Sign in with
Sign up | Sign in
Your question

Help - monitoring water temperature

Last response: in Overclocking
Share
July 4, 2006 3:25:39 PM

I'm planning on getting a Matrix Orbital for one thing, to measure
water temperature of a new rig I'm building.

Questions: I was going to place temp probes inside the tubing
before and after the radiator. I want to do this, because it's neat to
know how the system is behaving, plus the matrix orbital will be
controlling the fans on the radiator and the case.

But I thought I read somewhere that the water temperature will be constant throughout the loop when it reaches temp equilbrium.
So both probes will show identical temps. I'm trying to see where I
read that. Is this true? Who out there uses external probes for measuring water temps?

Another question: I have to use the probes that the matrix orbital
comes with. Meaning, I have to drill a small hole into the tubing,
heat wrap the temp probe, add glue to the probe and insert it slightly
into the tube (1mm) and then wrap some of the probes wires around
the tubing, to take stress off the probe, glue those same wires.
My system will be using the XPflud+ extreme which is non-conductive.
What kind of glue would you recommend?

Question: Where else would you place external probes?
Some useful information would be:
- Room temperature
- Case temperature - where?

My rig will have 6 - 120 mm Scythe Flex Fans (1600 rpm).
Three for the Thermochill PA 120.3 modded to the top of the case,
one intake, one exhaust, and one on the side near the graphics board.

All 6 will be monitored and controlled by the matrix orbital. So I will
be monitoring temperatures at a given load, to see how the fans can
be controlled for maximum effectness at lower noise levels.

The loop will be Reservoir -> Pump -> radiator -> cpu -> GPU (crossfire)
-> Reservoir

Pump - Swiftech MCP-655
Reservoir - Swiftech Micro
Tubing - 1/2" ID Tygon
CPU block - Swiftech storm revision 2
GPU block - Danger Den Tyee - crossfire

The case is a modified Lian Li PC-G70B
July 5, 2006 2:53:52 AM

bump
please help!
July 5, 2006 3:32:53 AM

have you thought about putting a temp sensor in your resevor instead of drilling one into your tubing?

i think it would give you a better idea of how well you rad is working if you had 1 temp sensor in the resevor and maybe one after the radiator.

just an idea tho.
Related resources
July 5, 2006 3:34:05 AM

In skimming through your article, You might want to put some probes on the exit from your waterpump (wrapped w/ electrical tape), from your reservoir *why do u have a reservoir* that sensor should tell you the cumulative temperature... i have a sensor in my radiator (in the fins)...


drilling holes in your tubing is the WORST thing you can do. W/ the pressure and temperature, the tubing will expand, and you're inviting a leak. I doubt any glue would help you. See if you can get the sensor by one of the barbs, w/ the tubing sliding over it... but u're more likely to break the wiring of the sensor.. then again, you're inviting a leak.

As for glue... use silicone glue... its about $3 @ your local hardware store... and get teflon tape, it doesn't hurt.
July 5, 2006 4:15:27 AM

Quote:
have you thought about putting a temp sensor in your resevor instead of drilling one into your tubing?

i think it would give you a better idea of how well you rad is working if you had 1 temp sensor in the resevor and maybe one after the radiator.

just an idea tho.


Putting it in the reservoir sounds like a good idea.
July 5, 2006 4:21:11 AM

Quote:
In skimming through your article, You might want to put some probes on the exit from your waterpump (wrapped w/ electrical tape), from your reservoir *why do u have a reservoir* that sensor should tell you the cumulative temperature... i have a sensor in my radiator (in the fins)...


drilling holes in your tubing is the WORST thing you can do. W/ the pressure and temperature, the tubing will expand, and you're inviting a leak. I doubt any glue would help you. See if you can get the sensor by one of the barbs, w/ the tubing sliding over it... but u're more likely to break the wiring of the sensor.. then again, you're inviting a leak.

As for glue... use silicone glue... its about $3 @ your local hardware store... and get teflon tape, it doesn't hurt.



The putting the hole in the tubing, came from the overclockers website
that reviewed the matrix orbital (I contacted the reviewer of the
article) Also, the matrix orbital website, mentioned the same idea.
July 5, 2006 4:21:20 AM

put a doc's style thermo into your res!

A small docs style thermo is short and filled with mercury (this is the most accurate tool to measure temps and is used to calibrate lab equipment) best of all it will fit into a res (my res is the size of a CD rom bay)... if you have a swiftec micro res its just a tad too small....

The temp of the water is differnt in differnt places!
eg: the water temp in the tube just after the cpu is hotter (obvious)
The water going into a rad is hotter and comes out cool (Well I hope it does with 6 fans and a Black ice extreme III)

You have confused pressure and flow rate these are constant in your tubes not the temps!

Dont put sensors into your tubes it will just restrict flow and decrease prefromance! TT makes sensors for water cooling and they just suck

Dont cut holes into your tubes its very difficult to make it water tight again even with silicone.... silicone once dry is not as flexable as your tygon tubes so it will just crack and leek!

your swiftec pump emits its estimated to be 15 wat per hour equates to less than one degre per hour..... a little bit of heat into the water from the pump is nuffink to worry about and cant be helped... The only pump that emits next to no heat is enhim pumps (used for fish tanks with sensitive fish) these pumps are piss week amd massive in size.

Whats the difff between rev2 storm and original storm????
July 5, 2006 1:23:13 PM

Quote:

Whats the difff between rev2 storm and original storm????



Revision 2 of the Storm water-block features incrementally improved flow rates thanks to redesigned outlet ports.



Revision 2 of the Storm water-block is shipping effective 2-17-06

The revision consists in additional ports machined in the middle and upper plates and converging via channels into the main outlet port. This revision incrementally improves flow rate (reduces the pressure drop) of the water-block.

I understand about water in cpu block heats up water - Rad cools it
down - I just read somewhere in a forum (of course - that doesn't
make it true ) - about equillibium occurring - where temps were equal -
July 5, 2006 9:54:02 PM

Quote:


The temp of the water is differnt in differnt places!
eg: the water temp in the tube just after the cpu is hotter (obvious)
The water going into a rad is hotter and comes out cool

You have confused pressure and flow rate these are constant in your tubes not the temps!



Here's some information from the overclockers forum:
thread1:
It's not a matter of opinion, it's a matter of fluid dynamics. On a single pass through a heat source, the cooling fluid only absorbs a small amount of energy per molecule. This is why the cooling is dependent on the volume flow rate and not the smount of fluid in the cooling system. The radiator only removes a small amount of energy per water molecule. The water temperature will come to an equilibrium and be roughly equivalent before and after the rad. Roughly by a factor of <1C.

thread2:
Now when you talk about water loops you have to understand that the water is moving rapidly over the blocks. So rapidly that water is unable to absorb high amounts of joules (heat/energy) because it does not touch the block long enough but this is compensated by having LOTS of water pass through the block. An example:
A block outputs 100 joules a second, 100 water molecules pass through the block in one second.
1 molecule absorbs 1 joule.
1 joule barely even raises the temperature of the water. Therefore, you can conclude that the water will not raise in temperature drastically in the loop.


Measuring water temperature:

I'm getting the swiftech Hi flow MICRO reservoir - I can drill a hole
in the top. place the probe in the reservoir, and seal up the hole.

The reservoir top is a fill-port - and normally is not filled to the top.
Meaning, no leaks.

From what I've been reading, having multiple probes in the water
at different points - won't accomplish much. At least, that's what
I'm planning to do first.
July 6, 2006 2:47:52 AM

whoa the physics thats interesting......

see if you can prove the oc fourm wrong!

micro res is an nice pice of work..... be careful!

do tell us what you find out
July 6, 2006 3:51:06 AM

Quote:
whoa the physics thats interesting......

see if you can prove the oc fourm wrong!

micro res is an nice pice of work..... be careful!

do tell us what you find out


Yes, the physics is neat. There's a lot going on, and there's
always a conflict of opinions. For example, the direction of
airflow through the radiator. I heard good arguments for both
cases. In my search through the OC forum, everyone seems to
agree that after a system has been on for awhile (no specific time
given) the temperature throughout the loop is almost constant
< 1C at a given constant load.
July 7, 2006 2:44:11 AM

I think the push pull system works well for rads 1 fan on both sides of the rad..... (a noizy solution tho)

I find that I dont have to switch on my rad fans on (takes ages for about 2 L of water to get hot).

The reason why nobody can tell you when to switch on the rad fans or how long it takes for the water to start heating up is because there are soo many variables!

eg: hotter chips including how much you have oced your CPU etc

the room temp

the size of your rad (bigger ones will disipate heat even without a fan)
July 7, 2006 6:08:14 AM

I'm buying a modified Lian Li case, that has a PA 120.3 thermochill on
top. The fans are blowing air from the case to the outside. Stating that
the air inside the case shouldn't get the hot, since the CPU and GPU are
on water blocks. They also quoted that pulling air in, there's a dead space
caused by the grill (I think), I also read that exhausting more, causes a
negative pressure which also helps temps. There's a lot of theory going
on here. My plan is to leave it the way it comes. Enjoy the computer and
finish playing HL2, this way I know my new build is stable. Run the
benchmark tests so I have a good baseline. The overclockers forum
is loaded with questions on how to measure good and accurate temps.
I said questions.... Not many answers that everyone agrees with.
Except to say, to treat the temps in relative terms. Meaning, are they
going up or down. For example, when I play with the fan controls to
make the machine more quiet. Or if I start playing with the fans, as
far as if they should be pulling or pushing the air. The external temp
probes will be telling me my water temperature (good for controlling
rad fans). The ambient room temperature, and differnt locations in
case to see how my air circulation is ( top of case feeding the rads -
since temperature will dictate how efficient my rad works) Maybe, the
exhaust. I don't know, whatever makes sense.

Afterwards, when I get to know my case and how it behaves under
varying loads. Does it need tweaking etc. Research overclocking,
as to where to start and how to do it safely. And have fun doing it.
Research and theory is one thing, but it all comes down to doing
it and analyzing the results.

Thanks for your input and interests.
!