• 24.04.2024, 20:16
  • Registrieren
  • Anmelden
  • Sie sind nicht angemeldet.

 

Lieber Besucher, herzlich willkommen bei: Aqua Computer Forum. Falls dies Ihr erster Besuch auf dieser Seite ist, lesen Sie sich bitte die Hilfe durch. Dort wird Ihnen die Bedienung dieser Seite näher erläutert. Darüber hinaus sollten Sie sich registrieren, um alle Funktionen dieser Seite nutzen zu können. Benutzen Sie das Registrierungsformular, um sich zu registrieren oder informieren Sie sich ausführlich über den Registrierungsvorgang. Falls Sie sich bereits zu einem früheren Zeitpunkt registriert haben, können Sie sich hier anmelden.

Help on temperature monitoring

Mittwoch, 4. Juni 2014, 16:18

Grretings to all :
I need your help on one aspect of the cooling liquid that has me somewhat obsessed. It is about controlling the temperatures of the various components..........which are, really, the temperatures should control and the best way to monitor them.
My liquid cooling circuit basically consists of 2 MCP35X pumps with EK dual top , a deposit Aqualis XT , 1x 480 Rad XSPC , 1x 360 Rad XSPC , 1 Aquacomputer Cuplex Kryos CPU block , 1 EK block for motherboard , 1 block EK for the GTX 780 , 2x Aquacomputer flowmeters , 1 filter of Aquacomputer ........ all controlled by a Aquaero 5 XT with its corresponding liquid cooling block.......also I was using 4x temperature probes Aquacomputer in-line............in short, a fairly restrictive circuit, in which I would like to remove restrictions were not necessary (filter?, in-line probes?).
I really wish control through normal analog probes the CPU, GPU; GPU VRM; VRM motherboard; Northbridge, Southbridge, etc ...... but the placement of such probes in liquid cooling block is like an impossible chimera. Furthermore probes in-line liquid are quite restrictive and not know if they are really useful and where should placing .
The only thing what I am sure is that I would like to monitor, in the best possible way, temperatures, either components or liquid cooling for it to verify the optimal performance of my liquid cooling and, in addition, make the most of the qualities and my Aquaero functions.
Well what do you say me about this issue? ........ What should I monitor temps and what would be the best way for you to do?
Thanks in advance

Dieser Beitrag wurde bereits 6 mal editiert, zuletzt von »dopax007« (4. Juni 2014, 16:43)

Mittwoch, 4. Juni 2014, 19:52

The system that I am currently working on has a single loop, 480 and 360 radiator, CPU, Dual GPU, Motherboard, MPS400 Flow Sensor, and RAM water blocks.

I have the Aquaero 6 XT and have installed In-Line Temperature Sensors on the input and output of the Radiators. Also Temperature Sensors on the CPU Block, rear of the CPU, MB Block, VRM, MPS400 at Pump output, and RAM. Additionally I use aida64 to monitor the CPU (Package, CPU, and individual cores), Motherboard, GPU Diodes, PLX, PCH, GPUs, and SSDs via Software Sensors.

I have tried numerous scenarios to control the Fan Speed to include Virtual Temperature Delta based on Rad Input/Output, Virtual Temperature Delta based on Ambient/Loop, Loop Temperature at various points, and component (CPU, GPU, RAM, VRM).

All have had their advantage, with some reacting too quickly to variations and others lagging too much to obtain the desired effect of loop and component cooling.

I have finally settled for Fan Control based on the CPU Temperature (based on a Probe on the side of the CPU itself) where the Fans begin to ramp up at a pre-determined CPU Temp and reach maximum speed long before marginal temperatures are reached.

With an ambient temperature of 25*C my i7-4770K temperature is 30.5*C. When the CPU temp reaches 31.4*C the fan speed on all Radiators and the Case begin to ramp up reaching 100% at a CPU temp of 34*C. This scenario pulls the Loop (and subsequently the component) temperatures back to acceptable values. The maximum CPU temperature that I've seen with PRIME95 running for several hours is a CPU Package temp of 50*C.

Hope this helps.

RodeoGeorge

Dieser Beitrag wurde bereits 1 mal editiert, zuletzt von »wa3pnt« (4. Juni 2014, 19:58)

Mittwoch, 4. Juni 2014, 22:39

Thank you very much for the info ......... is very useful.
If I understand correctly, you are using some common analog probes.........the truth is that I don't see that they have many options to be placed optimally ¿ can you give me more details about how you did it and what parts ?
On the other hand I find it very interesting what you said about the delta temperature .......... i'll inquire more about it.
Again, thanks for the help

Donnerstag, 5. Juni 2014, 07:28

I place a standard Temp Sensor Probe against the CPU itself before installing the water block. Just ensure that the probe is below the top plane of the CPU so that it does not interfere with the water block mating with the CPU. I use a dab of silicone to hold it in place, and then secure the wire near the CPU to prevent it being pulled from the mounted location.

On the In-Line temps I use this probe assembly (http://www.performance-pcs.com/catalog/i…oducts_id=38625) on the input and output of the radiators.

RodeoGeorge

Donnerstag, 5. Juni 2014, 12:20

I don't find the Aquacomputer inline temperature fittings restrictive at all; in fact they are straight through. I have 3 in my loop. Two are just for info and the 3rd is in the loop behind the cpu block and gpu blocks so that there is a meaningful temperature change to work with.

Also I base my curve controls on the Delta t of air/water. The air / water delta t requires making a virtual sensor....then I also make a virtual sensor of 2 ambient air temperature probes.

I have enough radiators that even when under heavy load the system never sees greater Delta t than 3* but it still seems to work fine for pump and fan control once you get the curve controls refined.

I didn't invent it but it works fine for me.
HTH

Donnerstag, 5. Juni 2014, 13:15

I place a standard Temp Sensor Probe against the CPU itself before installing the water block. Just ensure that the probe is below the top plane of the CPU so that it does not interfere with the water block mating with the CPU. I use a dab of silicone to hold it in place, and then secure the wire near the CPU to prevent it being pulled from the mounted location.
.
Thanks.......I'll take in account

I don't find the Aquacomputer inline temperature fittings restrictive at all; in fact they are straight through. I have 3 in my loop. Two are just for info and the 3rd is in the loop behind the cpu block and gpu blocks so that there is a meaningful temperature change to work with.

Also I base my curve controls on the Delta t of air/water. The air / water delta t requires making a virtual sensor....then I also make a virtual sensor of 2 ambient air temperature probes.

I have enough radiators that even when under heavy load the system never sees greater Delta t than 3* but it still seems to work fine for pump and fan control once you get the curve controls refined.

I didn't invent it but it works fine for me.
HTH
.
It was just an opinion..........I see inside the sensor, the flow narrows and considers a restriction . Anyway, thanks for your help in terms of advice on how to handle the delta temperature

Donnerstag, 5. Juni 2014, 14:06

There are many fittings that narrow down to 9mm, especially rotary type, so anything not smaller than that I don't really consider restrictive. Although I do use rotary fittings, I don't use any of what I call "hard 90" fittings; that is the flow is diverted immediately 90* instead of more gradually such as a 2 or 3 way rotary does.
All fittings of course impact flow to some extent. Since I have acrylic tubing I bend the tubing for 90* turns instead of using fittings to lessen the impact of the turns.

Donnerstag, 5. Juni 2014, 22:19

There are many fittings that narrow down to 9mm, especially rotary type, so anything not smaller than that I don't really consider restrictive. Although I do use rotary fittings, I don't use any of what I call "hard 90" fittings; that is the flow is diverted immediately 90* instead of more gradually such as a 2 or 3 way rotary does.
All fittings of course impact flow to some extent. Since I have acrylic tubing I bend the tubing for 90* turns instead of using fittings to lessen the impact of the turns.
.
Perfectly explained by you, mandrix.
I note with respect to the 90 ° fittings ......... thanks

Freitag, 6. Juni 2014, 23:05

By the way , mandrix.............I'm very interested in using , also , curve controls on the Delta t of air/wáter .
It would be correct to use , by example , 3x standard temp probes ..........the first two , one in each air intake of my two radiators , on the outside ( I,ve my fans in pull mode , taking air from the outside ), and one on the inside of the chassis ¿ or is better on the outside ?..and combine it with the use of two sensors in-line, each one in output of each one of the two radiators .......in this case ¿ would it be necessary any more in-line sensor somewhere in the loop ?

Samstag, 7. Juni 2014, 00:26

It's all down to personal tatse & needs really as to how many temp sensors are used & whether to go the Delta route or not to go the delta route :D .. Some go the Delta some don't, Personally i've always found just using two temp sensors per loop works just fine for me.

I put one temp probe on the outlet of each of the blocks I want to monitor (hotest the fluid will be) & a second probe on the Cold outlet of a the last Rad (coldest the fluid will be) & make a curve control just from those.. It's worked for me always for doing all my reviews & guides along with all the Block Test runs I do for Blocks, Rads e.t.c.. Sometimes it's all too easy to get all tied up with too many probes e.t.c (just because you can ;)) but as I said, just using the 2 probe per loop is just my personal choice & has always worked great for me :)

Others do of course use the Delta & find that works well for them, so I guess like most things in this little hobby of ours it comes down to what your personal choice you think will give you the best results for your particular needs & loop make up e.t.c... It is fun messing around with it all mind either way... :thumbsup: ...

N.

Samstag, 7. Juni 2014, 13:49

I'm one of the one's over the top on temp sensors, 13 hardware and 8 software. After six months I find the important ones are coolant temps. Sensors in rad inlets and outlets are the most important, and just as in air cooling ambient air temp effects everything. What has worked best for me is a constant pump speed with fan curves based on coolant temps. My DT'd run from 2-3C at low load to 6-8C at 100% system load.

Samstag, 7. Juni 2014, 22:21

It's all down to personal tatse & needs really as to how many temp sensors are used & whether to go the Delta route or not to go the delta route :D .. Some go the Delta some don't, Personally i've always found just using two temp sensors per loop works just fine for me.

I put one temp probe on the outlet of each of the blocks I want to monitor (hotest the fluid will be) & a second probe on the Cold outlet of a the last Rad (coldest the fluid will be) & make a curve control just from those.. It's worked for me always for doing all my reviews & guides along with all the Block Test runs I do for Blocks, Rads e.t.c.. Sometimes it's all too easy to get all tied up with too many probes e.t.c (just because you can ;)) but as I said, just using the 2 probe per loop is just my personal choice & has always worked great for me :)

Others do of course use the Delta & find that works well for them, so I guess like most things in this little hobby of ours it comes down to what your personal choice you think will give you the best results for your particular needs & loop make up e.t.c... It is fun messing around with it all mind either way... :thumbsup: ...

N.
.

I agree with you , Namron , in that , in the end , will be a personal choice from the various options that you guys are exposing all .............. in your case, however, if I'm not wrong , the coolant temperature in the loop tends to be uniform at all points by varying a few degrees ; according to what I read in some threads around 2-3 ° C.......¿ would not it be more strict , as it somehow , based on the delta temperature ?....... I think what madcratebuilder says goes , more or less , in that direction ............... is only my opinión ; as you very well say , what works well for everyone , is well chosen .
In any case , and above all , thank you for your answers and help
P. D. : Mandrix, I hope your opinion about this and the answer to what you ask two post before.
Greetings

Dieser Beitrag wurde bereits 2 mal editiert, zuletzt von »dopax007« (7. Juni 2014, 22:24)

Montag, 9. Juni 2014, 15:43

I just use two sensors placed on the side of my case to detect ambient temperature, and then I make a virtual sensor by averaging them. I then use an inline temp sensor just behind the gpu's in the loop.
The ambient virtual temp is used with the inline temp to create a second virtual sensor to give the Delta t of air/water. I have a curve control that uses the Delta t to ramp up my fans and pumps.
Might seem overly complicated but actually it's quite simple.

Since I run two Aquaero's my inline temp sensor behind the gpu's actually has two sensor leads, one to feed each Aquaero since temperatures can't be shared between the two. Likewise I also have a second set of ambient temp sensors.
Each Aquaero runs several sets of fans and 2 pumps. I have a total of 4x480mm radiators and 28 fans, 2xD5 pwm pumps, & MCP35x2 pumps.

Montag, 9. Juni 2014, 22:20

many thanks mandrix , you've clearly explained .......... I take note of it

Montag, 9. Juni 2014, 23:17

many thanks mandrix , you've clearly explained .......... I take note of it

:thumbup: