Thermal grizzly conductonaut worse then stock tim???

[Bill Dimwit]

I put some thermal grizzly conductonaut on my 980ti and help it run cooler along with Thermal grizzly pads on the ram and VRM.
Oddly now that card is running 7c hotter then stock
At first I thought I made a mistake and cleaned the card off and reapplied the conductonaut, but the temps stayed the same.
I used Cool labs liquid metal a lot in the past and applied it in the same way they recommend.
Dose conductonaut have a curing time? However the temps keep getting worse, after a fresh apply of conductonaut the card loads at 63c, but after about a day the max temps are closer to 67-68c. With the stock tim the card run no hotter then 60c. Im running the fan at full speed when I load the card up.

I'm just stumped, what could I have done wrong? I can't think of anything.

 

[Lochekey]

Did you use the correcct thickness pads on the ram and vrm, if they are to thick you could be having an issue with not making good contact with the conductonaut.

Do you have any pictures off the die after pulling the heatsink off?

I am not familiar with the cool labs method of mounting, have you tried the recommended technique from Thermal Grizzly?

 

[Bill Dimwit]

Did you use the correcct thickness pads on the ram and vrm, if they are to thick you could be having an issue with not making good contact with the conductonaut.

Do you have any pictures off the die after pulling the heatsink off?

I am not familiar with the cool labs method of mounting, have you tried the recommended technique from Thermal Grizzly?

I have the correct thickness for the pads, How does Thermal Grizzly say to use the stuff? It did not come with intrusions.
This is how I applied it.

I cleaned the card off and tried some thermal grizzly kyronaut and the card now load at around 83c

 

[EarthDog]

Website should have instructions...

 

[Lochekey]

Try this.

- - - Updated - - -

Oh and having the kryonaut run that hot is not normal either. Still sounds like you are getting a poor mount.

 

[Bill Dimwit]

Try this.

I tired applying it that way, its more or less the way I applied it the first two times.
I did a full clean of the card and heatsink with rubbing alcohol, no change in temps.

Oh and having the kryonaut run that hot is not normal either. Still sounds like you are getting a poor mount.

I thought so too, but after taking the heat sink off it was spread well and showed no signs of a bad mount. How ever I could have used a little more as it was not spread out over the full die, it had about 80-85% coverage. Its is about 85f in my computer room and the card sucking in the heat from the CPU heatsink.
Maybe I'm just expecting too much? But I could have sworn the stock tim never went over 60c.

Website should have instructions...

I thought so too but could not find it.

 

[EarthDog]

Was it 85f in your room when you recorded your initial temps? If the temp was 3c difference ambient, then your card will run 3c warmer.

I mean, changing tim is typically good for 1-2c... thats about it. Not a lot of difference between the majority of them really.

 

[Bill Dimwit]

Was it 85f in your room when you recorded your initial temps? If the temp was 3c difference ambient, then your card will run 3c warmer.

I mean, changing tim is typically good for 1-2c... thats about it. Not a lot of difference between the majority of them really.

It was hotter when I tested with the stock tim, about 90f, I just don't get it. How could liquid metal be doing worse then stock tim? I also have better fans in the case now too. They push more air then the ones I have in the system with the stock tim. It just doesn't add up.

 

[Bill Dimwit]

Ok now I feel dumb. I forgot I put better tim and lapped my CPU heat sink. The GPU is now sucking in a lot more heat then it was when I was running stock tim. That could be where the extra 7c is coming from.

 

[EarthDog]

It was hotter when I tested with the stock tim, about 90f, I just don't get it. How could liquid metal be doing worse then stock tim? I also have better fans in the case now too. They push more air then the ones I have in the system with the stock tim. It just doesn't add up.

too many variables changed...good luck.

Ok now I feel dumb. I forgot I put better tim and lapped my CPU heat sink. The GPU is now sucking in a lot more heat then it was when I was running stock tim. That could be where the extra 7c is coming from.

que? If you lapped the gpu and the only difference (aside the few above) is paste, it would be the same (better paste, 1-2c better temps). Im not understanding the chronology of things here.

 

[trents]

So you're saying you lapped the CPU and so more heat is being transferred to the CPU cooler such that inside the case air temps are higher than before and this is causing the GPU to run hotter?

 

[Lochekey]

is this a watercooled system on 1 loop?

 

[Bill Dimwit]

to enplane things better.

|When I put liquid metal tim on the card.

I put higher CFM fans in the system.
Room temp is 5f lower.
I put better Tim on the cpu and lapped its heatsink.
The GPU now runs 7c hotter.

Everything else is the same.

The CPU is near millimeters below the GPU fans intake. The GPU fan acts as both the CPU fan and GPU fan.
See atched image, they're old but will give you an better idea of what I'm trying to say. My only guess that the extra heat now being dissipated by CPU heatsink is now being sucked into the GPU. As you can see heat from CPU heat sink is being sucked into the GPU.
View attachment 199076View attachment 199077View attachment 199078

I added two 40mm fans on the CPU heat sink and covered the top of the CPU heat sink and ran the folding client on just the GPU, the GPU temps dropped by 3c. Its still 4c over what the stock Tim was, even with better fans and lower room temp.

I guess this will be the first and last time I buy anything from thermal grizzly. Looks like there stuff is crap like 90+% of the other over engineered crap from Germany. Kryonaut ran 20c over stock, conductonaut is 4c over stock. Both with better fans and lower room temps then when tested with the stock tim.
The back of the card where the VRM and Vram is, is running extremely hot compared to the stock pads. So the Thermal grizzly pads must be junk too.
I'll be buying some Coollaboratory Liquid Pro and Fujipoly SARCON X-E pads like I should have in the frist place

 

[trents]

That's a funky looking rig, Bill. Your issue makes more sense now with the pics.

 

[EarthDog]

My only guess that the extra heat now being dissipated by CPU heatsink is now being sucked into the GPU.

Thanks for the pics. I think i get it now.

That said, i dont buy it.....bear with me...

Even though you lapped your cpu etc, the load you are dissipating does not change. Temperature is not a function of wattage to cool. Think of it this way, which is hotter, a yellow flame bonfire or a yellow flame lighter? Answer: they are both the same temp (YELLOW flame). Clearly though, the bonfire has a lot more energy. Similar concepts here. Cpu is still putting out the same load and will need the same energy needed to cool it. Because it was lapped does not mean the load changed, just temps.

I guess this will be the first and last time I buy anything from thermal grizzly. Looks like there stuff is crap like 90+% of the other over engineered crap from Germany. Kryonaut ran 20c over stock, conductonaut is 4c over stock. Both with better fans and lower room temps then when tested with the stock tim. The back of the card where the VRM and Vram is, is running extremely hot compared to the stock pads. So the Thermal grizzly pads must be junk too. I'll be buying some Coollaboratory Liquid Pro and Fujipoly SARCON X-E pads like I should have in the frist place

Seeks unfair to blame the product especially considering the dozens of empiriy tested and sound reviews which rate it quite well and in the pack with others. When results are this off, clearly something else is going on.

 

[WhitehawkEQ]

to enplane things better.

|When I put liquid metal tim on the card.

I put higher CFM fans in the system.
Room temp is 5f lower.
I put better Tim on the cpu and lapped its heatsink.
The GPU now runs 7c hotter.

Everything else is the same.

The CPU is near millimeters below the GPU fans intake. The GPU fan acts as both the CPU fan and GPU fan.
See atched image, they're old but will give you an better idea of what I'm trying to say. My only guess that the extra heat now being dissipated by CPU heatsink is now being sucked into the GPU. As you can see heat from CPU heat sink is being sucked into the GPU.

I added two 40mm fans on the CPU heat sink and covered the top of the CPU heat sink and ran the folding client on just the GPU, the GPU temps dropped by 3c. Its still 4c over what the stock Tim was, even with better fans and lower room temp.

I guess this will be the first and last time I buy anything from thermal grizzly. Looks like there stuff is crap like 90+% of the other over engineered crap from Germany. Kryonaut ran 20c over stock, conductonaut is 4c over stock. Both with better fans and lower room temps then when tested with the stock tim.
The back of the card where the VRM and Vram is, is running extremely hot compared to the stock pads. So the Thermal grizzly pads must be junk too.
I'll be buying some Coollaboratory Liquid Pro and Fujipoly SARCON X-E pads like I should have in the frist place

After looking at your setup, that GPU is not suited for that MB/Backplane, even though it fits. You need a shorter GPU like GIGABYTE GeForce GTX 1080 that will work well with your setup.

 

[Bill Dimwit]

After looking at your setup, that GPU is not suited for that MB/Backplane, even though it fits. You need a shorter GPU like GIGABYTE GeForce GTX 1080 that will work well with your setup.

View attachment 199099

No vga or BNC. Will not work for me.
I also chose the 980 to help cool the 35w CPU I have. And anyway I fail to see this how a problem. Its not different the running SLI.


I put cool labs liquid pro and fujipoly pads on the card, Its now running max at 58c, that's a 9c drop. And the back side of the card is much cooler
I tried the Thermal grizzly crap on a few other heatsinks and its defiantly crap. Kryonaut was 3c higher then zalman zmstg2m on my cpu. zalman zmstg2m is well know for being a cheap mid range tim at best,
Conductonaut was 4c higher then the stock tim on my gd8200 laptop and liquid pro was 4c lower then stock.

There is nothing that should have made conductonaut run so much hotter then the stock tim on my 980. I had higher CFM fans and lower room temps when I tested the conductonaut.
I loaded up the GPU and blocked of the CPU heat sink from the GPU and it still ran higher then stock.
To make matters worse, when I took the heat sink off the thermal grizzly pads had dried out and cracked and crumbled apart. The conductonaut ate into the die of the GPU too. This was all after just 9 days.
Ether I got two bad tubes of thermal grizzly tim and two bad strips of pads and thermal grizzly has some major quality control problems or the reviews out there are straight up marketing lies.

 

[EarthDog]

You are the only one that calls it crap.

When properly applied with proper mounts, etc... its one of the best when its actually tested empircally. Not sure what the deal is on your end (hardware used, mounts, application, wtc), but ill hang my hat on review results where additional care is taken to minimize variability, etc.

Do you have an idea why your results are so off from every review on it? (And no, all reviews which show that result are not shills...)

 

[WhitehawkEQ]

No vga or BNC. Will not work for me.

You can get a DP to VGA adapter, I have 1 and it works great.

 

[Woomack]

Conductonaut is almost only made of gallium/indium. There is nothing special in it. Nearly every "liquid metal" is made of the same stuff. They only mix it and put different label.
Other thing is that using liquid metal on a GPU is a stupid idea. One thing is that you need perfect contact and other thing is that liquid metal reacts with everything near like aluminum and some other metals.