Unexpected results - higher flow = higher temps

[virtualrain]

I'm gonna look at the most logical explanation for this difference. First, how many days went by between test 1 and test 2? Was ANYTHING in the loop touched between first and second test? Was the CPU block remounted? Was tubing knocked? You see, if you even bump the tubing near the cpu block you could throw the block off and get different results - sounds crazy but true. I would bet that if you reran this test ON THE SAME DAY the results would be different. IDK why you have not done that - one test is meaningless without more tests to back it up.

Nothing was changed between tests. I ran the test at 12V on Sat. and I ran the test at 7V on Sunday. I started by leaving the computer idle all day while I was out running errands then I started the Orthos dual core test and noted the temps after 1hr. Same thing on Sunday except at 7V.

Incidentally, the reason I decided to do this more formal test is that last week I was experimenting less formally with the speed of the pumps and noticed that the temps in general (at idle or while gaming) were lower when the pumps were at 7V and higher when at 12V. Hence, I don't think there needs to be a retest... the observations I've made are consistent from day to day.

[Modzilla]

where's that beating the dead horse smiley?

[Hound53]

where's that beating the dead horse smiley?

dead horse or not I have gotten quite a bit of info from it. I realize it might have pushed a " which block is best" post off the 1st page but too bad.

[virtualrain]

dead horse or not I have gotten quite a bit of info from it. I realize it might have pushed a " which block is best" post off the 1st page but too bad.

LOL!!! I definitely appreciate all the input as well!

BTW, I found out from someone on the Aquacomputer forums that the Aquaero can only handle flow up to 3lpm... that explains why I get stupid readings beyond that. I'm not impressed!

[Radical_53]

But only if you use an old firmware version. The older Innovatek controllers had just the same problem.
Both of them are fine if you use the most actual firmware.

[virtualrain]

But only if you use an old firmware version. The older Innovatek controllers had just the same problem.
Both of them are fine if you use the most actual firmware.

I'm running the most current firmware... 4.10a.

[Radical_53]

Then there shoulnd't be any problems normally. How high did you have to set the pulse/liter settings?

[Tero]

Ok this is going to be a long post.

In response to Reinvented, Hound53, Fairydust, and Tero

As I will mention below, the heat dump into the water from the pump is mostly not from the impeller friction but rather the electronics and stator.

Is there any calculation you know offhand that will calculate that? I'm very interested in just how much heat impeller friction causes since all i've seen posted on it has said it causes some friction and it's the main cause of heat.

Thinking about it, it would be quite difficult to figure it out since every pump would have a different resistance coefficient based on the impeller and the housing the water is traveling through.

[Modzilla]

by asking about the dead horse smiley in my last post I only meant to imply that we have answered the original question posed in the thread, not to discount the value of everybody's post or the information contained in the thread.

I didnt say the main cause of the higher temps was friction, I said it was a cause of higher temps, and by listing 2 causes (pump input power and friction) and listing friction second, I meant to imply that the pump input power contributes more to the higher temps, and that friction contributes less to the higher temps.

[virtualrain]

Then there shoulnd't be any problems normally. How high did you have to set the pulse/liter settings?

The flow meter outputs 700 imp/l which is what I set the flow meter setting to on the Aquaero.

Here's a link to the Aqua Computer forum where the discussion is unfolding...
http://www.aqua-computer-systeme.de/e_forum-index.htm

[Marci]

Some references...

Pump Data: http://forums.procooling.com/vbb/showthread.php?t=10825 (pre-D5 / DDC+ hence ommitted)
CPU Block Performance vs Pumping (Hydraulic) Power: http://forums.procooling.com/vbb/sho...5&postcount=41
Radiator Analysis (Efficiency): http://www.swiftnets.com/Technical/A...erformance.pdf
Radiator Testing (the original baseline): http://thermal-management-testing.com/ThermoChill.htm


If monitoring flow with meters, ensure system is piped with tubing that's as close a match to the bore of the flow meter to minimise losses of that flow meter. Same goes for fittings. Suggest you repipe your system in 8mm ID 10mm OD tubing (5/16" ID) to make your flowmeter's results more "valid" without the pressure drop issues.

[Hound53]

thanks for the links Marci

[virtualrain]

Some references...

Pump Data: http://forums.procooling.com/vbb/showthread.php?t=10825 (pre-D5 / DDC+ hence ommitted)
CPU Block Performance vs Pumping (Hydraulic) Power: http://forums.procooling.com/vbb/sho...5&postcount=41
Radiator Analysis (Efficiency): http://www.swiftnets.com/Technical/A...erformance.pdf
Radiator Testing (the original baseline): http://thermal-management-testing.com/ThermoChill.htm


If monitoring flow with meters, ensure system is piped with tubing that's as close a match to the bore of the flow meter to minimise losses of that flow meter. Same goes for fittings. Suggest you repipe your system in 8mm ID 10mm OD tubing (5/16" ID) to make your flowmeter's results more "valid" without the pressure drop issues.

Given the problems I'm having with the flow meter providing accurate readings above 3lpm, its going to come out during my next loop rebuild.

[Serpentarius]

high flow cuts the time taken travelling from point waterblock to point rad .... well, maybe the rad is unable to dissipate the amount of heat effeciently before its recirculated back into the loop, ultimately back to the processer block

or maybe the rad is being repainted and the paint gets into the rad fins which is vital for dissipating heat?

[welshtom]

high flow cuts the time taken travelling from point waterblock to point rad .... well, maybe the rad is unable to dissipate the amount of heat effeciently before its recirculated back into the loop, ultimately back to the processer block

this has been covered already.
If the water goes twice as fast, it goes through the rad twice as often.

thus the overall time a water molecule spends in the radiator/block is the same, whatever the flowrate over a set time

[Xilikon]

Without considering it's a stupid thread ressurection...

But to add my grain of salt, if water get 2x more time in the rad, it also get 2x more time in the blocks collecting more heat so it even out.

[initialised]

Heat dissipation in the radiator is a function of water flow rate and air flow rate. If you increase the water flow beyond the capabilities of the fans then you wont transfer all the heat so dT (inlet temp-outlet temp) will increase.

[Jedda]

Heat dissipation in the radiator is a function of water flow rate and air flow rate. If you increase the water flow beyond the capabilities of the fans then you wont transfer all the heat so dT (inlet temp-outlet temp) will increase.

Do you want to look closely at what you just wrote?

[sbinh]

He watercooled PSU ???? More temp. cuz more power draws from PSU ?