New Swiftech H20-X20 Edge cooling kits

[gabe]

Now that's something truly unique and promising for a long time.

Just a few questions if you don't mind?

1. What kind of flow rate can we expect at the lowest pump speed for both 220 and 320 kits?


2. Can the radiator and pump operate vertically when mounted sideway like the MCR320 in the above render?

3. Perhaps powering the pump from a 4-pin molex, while routing PWM and RPM signals to the mobo header would be a more conservative approach? After all, a MCP355 draws up to 18W peak power.

4. Are the fill ports using standard thread sizes?

Phil

1. FLOW RATES: Sorry, this hasn't been done. email stephen@swiftech.com if you need more information

2. ORIENTATIONS: the pic below should answer your question. You can also download the install guide here



3. this is what were' doing.

4. yes, G1/4

hmmmmmmmmmmmmm does this 2 kits compete with H70 / ALC ???

no; these products are in a different league

3. MCP350 and MCP355 already have their power wires on a 4-pin molex connector and the tach wire on a separate 3-pin connector. They probably replaced the 3-pin by a PWM style 4-pin and kept the PSU style 4-pin molex for the power wires.

4. I think Swiftech said they were switching all their PG11 fill port to G1/4, probably the case here.

correct on both counts

Philwong, in regards to question number 1, at 1300RPM (0% PWM duty, a.k.a., minimum speed), the MPC35X/DDC3.25 should provide less pumping power than a D5 at setting 1.

Since we haven't done the curve of the 35X at minimum speed, it is really impossible to provide a scientifically correct answer. Let's wait for the curves before we speculate.

[Philwong]

Gabe, why not install another fill port for scenario *3 on top of the rez under that orientation?

Phil

[Vapor]

Since we haven't done the curve of the 35X at minimum speed, it is really impossible to provide a scientifically correct answer. Let's wait for the curves before we speculate.

Pressure scales with square of the RPM and flow scales linearly with RPM. Get a curve at any RPM and you can calculate a curve at any other RPM and it'll be really, really close to actual

D5 P5 (4800RPM): max head in your MCP655 tests is ~4.05 mH2O, max flow is ~20.7 L/min

D5 P1 (1800RPM): max head in your MCP655 tests is ~.58 mH2O, max flow is ~8.8 L/min

Theoretical D5 P1 (1800RPM) based on scaling from above: max head is ~.57 mH2O, max flow is ~7.8 L/min. Factor in the D5 lowers its RPMs a bit at high amp draw/flow at P5+12V (so that the 20.7 L/min actually occurred at ~4300RPM), and the max flow works out just fine. This scaling works on all fans/pumps and across the entire curve, not just the ends (math is one step easier at the ends though )

So taking that and skinnee's DDC3.2+EK V2 data, we can pretty accurately estimate the performance of a 1300RPM DDC based on the DDC3.2 being ~4300RPM.

4300RPM DDC + EK V2 top has a max head of 4.03 mH2O, a max measured flow of 14.2 L/min (at 1.96 mH2O TDH, so not literal max flow due to testbed limitations), and a max flow (via trendline ) of 18 L/min

1300RPM DDC + EK V2 top works out to be a max head of .37 mH2O, a max flow of 5.44 L/min (corresponding to the trendline max) and a point on the curve at 4.3 L/min and .18 mH20.

Max head (with one of the best tops on the market) is 35% lower than a D5 P1 and max flow is 51% lower than a D5 P1. Even with a 5% difference in testbed with your testbed vs. skinnee's, a DDC at 1300RPM is going to deliver flowrates some 30%+ lower than a D5 P1.

This isn't a knock on the MCP35X at all, I love everything about it--warranty on a DDC3.25 with top-notch performing top (I trust your products perform) and with PWM wiring on top of that is awesome. But at the lowest RPM, it will be extremely weak.

EDIT: just to come full circle to Philwong's question, I'll quote myself:

It'll work, but if you increase the speed, temperatures will improve.