BIP.1 to BIP.2 ... not impressed

[DVL73]

Hi,

I just replaced my BIP.1 with BIP.2 radiator. However, not impressed
with the temps improvement ... maybe 1-3C less than before. I'm
using the same fans as before (SilenX 120). I added the following parts
in the the water cooling loop:

Mcubed flow meter + water temperature probe
http://www.mcubed-tech.com/eng/produkt_tban_wasser.htm

ThermoChill ThermoTube
http://www.thermochill.com/thermotube.php

Does anyone knows in reality how much temps improvement 120.2
radiators can offer ? (comparing to 120.1) Is something wrong
in my loop ?

I'm using the Swiftech H20-120 Rev.3. Generally, I found that reservoir
from this kit is noisy (loads of water turbulence and flow noise). New
ThermoTube reservoir is working like charm.

My system specs are
P4 3.0 @ 4.0 , ABIT IC7-MAX3, 6800GT @ 435/1060

Temps: IDLE 45C - 46C LOAD 49C-51C (ABIT mobo reports higher temps
than other ones)

Just noticed that Thermochill is offering now the PA160 ?
http://www.thermochill.com/pa160.php
Anyone tried this one ?

[DVL73]

I forgot .... according to water temp probe, water temperature is in
the range of 29C - 31C. Ambient temperature is 23C - 24C, but because
my case is in the corner of the room (not much cool air around) and
radiator is mounted externally via radbox I would say that real ambient
temperature is maybe few degrees higher.

[bobkoure]

Are you cooling just the CPU?
Did you dismount/remount your wbs in this change? Which wb(s) are you using?
You mention the mcubed flow meter and water temp probes - but don't provide any data from it. Why rely on mboard-provided temps? Comparing the before and after flow numbers might tell you if you now have a restriction.

All that said/asked - you might find it interesting to replace your silenx fans with ones that push lots more air - just to see what happens to your temps. You should be able to do this without disturbing the rest of your system. Not seeing much change in temps would indicate that your "weakest link" is somewhere else - or that there's something wrong with your temp measurement methodology.

[JSimmons]

The PA160 was made available a day or so ago, and allsigns point to a general lack of availability anywhere except the UK for a week or two or three.

It's 8-1/2 x 6-1/2 x 1 inches, so it's kind of a weird size form most PC cases, but that doesn't keep me from wanting one.

[DVL73]

Originally Posted by bobkoure

Are you cooling just the CPU?

Nope. I'm colling the CPU + GPU

Originally Posted by bobkoure

Did you dismount/remount your wbs in this change? Which wb(s) are you using?

I didn't have to dismount wbs. Becuse the radiator is mounted externally
I just swap them (unplug the hoses from the old one and plug the new one).
Of course, I flushed the system before mounting and filled again with the
fresh mixture of distilled water/swiftech HydrX.

Radiator is placed as pictured here: (so that will give you idea about the
placement, I just replaced the single BIP with the dual version):

http://www.swiftnets.com/products/MCB-120.asp

CPU wb is Swiftech MCW6000 and GPU wb is Swiftech MCW50

Originally Posted by bobkoure

You mention the mcubed flow meter and water temp probes - but don't provide any data from it. Why rely on mboard-provided temps? Comparing the before and after flow numbers might tell you if you now have a restriction.

Water temp I gave you is actually from the water temp probe. Also, flow
meter is telling me that flow rate is 3.2l/min. Unfortunately, I didn't have
the flow rate measurement before (didn't have the flowmeter).

Apart from the radiator, I disconnected the old reservoir and installed
the ThermoTube. Also, I placed the water temp probe via T-Line on
one reservoir port (which is going to the Swiftech MCP-350 pump)
and attached in-line the flow meter on the second reservoir port.
So ... loop is like this:

ThermoTube (res) -> T-Line water probe -> MCP-350 pump
->GPU->BIP II radiator->CPU->flow meter->ThermoTube (res)

Is flow meter maybe responsible for the restriction ? I verified
the radiator ... no air or bubbles inside (warm equally) and
no visible bubbles in the water also .. all crystal clear.

Originally Posted by bobkoure

All that said/asked - you might find it interesting to replace your silenx fans with ones that push lots more air - just to see what happens to your temps. You should be able to do this without disturbing the rest of your system. Not seeing much change in temps would indicate that your "weakest link" is somewhere else - or that there's something wrong with your temp measurement methodology.

I tried the much powerfull Papst/Delta fans ... but that's too much noise for me.
That's the main reason why I changed the radiator to BIP II. It has thinner core
and doesn't require the powerful fan in order to penetrate the core properly.
CPU temperatures are measured from the motherboard sensors, so don't
think that something is actually wrong with the readings.

[Ruiner]

Since it would be easy to test, try removing the rad from the rad box, and see if your dT water/air changes. Perhaps airflow is limited. How well does the bip2 fit on the radbox?

[DVL73]

Originally Posted by Ruiner

Since it would be easy to test, try removing the rad from the rad box, and see if your dT water/air changes. Perhaps airflow is limited. How well does the bip2 fit on the radbox?

BIP2 fits just perfect in the radbox. It's attached via upper fan in the radbox.
There is enough spacing behind the fans for the decent airflow. However, because my PC is in the desk corner of the room all the heat from the PSU and back of the case is accumulating in that corner, and ambient temperature is higher on that spot ... but this was also the case with the previous setup.

Puzzle is how come that CPU/GPU temperatures didn't change so much with
this setup. BIP2 has larger surface area for heat dissipation.

[bobkoure]

Originally Posted by DVL73

I tried the much powerfull Papst/Delta fans ... but that's too much noise for me.
That's the main reason why I changed the radiator to BIP II. It has thinner core
and doesn't require the powerful fan in order to penetrate the core properly.
CPU temperatures are measured from the motherboard sensors, so don't
think that something is actually wrong with the readings.

The point of the more powerful fans isn't to leave them there but to see if your temps drop (your radiator will exchange more heat with more air moving through it). If they do not, look somewhere else for a problem.
From your comments, it looks as though you've changed from BIX.1 to BIP.2? Were you using the same fan type in both situations?

Even with low restriction cores like the BIPs, when using very-underpowered fans (like Panaflos at 5V) I've had better luck pulling air through than pushing it.
Motherboard readings are laughably bad and non-linear - one of the reasons folks buy mcubed gear is to get a better idea of their real temps (the rest of us guess - but we know we only have some connection to reality )

[DVL73]

Originally Posted by bobkoure

The point of the more powerful fans isn't to leave them there but to see if your temps drop (your radiator will exchange more heat with more air moving through it). If they do not, look somewhere else for a problem.

I tried to use the much powerful (Delta) fan in previous setup (BIP.1)
and temperature delta was not more than 2C ... so really not worth
the extra noise.

Originally Posted by bobkoure

From your comments, it looks as though you've changed from BIX.1 to BIP.2? Were you using the same fan type in both situations?

No no ... maybe you misunderstood me anyway I changed from
BIP.1 to BIP.2 (never had BIX)

Yes, I'm using the same fans (SilenX 120) in both setups.

Originally Posted by bobkoure

Even with low restriction cores like the BIPs, when using very-underpowered fans (like Panaflos at 5V) I've had better luck pulling air through than pushing it.

Hmmm .... I may try to position the fans in the pull, instead of push
configuration. Also, it may happen that with the current (push) setup
all actual heat from the back of the PSU (and case) is sucked directly
to the cure. Pull configuration would be maybe more benefical.

Originally Posted by bobkoure

Motherboard readings are laughably bad and non-linear - one of the reasons folks buy mcubed gear is to get a better idea of their real temps (the rest of us guess - but we know we only have some connection to reality )

I also attached the mcubed thin foil probe (under CPU) and it's showing
approx 7.5C less then motherboard sensor (so ABIT temperature reading
is really, definitely messed up).

Of course, problem is that I didn't have the mcubed in the previous setup
so can't compare the temps via mcubed (had to use MB sensor).

[snowwie]

i don't know if this has been mentioned yet, but it's no wonder that the addition of a flowmeter/temp probe in the loop would likely hurt performance.

[nikhsub1]

Originally Posted by snowwie

i don't know if this has been mentioned yet, but it's no wonder that the addition of a flowmeter/temp probe in the loop would likely hurt performance.

remove the flowmeter

[MadHacker]

Originally Posted by nikhsub1

remove the flowmeter

I put a flow switch into my system and it hurt my performance real bad...
I removed it within 24 Hours.
it was a Paddle type
http://www.gemssensors.com/pdf/IOM_B...s/RFS-2500.pdf
a pitty cause I bought 2 of them and now I don't have a use for them...

[maxSaleen]

I'll second the concept of mounting a large radiator via radbox. I have a BIP 3 mounted via radbox on the back of my system. PITA to route the cables. Try mounting the radiator so that no part of it is in line with the exhaust from the PSU. Pull configuration should help, but probably only 1c or so (especially if deltas only helped 3c). I doubt the flowmeter would drage temps down that much. Still might try remvoing it to see what you get for temps.

Mobo readings are wack. That's why I've never posted temps. I have a caterwood based epox in one of my systems cooled via an mcx 478V and a 92cm fan good for about 55cfm. Airflow through the case is solid. Processor is a 2.6C at 3.2. Guess what my mobo reports as temps? Guess. Come on Guess.

92C! LOL! How's that for accuracy? If I get time I'll post a screen shot.

[maxSaleen]

Just took a look at that flow meter. Major flow killer. I say pull it and watch your temps drop.

[Ruiner]

Fans 'pulling' is also much quieter than 'pushing'.

[bigben2k]

The water temp is 'allegedly' 6 or 7 degrees above ambient; I say it's pretty much maxed out there. What's left, as everyone else pointed out, is to increase your coolant flow, to reduce the only thermal layer that you can affect; the one inside the block.

Drop the res and meter, see what happens.