mcp355+top might work for a xbox 360

[Taylor01]

Wow those are some nice temps I'm getting around 60c load with a q9550 @3.6 and thats with both the the 184 & 160. I was hitting 72c with the gtx260 in the loop. I wonder why my temos are so high? Here is a pic if any one is interested sorry I couldn't get it to upload.

http://www.xtremesystems.org/forums/...1&d=1237172964

[b@llz0r]

Wow those are some nice temps I'm getting around 60c load with a q9550 @3.6 and thats with both the the 184 & 160. I was hitting 72c with the gtx260 in the loop. I wonder why my temos are so high? Here is a pic if any one is interested sorry I couldn't get it to upload.

http://www.xtremesystems.org/forums/...1&d=1237172964

what fans r u using?

[Taylor01]

For the 92mm's push I used extremely loud panaflo H1 @ 56cfms and Zalman ZM-F2 for the pull I don't know what the cfm for them are but there are a real nice fan quiet and move a good amount of air. As for the 80mm's pushed with Cooler Master Y720D8D-29T1-GP @45.8 cfm these fans and pulled the the stock silverstones I think their about 25cfm.

[b@llz0r]

what RPM?

These little thermaltake fans run at 2000 rpm at 6v... you need a lot of pressure.

It might help trouble shooting if you had temp sensors for coolant, air in and air out

[Taylor01]

Coolermasters @2900rpm zalmans 2800rpm & panaflos @2850rpm all 12v

I guess I should add that I have moved on to a new rig this current HTPC is for sale one ebay right now.

http://cgi.ebay.com/ws/eBayISAPI.dll...m=230343569906

[Waterlogged]

Wow those are some nice temps I'm getting around 60c load with a q9550 @3.6 and thats with both the the 184 & 160. I was hitting 72c with the gtx260 in the loop. I wonder why my temos are so high? Here is a pic if any one is interested sorry I couldn't get it to upload.

http://www.xtremesystems.org/forums/...1&d=1237172964

What season and time of day did you take your temp readings? If I'm not mistaken b@llz0r is in fall or winter about now.

[Taylor01]

well I finished the build around feb. here in Mich. it was around the -20F and had it running a few days ago when it was 80f. It doesn't that ambeint temps are effecting it that much.