Watercooling GPU!

[Elie]

Image of rig attached and screenshots from video taken literally one second apart with 30 degrees Celsius increase. I am running a i7 6700k , MSI Z170A Gaming M5 motherboard, 16GB of DDR4 3200Mhz ram and a Asus 2070 super graphics card. I am using an EK waterblock for the gpu. 

 

Please let me know what I can do. I have taken the water block apart and reapplied thermal paste along with new thermal pads to ensure they fit perfectly. The liquid is also running well and the card is not hot to the touch at the apparent 87 degrees. Could this be a gpu temperature sensing issue?

 

Thanks in advance!

[Mick Naughty]

What are those pumps? Looks low end going by the wires sticking out. Unless that’s for lights I don’t see. 

 

If flow isnt the problem I assume contact. You applied paste again but how was the spread? Using the correct pads as described in the instructions?

 

Card should be hot to the touch on the backplate side assuming it’s designed right. 

[Bombastinator]

So The problem is not described  but must be derived from examining a bunch of shots from a cell phone?

[For Science!]

8 hours ago, Elie said:

 

 

What radiators are you using? If its the Thermaltake Pacific Aluminium radiators you probably have bad clogging.

[Elie]

13 hours ago, Mick Naughty said:

What are those pumps? Looks low end going by the wires sticking out. Unless that’s for lights I don’t see. 

 

If flow isnt the problem I assume contact. You applied paste again but how was the spread? Using the correct pads as described in the instructions?

 

Card should be hot to the touch on the backplate side assuming it’s designed right. 

 

12 hours ago, Bombastinator said:

So The problem is not described  but must be derived from examining a bunch of shots from a cell phone?

 

The pumps are Thermaltake pr 22 pumps (bigger ones and a D15 pump in the centre). Excuse the wires All the pc cables are locked in the case slots although I quickly connected the peripherals to test the system thats why it is messy. The thermal paste was applied correctly and spread evenly. The thermal pads are also definitely touching the GPU. This is my first time posting so im sorry if I have done something incorrectly. Do you have any recommendations that may help?

[Elie]

9 hours ago, For Science! said:

What radiators are you using? If its the Thermaltake Pacific Aluminium radiators you probably have bad clogging.

https://www.thermaltake.com.au/pacific-rl480-radiator.html

 

I think I am although they are brand new and the flow is actually quite fast.

[For Science!]

1 minute ago, Elie said:

 

Doesn't matter how new or old they are aluminium parts with opaque coolants is a guarantee for precipitation and clogging. You can read around the topic but I guarantee in a few days to 2 weeks you will notice that the coolant becoming opaque.

[Bombastinator]

I’m still trying to figure out what question you want answered.  There was some sort of implication that there was a problem that was evidenced by a series of snap shots but I can’t see the snap shots and don’t want to spend 10 minutes trying to decode clues.  What sort of problem do you think you are having?  I don’t watercool so I can’t speak to that radiator thing.   I do know that there are several chemicals that do weird things in the presence of aluminum though.  One of them is copper.  You don’t want copper and aluminum to touch because it creates a sort of battery that produces wild amounts of corrosion.  It’s very slow though.

[Elie]

10 minutes ago, Bombastinator said:

I’m still trying to figure out what question you want answered.  There was some sort of implication that there was a problem that was evidenced by a series of snap shots but I can’t see the snap shots and don’t want to spend 10 minutes trying to decode clues.  What sort of problem do you think you are having?  I don’t watercool so I can’t speak to that radiator thing.   I do know that there are several chemicals that do weird things in the presence of aluminum though.  One of them is copper.  You don’t want copper and aluminum to touch because it creates a sort of battery that produces wild amounts of corrosion.  It’s very slow though.

Hi, The question I want answered is why is my GPU jumping up and down a significant amount of temperature in a matter of seconds. Could my sensor be broken?

[Elie]

34 minutes ago, For Science! said:

Doesn't matter how new or old they are aluminium parts with opaque coolants is a guarantee for precipitation and clogging. You can read around the topic but I guarantee in a few days to 2 weeks you will notice that the coolant becoming opaque.

So you are suggesting that my fluid is opaque although the P1000 Pasetel coolant from thermaltake comes like this?

[Bombastinator]

Hmm...  Could possibly be the high thermal mass nature of water cooling.  I leave that to others who know watercooling.

 

There are some that could have nothing to do with the cooling though.  It would mean that the. GPU is ramping up and down very fast.  There are a couple of reasons it might do that.

[Elie]

7 minutes ago, Bombastinator said:

Hmm...  Could possibly be the high thermal mass nature of water cooling.  I leave that to others who know watercooling.

 

There are some that could have nothing to do with the cooling though.  It would mean that the. GPU is ramping up and down very fast.  There are a couple of reasons it might do that.

I judt dont believe a card can go from 35 Celsius to 88 Celsius in a matter of seconds then back down. The card is not even hot at 88 Celsius 

[Bombastinator]

Well if the card is still under warranty simply RMA it.  There’s no fixing such a thing if it is a problem. A watercooling guru might have a more specific answer.  I’m not one I’m afraid.

[Elie]

2 minutes ago, Bombastinator said:

Well if the card is still under warranty simply RMA it.  There’s no fixing such a thing if it is a problem. A watercooling guru might have a more specific answer.  I’m not one I’m afraid.

The card is brand new thats why it is weird

[For Science!]

15 minutes ago, Elie said:

So you are suggesting that my fluid is opaque although the P1000 Pasetel coolant from thermaltake comes like this?

Yes, i've seen it over and over and over again.

 

[For Science!]

More down to the point of the actual issue with the GPU temperatures, I see you have installed the waterblock with the factory backplate. Are you sure this is not causing some issue with the mounting? afaik most of the EKWB blocks don't let you keep the backplate.

[Elie]

4 hours ago, For Science! said:

More down to the point of the actual issue with the GPU temperatures, I see you have installed the waterblock with the factory backplate. Are you sure this is not causing some issue with the mounting? afaik most of the EKWB blocks don't let you keep the backplate.

I could not see as to why that would affect it. I did remov e the back plate with no success. I have attached images to show the thermal pads are touching the areas.

[Mick Naughty]

7 hours ago, Elie said:

 

 

The pumps are Thermaltake pr 22 pumps (bigger ones and a D15 pump in the centre). Excuse the wires All the pc cables are locked in the case slots although I quickly connected the peripherals to test the system thats why it is messy. The thermal paste was applied correctly and spread evenly. The thermal pads are also definitely touching the GPU. This is my first time posting so im sorry if I have done something incorrectly. Do you have any recommendations that may help?

Then it’s a flow issue. As you don’t have anything else in the loop to see it and no temp sensors for some reason. 

 

Id use another program to see the gpu temp to narrow that down. Gpu-z or something and see if the sensor is faulty. Should be able to see something similar when not gaming or in a benchmark. 

[For Science!]

1 hour ago, Elie said:

I could not see as to why that would affect it. I did remov e the back plate with no success. I have attached images to show the thermal pads are touching the areas.

Doesn't really show anything, only a dismount and a picture of how the thermal paste has spread is proof that the mounting is sufficient.

[Powell]

On 11/13/2019 at 4:04 PM, For Science! said:

Yes, i've seen it over and over and over again.

 

Oh have i got a story for you. So after i decider to buy the ek radiators and recycle the thermaltake ones. Ive pulled down the gpu loop to scratch fitting the new radiators all going well, next minute ka poosh! The mobo/cpu loop blows open. Coolant on my monitior, chair, keyboards, mice, wall, cupboard just about the whole room! That thing had pressure. So its been sitting for about three weeks with no power. This us as far as im concerned that the pressure build up is from the mixed metals as the loop has been turned off for weeks. 

 

Needless to say im glad it had now power and i had already received the ek radiators. 

[For Science!]

10 minutes ago, Powell said:

Oh have i got a story for you. So after i decider to buy the ek radiators and recycle the thermaltake ones. Ive pulled down the gpu loop to scratch fitting the new radiators all going well, next minute ka poosh! The mobo/cpu loop blows open. Coolant on my monitior, chair, keyboards, mice, wall, cupboard just about the whole room! That thing had pressure. So its been sitting for about three weeks with no power. This us as far as im concerned that the pressure build up is from the mixed metals as the loop has been turned off for weeks. 

 

Needless to say im glad it had now power and i had already received the ek radiators. 

Sorry to hear that. This sounds similar to the guy who added mayhems pastel premix to the EKWB 240G kit; the dude had tubes swelling overnight.

 

I will add this onto my long growing list of confirmed cases of pastel/opaque fluids + ammonium causing issues