Little River Storm G4 Review Posted
 

As good/better performance than the Cascade SS and less than half the price. What's not to like? Perhaps the fact they are all already spoken for. At any rate, here's Procooling's

Storm G4 Waterblock Review

Very nice results. Shame I was away when they came up for order, sigh sigh.

Surprised about the very low flow performance, although incredible, I'd have though it'd still beat the cascade. Not like it matters to me....

Cathar
I am surprised by the "1.7x" in
"jet velocities for the Storm are typically about 1.7x that of what occurs within the Cascade when given the same pump."

An extension of the discussion would suggest a lower number. :-

http://www.jr001b4751.pwp.blueyonder.co.uk/Storm1.jpg

Does not(yet) include "tubing + radiator" pressure drop . However, suggest, their inclusion further reduces the effect of "Total Orifice Area" area

Don't know what to say Les. I'm just reporting what I measured, and not extrapolating from theoretical pressure drops.

Here's a chart of the 50Hz MD-30RZ jet velocities with a singular diameter hole with a flattish (67° from the pepenpendicular) straight chamfer. These are measured values, not theoretical extrapolations. The Storm's jet inlets are significantly more optimised than either the following test results, or the Cascade jet inlets.

Iwaki MD30Z

2.00mm 2.25 LPM 11.9m/s
2.78mm 4.20 LPM 11.5m/s
3.18mm 5.30 LPM 11.2m/s
3.50mm 6.40 LPM 11.1m/s
4.00mm 8.00 LPM 10.6m/s
4.37mm 9.00 LPM 10.0m/s

1.7x was vs a full system where the pressure drop of tubing + radiator is more significant for the pump to work against for the Cascade vs the Storm. Could get as low as 1.55x for some weaker low-head pumps I guess, but pretty close to the 1.7x mark for the higher head style DC pumps.

Jet orifice area for the Storm is ~53% that of the Cascade that Phaestus tested. As you can see for Phaestus, peak flow for the Storm in a full system was ~1.72gpm, and for the Cascade ~1.98gpm, or ~87% of the peak flow rate. 0.87 / 0.53 = 1.64, meaning that the Storm had 1.64x the jet velocity of the Cascade for Phaestus.

Maybe "typically 1.6x" would be a better number to quote to be safer.

Unsure of test conditions, but have included in graph:-

http://www.jr001b4751.pwp.blueyonder.co.uk/Storm2.jpg

Possibly fair agreement, but tells us little


Applaud the "total orifice area" optimization , but still question the "1.7x"

Do not argue with those numbers.
Possibly nozzle design has a greater influence than I envisaged.
On pure area consideration would not have put above "1.35x"
Hat-off to your improved nozzle design

Seems to be a clash of theory vs practise.

Perhaps dP of each block may help somewhat here for the theoretical angle analysis.

At 10LPM, dP of Storm/G4 is ~7.8mH2O, and for the basic Cascade is ~4.8mH2O.

Would suggest perhaps plotting a dP curve as suits against a few pump PQ curves, including dP for a heatercore and tubing. Determine the final flow rates, and then factor in the 0.53 sized orifice area for the Storm over the Cascade to arrive at jet velocity boost. I think you'll find that even for the weak Eheim 1046, or the E1250 which seems to be causing the most troubles in those graphs, that the boost is very substantial.

Both Cascade and White Water really did have pretty poorly done jet/nozzle inlets. A LOT of wasted pressure drop there. Cascade was also a lot more restricted after the jets as well. A fair amount of gain was found there too. This is also seen with the Cascade SS vs the normal Cascade, where the large bulk of the SS's lower pressure drop came from improving the ease with which the water could exit the cups. At a machining level the difference is positively tiny, being barely perceptible with the naked eye, but it's clear that the Cascade was needlessly hampering cup discharge. The Storm design also built on those lessons.

http://www.jr001b4751.pwp.blueyonder.co.uk/Storm3.jpg
(No Eheim1046 data handy)

IwakiMD30-RZ 8.7/10.2 x1/0.53 = 1.61
SwiftechMCP-600 5.1/6x1/0.53 = 1.60
Eheim1250 4.55/5.55x1/0.53 = 1.55
Eheim1048 3.75/4.55x1/0.53 = 1.56

Going to the pub soon.

Edit:
Added
Iwaki MD-20RZ(M)-N 60MHz 7.8/9.05x1/053 = 1.63
Eheim1046 3.0/3.5x1/0.53 = 1.62
Iwaki MD-20R(M)-N (60Hz) 6.9/8.5 x1/0.53 = 1.53

Righto, maybe 1.6x would be a fairer figure and I'd accept that. I measured 1.65-1.67x myself with the 30RZ, the MCP600, and the MCP650, and so just rounded to 1.7x for convenience. If Phaestus is willing to update change the 1.7x to 1.6x in the review, as well as fixing up the block's cost from $91US to $85US (116 AUD x 0.735 = ~$85 US), and we'll call it even.

Man, the US dollar is falling in value so rapidly. Just 3 weeks ago the exchange rate was 0.68, and people who bought in the first batch were picking them up for ~$79US (plus shipping and payment overheads).

"Man, the US dollar is falling in value so rapidly."
well we have to print more to pay for the oil, no ?
/off topic

it's great if you're using foreign currency to pay off US debt though

sorry for the delay but USD cost and the change from 1.7x to 1.6x have both been made now. Cable modem died and was without internet most of the day. Got wine bottled though so the day was not a total loss.

Les,

Could you please add to your graph the pump pH uses & the MD20Z (60hz) as well? Thanks.

pH, I looked for your pumps flow/head rate figures in your methods link, but didn't see them.

May I suggest adding that info for those interested.

I don't think pH's pump info could be much other than misleading, too much unquantified hl stuff in the loop

Yes a very good point Bill, his pressure drop has to be much higher than "normal" (if there is such a thing) water loops. I was surprised by the reviews max flow of 1.7 gpm.

I'm trying to get a reading of how well my MD-20Z can hope to do.

Also getting a new rad that I hope will offer a improvement in performance over my 2-342 core. 2-192, will require some alteration of my rad box to allow for it's greater size.
Recent debate of how to get max performance & lower noise got me looking, I guess this is my answer to the thread regarding what peeps regard as a good low noise solution. With such a large increase in area, it will still cool at least as well with less noise (I Hope ;) ).

One other quesiton for Cathar,


Do you expect to see a higher flow rate with the Storm G5 as you did with the SS Cascade vs the "standard" Cascade?

Thanks

If it's at 60Hz (most likely is since you're in Michigan) with the MD-20RZ I would be expecting around bang on 2gpm with a Storm/G4 with an otherwise 1/2" tubed system + heater-core(s).

No. The Cascade SS was an optimised Cascade. The lessons learned with the SS vs Cascade were taken to the Storm design. The G5 also has a greatly more dense internal structure over the G4. You can tell which level the design is by counting the number of cups along the top row.

http://www.employees.org/~slf/storm/bpl.jpg

http://www.employees.org/~slf/storm/g5-03.jpg

Hmmm VERY, VERY NICE !

Thanks for the pic of the two bases along side one another. But study of the pic leaves me with more questions than it answered! :D

59 vs 35 jets but appear to be smaller jets in near the same area (deeper cups as well?). Isn't this 59 cup design even more cups than the SS Cascade had? G5 also shows a slight step in the cup walls, to reduce pressure drop at a point where most/all heat has been removed? Can you share any data yet regarding the G5's differances in jet velocity & flow rates? I recall you posting regarding a estimated temp differance of .8C, was this confirmed in testing?

A large number of optimisations indeed Cathar, great efforts man! :drool:

Thanks again for the pic & any info you can share! !

Digging for pics & other info from other threads.

I noted that the Storms jets have a much larger intake side vs outlet than did the Cascade. Sort of a venturi effect gained to help allow the higher velocitys?

The longer I study the Storm the greater the number of differances and details I see vs the Cascade. Quite a few differances from G4 to G5 as well..

Many interesting alterations Cathar. :D :drool: :D

35 vs 59 jets in the same area, yes. I do have a G6 design plans in hand with 89 jets, but it's really pushing what can be done with a basic CNC in the metal, and I need to talk to the machinists about if it can be made. The G6 would only ever be a one-off, much like the Cascade XXX was.

The jetted area is around 1.5x larger on the Storm design than on the Cascade. It's a rough 7/8" diameter circle effect, and the geometric level just adjusts the size of the cups to fit into that same sort of area.

Cup depth is not something I'm willing to discuss.

Cascade(/SS) had 52 cups in 2/3's the area of the storm design. On a cup density basis the Cascade is significantly more dense than even the Storm/G5.

I see what you're talking about in the pictures, but it's just a lighting illusion. There's a lot more clearance between the cups and tube walls than there was with the Cascade, also as a result of going to the round tubes, and so it wasn't necessary to contour the walls to facilitate water discharge.

Jet velocity per flow rate is almost identical. G5 has a very marginally higher pressure drop in comparison to the G4. Would typically expect around 5% lower flow rates when using a G5 over a G4. This is a natural result of shrinking the jets even smaller and the pressure drop increase to force the water into the smaller jets.

Across repeated mounts 0.8C is the least difference I have seen between the G4 and the G5 (best mount of G4 vs worst mount of G5). I have seen a significantly larger differences than that in my testing, but I'm keen to get the block reviewed by Phaestus to confirm. Phaestus does normally see around half the temp delta differences I see due to the nature of our respective testbeds and heatloads. Despite my pleasure with the results I have avoided making a fuss about them because they are unexpectedly good, and I'm wondering if I've done something wrong (even though I have been able to reproduce many times). This has also thrown a total spanner into anything I've ever written about being able to predict what the design is doing, and leaves me more open to getting the G6 made up, even if only to explore the temp differences at that level.

Thanks for all the info Cathar!

LOL, the fact that in this case your testing is showing what you expected is the GOOD news! !

Glad to see it going very well.

I'll be glad when you feel you can discuss all the details of your work.

iwaki curves at 50hz and 60hz can be found here (page 6 of 8) if you wanted to DIY it....
http://www.iwakipumps.jp/products_e/...f/md_w2003.pdf

Added Iwaki MD-20RZ(M)-N 60MHz (hope is right Model) and Eheim1046.

Well I ripped the Storm/G5 silver block out of my testbed, cleaned it up a little, and it's now on its way to Phaestus. If it arrives as fast as the Storm/G4 did, it should be within him next Tuesday.

That sounds painful

Oh btw I found the white balance on my camera. I assumed "auto" would be ok but it looks like if you are taking pics with no flash you have to manually set it to the type of lighting.

The things I do to get stuff through customs...