Thermal Putty Roundup von Pink oder Grau bis Blau und von Halnziye oder Laird bzw. Dow - HY234, HY236 und HY268 vs. Hardwareliebe Extreme64 im Test

[Redaktion]

Heute teste ich viermal Thermal Putty für Euch, wobei diese vier Produkte exemplarisch für große Firmen und deren Produkte bzw. Mischprodukte stehen. Wenn es mir gelingt, noch weitere Produkte zu organisieren, werde ich die Datenbank auch um diese Sparte erweitern, denn wer die Umstände kennt, wenn auf Grafikkarten die Padstärken nicht bekannt sind, wird gern […] (read full article...)

 

[Besterino]

Danke für den Test, die Datenbank wird was ganz Feines werden! =)

 

[Igor Wallossek]

@Snarks Domain Thx for posting.

One Question: How thick was the measured gap and how much pressure (estimated)?

 

[Snarks Domain]

@Snarks Domain Thx for posting.

One Question: How thick was the measured gap and how much pressure (estimated)?

The GAP between VRAM and cooler is 1.8mm on my Zotac 3070ti (GDDR6X). Not positive on mounting pressure, but it's lower than other card since it uses 4 spring screws, instead of an "X" bracket.

he my tests the fan is locked at 85% and an overvlock of +1100Mhz is added using MSI Afterburner.

i should note that since it's a Nvdia card the Memory Junc5ion Temperature o ly goes up or down in 2 degree increments within HWINFO and other software. I'll start using a new Testbench soon using an AMD 6800 videocard. The AMD 6000 series shows Memory Junction Temperatures in 0.1 degree increments, so will be a nice improvement for better conversions. I'll end up retesting everything again on the new testbench. Should keep me busy for the Winter.

 

[Igor Wallossek]

The intention to buy such an equipment is the problem to be flexible (and because I'm so curious ). I've tested in the past also on VGA cards and CPUs but I saw very fast all this ugly limitations and I damaged on the end 2 cards (screws).

For the readers it's very interesting to know, on which thickness a gap filler works optimal and how deep I can it press it down. And I don't need any deltas, because I can control all temps and keep it constant (+/- 0.05 K) and I can also play with the pressure.

For all the PTM7950 lovers: after 10.00 cycles it seems totally damaged

 

[Snarks Domain]

Sounds like a nice test setup. I hope to have something similar one day, but that will have to wait for now.

Most thermal aputties can be compressed down to around +/- 0.1mm Minimum Bondline Thickness.

 

[Igor Wallossek]

It is important to know the real numbers



(The costs of such equip is equal to a MB E-class T-model )

 

[Xaphyr]

Mega, danke für den Test @Igor Wallossek. Sag mal, was das säubern betrifft, wie bekommt man den Kram am besten wieder runter? Ich nehme an, das geht nicht so easy mit Wattestäbchen und Isopropanol, oder?

 

[Rizoma]

Mega, danke für den Test @Igor Wallossek. Sag mal, was das säubern betrifft, wie bekommt man den Kram am besten wieder runter? Ich nehme an, das geht nicht so easy mit Wattestäbchen und Isopropanol, oder?

https://www.igorslab.de/community/t...e-wärmeleitpaste-aus-fernost.8212/post-252831

 

[roccale]

For all the PTM7950 lovers: after 10.00 cycles it seems totally damaged

What?
How can I avoid 10 Phase Transition cycles if the ptm7950 changes if I remember correctly now between 40 and 50 degrees celsius?
I basically play 10 times and I have to change the PTM?
I'm confused.
Mine seems to be fine for now.
Laird seems to have a higher hotspot:
<<Softening Temperature Range 50°C to 70°C>>

 

[Besterino]

I hope he meant cycles = removal+(re)apply of cooler/measuring equipment, i.e. "physical re-use" (not phase transitions)...

 

[Igor Wallossek]

What?
How can I avoid 10 Phase Transition cycles if the ptm7950 changes if I remember correctly now between 40 and 50 degrees celsius?
I basically play 10 times and I have to change the PTM?
I'm confused.
Mine seems to be fine for now.
Laird seems to have a higher hotspot:
<<Softening Temperature Range 50°C to 70°C>>

Typo 10,000 cycles (phase transitions)

 

[roccale]

Typo 10,000 cycles (phase transitions)

Oh ok good. 10,000 should be enough

 

[Igor Wallossek]

My todays biggest problem:



The maximum 500N force of the TIMA were not enough to cut off this shit after pressing and measuring it and drive up the test body. This was the end of both test bodies

Positive effect: I've talked with Nanotest and they will send me extra hard bodies with extra coating of chromium nitride and titanium nitride. That's perfect to measure liquid metal shit

 

[roccale]

I hope he meant cycles = removal+(re)apply of cooler/measuring equipment, i.e. "physical re-use" (not phase transitions)...

Impossible. You can't do it. It destroys itself.

 

[roccale]

My todays biggest problem:

Anhang anzeigen 34173

The maximum 500N force of the TIMA were not enough to cut off this shit after pressing and measuring it and drive up the test body. This was the end of both test bodies

Positive effect: I've talked with Nanotest and they will send me extra hard bodies with extra coating of chromium nitride and titanium nitride. That's perfect to measure liquid metal shit

WTH is that??????

 

[Igor Wallossek]

Parker Gel (HF60) - Not bad at all, but a real glue. 100% no pump-out

 

[roccale]

Parker Gel (HF60) - Not bad at all, but a real glue. 100% no pump-out

Hahahaha I saw, I can imagine.

 

[IXI_HADES_IXI]

Zumindest weiß ich jetzt das ich etwas falsch gemacht habe wenn das HWL64 Extreme doch so gut performt bei dir. Danke für den Test!

Da dasPutty von HWL extrem leicht abgeht vermute ich das ich bei der Installation nicht ganz aufgepasst habe und ich eventuell das Putty verrutscht habe oder wie auch immer. Zumindest geht es wesentlich leichter ab als die HY Varianten und lässt sich auch 1A verteilen.

Hab mal ein Bild mit HY234 daneben eingefügt. Das klebt, lässt sich schlechter verteilen und auch wieder richtig schlecht entfernen. HY234 ist sogar von Isopropanol relativ unbeeindruckt. Da braucht man schon eine kleine Bürste und viel Geduld um wieder alles restlos runter zu kriegen.

Die besten Erfahrungen habe ich aber doch wieder mit den PADS gemacht. Nutze momentan Artic TP3 Pads. Dafür braucht man aber trotzdem die richtige Dicke, denn vor allem in Verbindung mit Kryosheet, wenn das zu wenig Druck bekommt, geht der Hotspot durch die Decke...

 

[IXI_HADES_IXI]

Hier das Bild mit den Artic TP3 PADS... Hatte im Memtest Vulkan damit die besten Temperaturen. Wenn man aber eine schiefe Karte, einen schiefen Kühler bzw einfach unterschiedliche Höhen aufgrund solcher Toleranzen hat wirkt Putty wahre Wunder, da es den großen Vorteil hat, dass es sich selber an verschiedene Höhen optimal anpassen kann.

Raumtemperatur glaube ich hatte ich da fast 27 Grad, also ziemlich warm. Ich hatte die Karte mehrmals repastet weil mein Vram mit 105 grad einfach zu heiß wurde. (Sapphire Nitro 7900XTX). Mittlerweile läuft aber wieder alles einwandfrei.

 

[Snarks Domain]

At least now I know that I did something wrong if the HWL64 Extreme performs so well for you. Thanks for the test!

Since the putty from HWL comes off extremely easily, I suspected that I wasn't paying attention during installation and I may have slipped the putty or something. At least it comes off much easier than the HY versions and is also easy to spread.

I've included a picture with HY234 next to it. It sticks, is harder to spread and is really hard to remove. HY234 is even relatively unaffected by isopropanol. You need a small brush and a lot of patience to get everything off completely.

But I've had the best experience with the PADS. I'm currently using Artic TP3 pads. But you still need the right thickness, especially in combination with Kryosheet, if there's not enough pressure, the hotspot will go through the roof...

It looks like you have close to the correct amount of putty. If you try putty again I recommend rolling it into logs/cylinders and letting the boutty get the final compression during reassembly.

here is a guide I made which you might find useful. There is also a Thermak Putty Calculator for log diameter and grams needed (if you know your putties density). Using a pair of digital calipers to measure OEM pads or gap thickness, and the calculator, you can get the perfect amount.