Swiftech MCW6000 reviewed at O/Cers

pHaestus · 05-20-2004, 09:08 AM

Right here

I am waiting for someone to show proper comparison of the 3/8" vs. 1/2" MCW600x models over the full flow range. I wanna see how well the nozzle inlet does compared to the 3/8" inlet. (Bill??)

And before you ask: I don't have these wbs so "wait for pHaestus to do it" isn't going to be very rewarding

killernoodle · 05-20-2004, 09:59 AM

Seems like a good block to me. Looks like a good bargain.

Wildfrogman · 05-20-2004, 10:41 AM

Interesting block, though with the small barb~hose connection size its geared towards the lower flow group. Not that low flow is bad...small hoses,small pump, small rad in a tiny shuttle type box work pretty good actually from what I have seen. You cant stick an iwaki pump and a heatercore inside a sff pc

pHaestus · 05-20-2004, 10:44 AM

Quote:

though with the small barb~hose connection size its geared towards the lower flow group.

Not so. The MCW500x was far more geared to the low pressure pump crowd. This block looks to me like it was redesigned to take better advantage of the higher pressure offered from Swiftech's MCP600 pump. Having said that, it looks like the MCW6000 will perform better than the MCW5000 across the board:

http://www.swiftnets.com/products/mcw6000.asp

Ruiner · 05-20-2004, 10:58 AM

It looks like the lower end of the barb of the 6002 is tapered anyway (look at the cross sectional diagram). The only flow difference compared to the 6000 may be in outflow then.

As an aside, how do you think the eheim 1048 would handle this block?
It does look like this block was tailored for the mcp600, an expected and smart move on swiftech's part.

pHaestus · 05-20-2004, 11:15 AM

Looking at the graphs on the Swiftech site, head loss is pretty similar to the MCW5002 at 0.3, 0.5, and 1.0 GPM flow rates (first 3 points). So I'd say the MCW6000 would be fine w/ 1048. A good match in fact for those wanting a quiet system. If you are performance minded, though, it seems like there is some additional performance from using the higher pressure pumps.

BillA · 05-20-2004, 01:11 PM

a spot of confusion presently re the head loss value
JoeC's is ~30% higher than our in-house value (for which we have many measurements)
- it is being investigated

what could cause a wb to have a higher head loss ?
blockage, or . . . . . ?

will update

the MCW6002 has a tapered inlet; exactly the same C/W, slightly lower head loss
therefore overall slightly better, but the 1/2" barbs are for the convience of those with 1/2" ID tubing

Wildfrogman · 05-20-2004, 01:54 PM

wildfrogman Quote:
though with the small barb~hose connection size its geared towards the lower flow group.

Quote:

Originally Posted by pHaestus

Not so. The MCW500x was far more geared to the low pressure pump crowd. This block looks to me like it was redesigned to take better advantage of the higher pressure offered from Swiftech's MCP600 pump. Having said that, it looks like the MCW6000 will perform better than the MCW5000 across the board:

http://www.swiftnets.com/products/mcw6000.asp

Thanks for catching me on that, I was thinking higher flow rate restriction from smaller barbs, smaller hose, pin fin=lower flow rate. Not lower pressure exactly but just would perform very good even at low flowrates.

Interesting about Joe's differing test results for the headloss.. , thanks for the heads up.

AngryAlpaca · 05-20-2004, 05:21 PM

That is a HIGH pressure drop. Performance seems predictable, but with a pressure drop that big (I believe you, Bill, that it is lower) it would seem the MCW-5002 would be superior.

BillA · 05-20-2004, 05:33 PM

Quote:

Originally Posted by AngryAlpaca

That is a HIGH pressure drop. Performance seems predictable, but with a pressure drop that big (I believe you, Bill, that it is lower) it would seem the MCW-5002 would be superior.

the 2 wbs are compared on the site, no need to surmise

AngryAlpaca · 05-20-2004, 07:48 PM

It doesn't show flows above 1GPM. Most pumps can push 1GPM through 34" of resistance, so it's largely irrelevant. At 1 GPM the 6000 beats the 5000 and 5002.

BillA · 05-20-2004, 07:52 PM

www.swiftnets.com
not OCers, sheesh

agiacobbi · 05-20-2004, 09:25 PM

Kinda off topic but does any one know why Swiftech's site all of a sudden says "distribution only" for the 6002 series blocks. BillA care to answer.

deathBOB · 05-20-2004, 10:33 PM

Block looks really nice... Nice performance for its price... Makes me think about a 3/8th system more and more...

Cathar · 05-20-2004, 11:29 PM

Water blocks, for a while there, were extremely low restriction.

We really only need around 0.75gpm (~3LPM) flow rates to do decent water cooling in the sense that the thermal resistance due to the volume of the water is not a significant factor in the level of cooling for which the water block can offer. The C/W of 0.75GPM of water is almost bang on 0.005, or low enough to be insignificant. Doubling the flow rate to 6LPM only yields a 0.0025 C/W improvement for the water mass, or about 0.25C for a 100W CPU.

So if we set our sights on 3LPM as being a minimum desirable flow rate, then this means that 3/8" ID tubing can make quite a lot of sense. 1 meters of 3/8" ID tubing at 3LPM provides around 9cm of pressure drop.

I'd like to make a block that peaks out at around 4LPM with an MCP600 attached to it. Only problem is with the public perception. It seems that the public needs to be eased into such an understanding. One of the prototypes I designed (on paper) did just that - as in it was restrictive enough that it would cause an Eheim 1250 to only run at slightly above 0.75GPM. Problem is that people would freak out at a block that had ~4PSI of pressure drop at 1GPM. I can imagine JoeC's comments at OC.com even now - "Absurdly high pressure drop block requiring a monster pump!" - when really all the block needs is just your every day pump that everyone uses to achieve the cooling purpose for which it was designed.

I find JoeC's comments somewhat contradictory. He comments on the need for powerful pumps for blocks that will happily run at the 1GPM that he tests at when driven with even quite weak pumps.

I believe that there is a place for 3/8" ID tubing - it's just that the market needs to be eased back into that understanding. There is an obsession with monster flow rates, because the very low pressure drop blocks often required those flow rates to perform well. The high flow rates (>5LPM) necessitated the need for 1/2" ID tubing, and it's not that 1/2" ID tubing is needed to push the flow rates that are required for adequate cooling at a fundamental level.

[Edit: fixed units in one location]

AngryAlpaca · 05-22-2004, 09:00 PM

Quote:

I find JoeC's comments somewhat contradictory. He comments on the need for powerful pumps for blocks that will happily run at the 1GPM that he tests at when driven with even quite weak pumps.

I concur. Correcting his comments at the end of his article is becoming a daily event...

unregistered: I was thinking you meant that, but some of your results have begun to surprise me. It started here:

Quote:

It can bee seen in the above graph that the flow resistance of the MCW5002™ is higher than the earlier version of the MCW5000™, due to the greater flow being forced through the Diamond Pin Matrix ©. This has the effect of increasing the flow velocity and improve heat transfer.

The pressure drop does seem to actually be LOWER than the 5000, according to Joe's setup, and common sense. Also, there's that 34" drop on Joe's testbed, compared to your 20". Maybe Joe's testbed has gone wacky... I feel somewhat obligated, as a consumer, to not trust the results of the corporation that made the product.

pHaestus · 05-22-2004, 09:08 PM

I am certain that Bill has accounted for the pressure drop of the fittings and pvc crosses he uses to plumb in the pressure transmitter. I would guess that JoeC has not (hence across the board higher pressure drops).

snowwie · 05-23-2004, 02:58 PM

Quote:

Originally Posted by AngryAlpaca

I concur. Correcting his comments at the end of his article is becoming a daily event...

unregistered: I was thinking you meant that, but some of your results have begun to surprise me. It started here: The pressure drop does seem to actually be LOWER than the 5000, according to Joe's setup, and common sense. Also, there's that 34" drop on Joe's testbed, compared to your 20". Maybe Joe's testbed has gone wacky... I feel somewhat obligated, as a consumer, to not trust the results of the corporation that made the product.

in that quote he is probably referring to the first version of the mcw5000, where the top alu piece has the extra space for water to flow, contrary to the mcw5000 v2 and mcw5002 (which differ only by the fittings, kinda like the 6000 and 6002), which eliminate that "space", high flow velocity etc.

i think joeC's comparison (and the one you see in your head) is between the second mcw5000 and the mcw5002

BillA · 05-23-2004, 03:27 PM

correct . .

05-20-2004, 09:08 AM	#1
pHaestus Big Player Making Big Money Join Date: Aug 2001 Location: irc.lostgeek.com #procooling.com Posts: 4,782	Swiftech MCW6000 reviewed at O/Cers Right here I am waiting for someone to show proper comparison of the 3/8" vs. 1/2" MCW600x models over the full flow range. I wanna see how well the nozzle inlet does compared to the 3/8" inlet. (Bill??) And before you ask: I don't have these wbs so "wait for pHaestus to do it" isn't going to be very rewarding __________________ Getting paid like a biker with the best crank... -MF DOOM

05-20-2004, 09:59 AM	#2
killernoodle Thermophile Join Date: Sep 2002 Location: Charlotte, NC Posts: 1,014	Seems like a good block to me. Looks like a good bargain. __________________ I have a nice computer.

05-20-2004, 11:15 AM	#6
pHaestus Big Player Making Big Money Join Date: Aug 2001 Location: irc.lostgeek.com #procooling.com Posts: 4,782	Looking at the graphs on the Swiftech site, head loss is pretty similar to the MCW5002 at 0.3, 0.5, and 1.0 GPM flow rates (first 3 points). So I'd say the MCW6000 would be fine w/ 1048. A good match in fact for those wanting a quiet system. If you are performance minded, though, it seems like there is some additional performance from using the higher pressure pumps. __________________ Getting paid like a biker with the best crank... -MF DOOM

05-20-2004, 11:29 PM	#15
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Water blocks, for a while there, were extremely low restriction. We really only need around 0.75gpm (~3LPM) flow rates to do decent water cooling in the sense that the thermal resistance due to the volume of the water is not a significant factor in the level of cooling for which the water block can offer. The C/W of 0.75GPM of water is almost bang on 0.005, or low enough to be insignificant. Doubling the flow rate to 6LPM only yields a 0.0025 C/W improvement for the water mass, or about 0.25C for a 100W CPU. So if we set our sights on 3LPM as being a minimum desirable flow rate, then this means that 3/8" ID tubing can make quite a lot of sense. 1 meters of 3/8" ID tubing at 3LPM provides around 9cm of pressure drop. I'd like to make a block that peaks out at around 4LPM with an MCP600 attached to it. Only problem is with the public perception. It seems that the public needs to be eased into such an understanding. One of the prototypes I designed (on paper) did just that - as in it was restrictive enough that it would cause an Eheim 1250 to only run at slightly above 0.75GPM. Problem is that people would freak out at a block that had ~4PSI of pressure drop at 1GPM. I can imagine JoeC's comments at OC.com even now - "Absurdly high pressure drop block requiring a monster pump!" - when really all the block needs is just your every day pump that everyone uses to achieve the cooling purpose for which it was designed. I find JoeC's comments somewhat contradictory. He comments on the need for powerful pumps for blocks that will happily run at the 1GPM that he tests at when driven with even quite weak pumps. I believe that there is a place for 3/8" ID tubing - it's just that the market needs to be eased back into that understanding. There is an obsession with monster flow rates, because the very low pressure drop blocks often required those flow rates to perform well. The high flow rates (>5LPM) necessitated the need for 1/2" ID tubing, and it's not that 1/2" ID tubing is needed to push the flow rates that are required for adequate cooling at a fundamental level. [Edit: fixed units in one location] Last edited by Cathar; 05-22-2004 at 08:27 AM.

05-20-2004, 10:41 AM	#3
Wildfrogman Cooling Neophyte Join Date: Jan 2004 Location: USA Posts: 31	Interesting block, though with the small barb~hose connection size its geared towards the lower flow group. Not that low flow is bad...small hoses,small pump, small rad in a tiny shuttle type box work pretty good actually from what I have seen. You cant stick an iwaki pump and a heatercore inside a sff pc

05-20-2004, 10:58 AM	#5
Ruiner Cooling Savant Join Date: May 2003 Location: New Smyrna, FL Posts: 258	It looks like the lower end of the barb of the 6002 is tapered anyway (look at the cross sectional diagram). The only flow difference compared to the 6000 may be in outflow then. As an aside, how do you think the eheim 1048 would handle this block? It does look like this block was tailored for the mcp600, an expected and smart move on swiftech's part.

05-20-2004, 01:11 PM	#7
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	a spot of confusion presently re the head loss value JoeC's is ~30% higher than our in-house value (for which we have many measurements) - it is being investigated what could cause a wb to have a higher head loss ? blockage, or . . . . . ? will update the MCW6002 has a tapered inlet; exactly the same C/W, slightly lower head loss therefore overall slightly better, but the 1/2" barbs are for the convience of those with 1/2" ID tubing

05-20-2004, 05:21 PM	#9
AngryAlpaca Cooling Savant Join Date: Jan 2004 Location: Alberta Posts: 631	That is a HIGH pressure drop. Performance seems predictable, but with a pressure drop that big (I believe you, Bill, that it is lower) it would seem the MCW-5002 would be superior.

05-20-2004, 07:48 PM	#11
AngryAlpaca Cooling Savant Join Date: Jan 2004 Location: Alberta Posts: 631	It doesn't show flows above 1GPM. Most pumps can push 1GPM through 34" of resistance, so it's largely irrelevant. At 1 GPM the 6000 beats the 5000 and 5002.

05-20-2004, 07:52 PM	#12
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	www.swiftnets.com not OCers, sheesh

05-20-2004, 09:25 PM	#13
agiacobbi Cooling Neophyte Join Date: Apr 2004 Location: TN, USA Posts: 19	Kinda off topic but does any one know why Swiftech's site all of a sudden says "distribution only" for the 6002 series blocks. BillA care to answer.

05-20-2004, 10:33 PM	#14
deathBOB Cooling Neophyte Join Date: Jan 2004 Location: Atlanta, GA Posts: 60	Block looks really nice... Nice performance for its price... Makes me think about a 3/8th system more and more...

05-22-2004, 09:08 PM	#17
pHaestus Big Player Making Big Money Join Date: Aug 2001 Location: irc.lostgeek.com #procooling.com Posts: 4,782	I am certain that Bill has accounted for the pressure drop of the fittings and pvc crosses he uses to plumb in the pressure transmitter. I would guess that JoeC has not (hence across the board higher pressure drops).

05-23-2004, 03:27 PM	#19
BillA CoolingWorks Tech Guy Formerly "Unregistered" Join Date: Dec 2000 Location: Posts: 2,371.493,106 Posts: 4,440	correct . .

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)