Honeywell PTM7950 vs Thermal Grizzly Conductonaut EXTREME - Ai was wrong about the Honeywell pad
Вставка
- Опубліковано 14 лис 2023
- Ebay store: www.ebay.co.uk/usr/sorinelect...
WebSite: sorinelectronics.com/
Patreon support: / electronicsrepairschool
UA-cam Members: / @electronicsrepairschool
Forum: forum.sorinelectronics.com/
Discord: www.sorinelectronics.com/discord
My tools: www.amazon.co.uk/shop/electro...
PayPal me: www.paypal.com/paypalme/Elect...
I'm so happy you read the comments from the last video
The heating cycle, or "curing", is important. Even Honeywell's recommended application includes an instruction to preheat the heatsink surface before applying!
Huh, didn't know about preheating the surface, but I did know running it through a few heat cycles was required for best effect. I did this by running a few 20 minute gaming sessions on something that I know taxes the system when I applied it to my Steam Deck. It seemed to do the trick for me.
7950 is precured
@@dun8632 yeah, I think people have been calling it curing because that's what other processes of breaking something like this in would be called, even though that's not what is exactly happening.
But yeah, heating it up so it changes phase to allow it to better conform to the shape it needs is what is actually happening. Usually doing it either for a long period or several times is necessary because in liquid form this is a very viscous material.
@@dun8632 the viscosity also prevents pump out and instead allows it to just change as the other materials warp.
personally think that is fake if its on amazon and not sold by Honeywell
Sorin, you tested the pc without the rear cover so we don't know how much air the fans was getting sitting directly on the desk. Also the black mat on your desk may have reduced the air flow in all the tests.
You lack of logic it is all comparisons in the same air flow same everuthing
It is indeed necessary to place a computer on a high perforated platform for testing fans.
Reminds me of a time when a new computer came to me less than a year for repair, the customer loads after 20 minutes of playing, the computer freezes... The genius customer removed the black silicone feet underneath.
@@mauricemenard2243he tested the pad with backcover on and liquid metal with backcover off.
The airflow should be even better without the cover
@MykolaInc not if the computer was placed on a mat that touched and blocked the fan air inlets. Also, covers are usually designed in such a way that it makes the air flow through all the correct places of the heatsink, which includes places that are cooling the VRM, VRAM, etc. It is an example of aerodynamics. There are lots of cases where customers make a lot of additional holes in the bottom case thinking that it will enhance cooling, but what really happens is the opposite, they destroy the correct airflow, for example VRMs stop getting the airflow at all because the flow path gets destroyed and air gets sucked in right into the fan without getting around VRM or other important region that needs cooling first.
Really Thank so much for Video I have applied I got 7900x to 5.4ghz on all core vth liquid metal
Nice compare!
PTM is solid only at room temperature, but above 35C it starts to melt and act like a paste without pumpout. the cycling of this state can push out any air from the two surface ending in a perfect contact fill, after 500 cycles the thermal impedance lowers according to the Honeywell measurements, so this thermal interface gets better over time not worse, basically on a laptop you need to clean the dust with the Honeywell and you are good to go.
Phase Change Material is quite good indeed. Definitely gets better over time, even weeks/months later. PCM's can melt and squish down to 0.038mm or 0.0015" (1.5 Thousandths of an Inch). Which Thermal Putty are you using on the VRAM/VRM/etc? Or is that the factory Thermal Putty?
Best tim for rdna 2 and ada imo
Hello Sorin,how are you? Plese make a test and maybe video with KingPin thermal compound - its non elektrik and its 18Wmk which is the best after liquid metal wihis is risky to use.Have a great day!
Sorin try dowsil tc-5888.Same performance as ptm and easy to apply.This paste is used for servers or high end consumer devices.One more thing you should repaste thermal putty on your vram,vrm..etc.Once you open it..same drill as thermal paste.
From my experience with liquid metal you can use too little, and the amount Sorin is using is too little. It was just the right amount he used on the cpu before he started to move it to the GPU.
I have gone through 3 packs of liquid metal. Both on cpu, gpu, desktops laptops, playstations.
Every time I saw similar performance with regular thermal paste, i opened up again and applied a little bit more and the magic happend.
But keep in mind too much is dangerous, therefore I always cover the rest of the cpu around the die with nail polish.
It will catch any spilled liquid metal.
I personally won't touch the stuff, but I don't care as much for the performance increase and I'm not a professional technician. I like to repair things as a hobby, but currently my personal motto is don't try to fix something you're not afraid to break even further beyond repair. It's been a good philosophy as I learn on my own.
Is liquid metal better for thermals than paste?
@@mordaeu1411 Yes
For me Honeywell PTM7950 is better than liquid metal on laptop because PTM7950 is safe to use and last longer and don't have to change thermal paste ever so only fans to clean. I'm still using it more than 4 months the performance are still the same and performs well and don't have to worry about risking my laptop to get damage by liquid metal.
@@reshiram8747 Does it have better "properties" than other liquid metals? Cooler? I've never used it so I don't know how it works.
Why you did not compare temperature?
For what purpose as after a few seconds the limits of PL1, PL2 enter and the temperature drops to, for example 70°C. The best is to have hwinfo running in the background and run a 10-minute cinebench stress test and compare the average values of cores frequencies, cpu package power and the number of points scored in the test.
Because on all thermal solutions they are thermal throttling at 95°C at some point during the test. The only way he could compare temps is by charting them on a graph against time during the test. On many laptops running heavy loads the difference between performance has less to do with just the chips themselves and more to do with how efficient the chips are as well as how good the cooling is, because the system *WILL* thermal throttle, the question is how long does it take to throttle and by inverse how long does the system get to take advantage of its full processing power.
hi sorin try also direct all fan to 5volts and see what the deference 😊😊
I also orderend PTM7950 for my Lenovo P16, because it only reached a score of about 15000 on Cinebench23 and was overheating even in idle mode. Before that, I took it to the Lenovo service and they replaced the cooling unit. That didn't really help, but made it even worse. I was also thinking about liquid metal, but that seemed to be too dangerous for a 4000+ Euro workstation. So I ordered a piece of 30mm * 20mm which covered exactly both the A3000 GPU and the 12950HX CPU. Now I can reach a score of about 20500. Although it also begins to throttle after some minutes under hevay load, it's only PL1 holding it back now. But that's okay.
Apart from that there is still a "0,5GHz on battery bug". If you ever heard about that one, I'd be glad if you have any technical explanation for that. The CPU will throttle to 0,5GHz on battery if any higher load is requested. Temps are about 50° then.
What processor?
@@hansstam5835 12950HX
Usually the pressure between the heatsink and chips on laptops is very low, so what happens is that the contact is very poor. You can test this contact appying the thinnest layer of thermal paste that you can, installing the heatsink and removing to see the amount of paste that is on the heatsink. You will see that in most cases the heatsink is contacting only 2/3 of the chips.
That is a problem with liquid metal and low pressure mounts. Since the liquid metal layer is very thin, it does not allow for full contact between gpu and heatsink. The honeywell thermal pad is 0.2 or 0.25mm thick, and will flow with heatcycles, allowing full contact because it fills the gaps perfectly.
If you want to test a temperature difference between different interface materials on laptops, put the fans at full speed(or at a fixed speed, but some laptops don´t allow this so you have to use full speed), and disable boost on the cpu. If you don´t disable boost the cpu will boost until it reaches 95 or 100ºC depending on the cpu.
I used to use Thermal Grizzly pastes then I used PTM and got amazing results of 10-14C drops. Left Thermal Grizzly in the dust. When I mentioned this on forums all I got was Grizzly fans making excuses for it and that its not meant to be used in this or that or this way and that. Yet you look at the official marketing and...doesn't say any of that. Thermal Grizzly is obsolete.
Same here. Used all sorts of the popular brands like Grizzly, MX-4 and Noctua but they pump out/degrade in a a couple of weeks on agressive superturbo laptops/vga which makes them suitable for desktop CPU's only with heatspreader and high contact pressure. If the fans cannot be removed easily (Stupid design) and there is PTM i clean light fan dust with a soft brush and blow it out with the fans blocked and disconnected unless it's really caked.
I switched to PTM for VGA cards and hot laptops and basic office laptops get GD-2 viscous paste from Ali. Before GD-2 i used Shin-Etsu X23-7921-5 industrial paste but it's very expensive like 800/1200€ for 1KG
(This week, i'll receive around 250 laptops for refurbishing)
250 computers and you have time to comment on youtube?@@harrydijkstra9936
@@harrydijkstra9936 I also love GD-2, use it lots. There are is also a test on the OC page where it outperforms Kryonaut and it's so much cheaper. Want to use PTM but worried about reports of 'fakes'. Can you point to a decent Ali store to buy PTM? What is your experience?
How about you be more specific as to which paste you actually used rather than being vague as possible to try and make a point, and I'm willing to bet you more than likely used Kryonaut which literally has about the same W/mk but no doubt the PTM would last longer so if that's the "Grizzly" you've been using then the results are quite obvious
@@wolfstorm5394 Ahhh I'm sensing another Grizzly fan telling me whichever Grizzly paste I used it was the wrong one due to...excuses that Grizzly never mention anywhere in their specs or instructions.
hello sir, i like your videos i want to know if pad need to be repaste same time when you repaste cpu and gpu some new laptop use the k5 pro when you look hot spot in gpu the temp is very high 96 and 99 let me know your advice if it normal. test for us artic mx-6 and some popular paste on your channel to what the best thermal paste thank you.
K5 Pro is the worst thermal putty available. Use Upsiren U6 Pro or the more powerful UX Pro for the power section, vrams. PTM7950 is a paste replacement and you apply it to the CPU and GPU core.
@@goorthiss thank you
@@goorthiss As a repair store i have seen many user "improved" devices with K5 Pro and i fully agree. Silicone outgassing all around the PCB's and bubbly air pockets.
The rule is. The cooler the name, the cooler the product, the cooler your CPU.
Good evening, thank you for sharing with us 👍
There may be several options for repair 1. Disassemble the left side FAN completely, clean + put grease, enter the bios and monitor the left FAN speed. 2 Although the processor surface looks clean, it is still gray in color.ua-cam.com/video/3nb_6Ihq2EI/v-deo.htmlsi=bstfkHnSSc9BvLgv&t=295 .. it should be copper colored, light polishing, 600 grain sandpaper, after that 1000 grain, final polishing 2000... 3 It could be CPU pipes, cooling without gas is rare, but it may be possible to disable a temporary video card Place other ribs on the CPU and check temperature
You don't need to disturb ptm 7950 on some systems. Some gaming laptops allow you to remove fans without removing the heatsink.
It can be reused. A have an MSI like the video poster, and I've heard one user with same system (as myself) reused 7950 after removing heatsink to clean their fans. They scooped up a bit of the squished off material from besides the die and put on top of the die before reassembling. It required a couple heat cycles and then tests showed it was performing the same.
No hate, you change the paste to have better cooling for the chips, not to have a better score in benchmarks. Use the liq metal for a long while and monitor the temps.
Also chips are throttled so they dont overheat.
Also, laptops don't have good thermals. On Alienware laps is normal to have close to 90 degrees C on the CPU (insane, I know)
Scores change for little reason...we want thermal values/results.
It's expensive but worth it. Honeywell stuff the real thing
After applying the liquid metal, everything around needs to be coated with a dielectric varnish for protection ! Damn ! And you have 348k subscribers...
"Attach your fans to the 12V line instead of the 5V one." People here is so funny 🤣
Poate ca gresesc eu dar cred ca trebuia mai mult lichid pe cpu. Eu pun si intind lichidul ala pana se face ca o oglinda. La mine a dat rezultate foarte bune la desktop cpu. Insa mare atentie ca daca e prea mult lichid poti face un scurt de toata frumusetea 😂
ptm one has seen max 95c and lm one seen 98c. clearly wrong amount of lm used.
my thermal paste took one week to cure, it dropped 10c.. if you just change the thermal paste and start to use it won't change that much until it cure..
Try lt-cooling lt-100, it is better than grizzly metal and much cheaper, a friend tested it against conductonaut and it's not a lie, the efficiency was better.
I do not recommend using liquid metal for beginers.A drop of liquid metal in wrong place can kill the hardware.For safety if want something from the same brand just use Thermall Grizzly Kryonaut,its non conductive.Electronics Repair School is very carefull and looking for not using to mouch licquid metal,but there is an one important thing missing. Arround gpu core there is a lot resistors. If the licquid metal will flow on those resistors the gpu can be done ( whole core for replacement).Always protect those resistors for example with varnish,capton tape,or combination with both together.Also,it may be great idea for an video to show it can happen :D
You have to close the laptop completely, and use on hard surface plus fan max
Honeywell PTM7950 link?
moddiy is good
cargo money?@@1syncgg
"FEELSBADMAN" 👈😂
I've heard that you want to apply liquid metal in "layers", you put on a layer then leave it to rest for 15-20 min then another layer or 2 same amount of time waited then you finally put the whole thing together, little by little and allegedly this will improve the performance of liquid metal. Its why manufacturer applied liquid metal applications suck. Would be interesting to see a test like that where you apply in layers and see if there is a performance difference.
I'll say PTM7950 are still better for laptop because of longevity and performance are the same as liquid metal, the liquid metal is conductive and risky to use especially for laptop, even with protective coating theres a tiny gap that have a chance to spill the liquid metal to other components when carrying, keep in mind on this test the performance on liquid metal is slightly worst than PTM7950 with no back cover meaning the fan is full blow yet somehow loses the score slightly than PTM7950.
@@reshiram8747 I don't disagree that liquid metal is more risky vs PTM7950, the point of my comment was that it would be interesting to see liquid metal applied in "layers" to see if it improved LM performance over that 23200 score that we saw.
Why would that help? By what mechanism? Makes no sense, so without some professional tests, I don't buy it.
@@reshiram8747Dude don't kid yourself, PTM is not as good as liquid metal, it's way worse, especially for higher powered components that give off more heat, I'm not sure what sorin did here but I'm also thinking it has to do with that Razer laptop, I'd never buy one of those things for gaming, the cooling system is never really adequate to cool the internals so no matter what paste you use and if that's all you're relying on then you'll more or less get the same results unless you just use something that's really bad then you'll get even worse results
you should try the Thermal Grizzly Carbonaut graphene pads, works very well and never have to change pads
Nah!. After less than 1 year is deforming and performs very poorly right from beginning and getting worse. Its worse than a thermal paste in any aspect. In fact i want to do a bad advertising of the product because was expensive and worthless.
PM7950 is performing better than any thermal paste i have tried so far(MX2 and Silver5) at beginning. I don't know yet how will be after 1 year because i just put it(1 week ago), but the experience looks fantastik until now. For me is a woow product since 1st minute until now. And i bought it from aliexpress, not even sure that is original. Not sure how will perform in the end, to recommend it, but i can recommend it for sure if you want to test.
Me got very hot pch chip on laptop Acer i3 4000m with 920m graphics .. if I put thermal PAD on the pch do will have better temperature???
If the chip is located underneath an heatsink then yes it will improve thermal transfer, i recommend in that case a soft thermal pad.
Liquid M. needs to get a burn-in it performs better after a few hours
Best to lap the die and heatsink
Sorin, isn't it dangerous for the cpu or the laptop in general to reach a temperature that high for gaming laptops? 98c or 99c is crazy imo
At this temperature the limits PL1, PL2 are activated and the temperature drops to ~70°C.
@@goorthiss i use msi katana gf76 12700h, the heating is a real issue with it, i have to lower the pl1 and pl2 manually to lower the temp, without it, its stuck on 95c during games
Hello Sorin, could you please test if a regular thermal pad, will cool a CPU just as well as the Honeywell pad. And I wonder if using a thermal pad on a PS5 is better than liquid metal.
PTM7950 is a compressed paste on a sheet for more convenient application. There is also a version of PTM7950-SP in a tube, only that after application you have to wait a few hours for it to "dry" before reinstalling the heatsink, and it spreads poorly.
Thermal pads or thermal putty are applied to the power section, vrams not to the CPU or GPU core. Traditional pastes in most cases at a thickness of >0.5mm lose their thermally conductive properties and a pump-out effect may occur.
You can use PTM7950 for PS5, the console will not overheat.
temps with the liquid were 30c cooler
Who's Ai?
You need to apply LM on both the die and cooler. And I think you used too little.
he did that, but surely used too little.
Not likely. I don't think either of you understand how little Tim it takes to close the air gap, especially at that viscosity
If there wasn't thermal throttling already with the Honeywell pad it is normal that there has not been an improvement with the liquid metal.
It would have been useful to see temperatures while performing the test.
Yeah, or if it is thermal throttling, time to thermal throttle would be useful too.
@@CreativityNull or even better: clock over time chart
@@lavix5 yeah, that would more accurately reflect the performance. Having both would be better since you could also see how the laptop behaves overall when reaching certain temperatures by comparing the temperature to the clock speed at certain intervals.
na, this can't be right, he should than atleast get the same performance. he did something wrong there, like not enough LM because he is afraid of shorting the gpu DIE's sorrounding SMDs, thats all.
@@lavix5Yeah because that test was way too short, should have just done a Cinebench run
The performance limiting factor is probably voltage/power. Not thermal.
But with higher temps is higher draw, no? Or is that just for circuit board components, not CPU/GPU chips?
Or it is simply the limit of the cooling system used and less thermal throttling at a given cpu package power cannot be achieved.
For me Grizzly Extreme îs the best. Compared with thermal paste that i used on a cpu, liquid metal reducess the max temperature by 10-15°C
LT-100 is even more efficient
T-1000 left chat :D
like ur accent)
liquid metal will always be better than those pads, something was done not right on ur side. let me guess, to less liquid metal because you are afraid to short the SMDs around the gpu die. just use some nail polish for those SMDs and apply more liquid metal.
👍👍👌👌
😊
Try out gelid gc extreme in laptops , other thermal pastes pump out and perform worse over time , GC extreme is thick and dense.
In gc-extreme the composition has changed in recent months is no longer as efficient as it once was.
@@goorthiss I see , i have read quite some reviews and they seem good , mine is going to arrive tomorrow , im currently on NTh1 and it pumps out starts at 88c on day 1 and 98 after a week
Who idea say PTM works better that LM in some comment? That be so stupid claim.
i think the liquid metal will show something if you apply stress on both cpu and gpu,
gpus are getting much hotter for a shorter time and thermal pad/paste would not keep up with that.
You have just demonstrated why liquid metal is a waste of time for most people. Not because it doesn't work - it is better than a pad or paste. But you have to put the right amount on, too little and you get your result, too much and disaster. It's too messy and tricky to use for most people and just not worth the money, time and effort. For the few who can use it properly it's probably worth it in the right use case. I don't have the patience to mess about trying to gain small operational advantages but a dedicated gamer could benefit if it is used in the right way. I think this is a great video to show people for this reason.
Liquid Metal performs its best on high tdp chips
I don’t remember if I saw it on gamers nexus or jayztwocents
Sorin apologized being wrong about bad performance of Honeywell PTM 7950, and wanted to outsmart us by using crappy liquid metal, yet again was proven wrong. Leave performance testing to Gamer Nexus, and do what you know electronics repair.
Plz try the Thermalright TF8.. you won't regret it i promise
it sucks
@@pegatrisedmice no it doesn't, i have tried it with core i7 11th gen H and performance was amazing.. Cpu didn't throttle and temp didn't exceed 80c under load
I used TFX. I became a rock solid in less than a year.
@@tarek3006nope, its shit. try some decent chinese paste like GD2 and u qill think the same.
If you are not thermal throttled, you compare temps :)
Hello🤝👍👍👍👍👏👏👏👏👋👋👋👋👋
not enough liquid metal bud thats why your score was bad compared to the pad
because PTM7950 is really good performance thermal pads and safe to use also last longer compare to liquid metal
It would be great to see a test with another high end thermal paste like Arctic Silver
I think all those things they sell is a marketting thing ,our paste is better no ours is better bla bla bla ,they give it to you in 4-5 grams pouchs like it is so valuable thing ,it is funny actually and a drama I think ,only important thing is the design of gpu-cpu and it's steppings are important if it is meant to heat it heats no matter what you do or it have flaw in stepping or TDP issues you cant stop it lol even their designers see the as black product and they make better in next generation ,and they have no regret no excuse nothing.last cpu that have flaw is 12.th series of intel P series cpus it have less battery life and heats ,all they do is selling tiny and thin laptops with this cpu is selling them cheap it's their excuse.
The difference in performance between low-end and high-end pastes can be as much as 8°C. Just as there is a big difference between liquid metals from different manufacturers. So this is not marketing gibberish. Thermal throttling in laptops is because they use cooling not adequate to the TDP of the CPU and TGP of the dGPU. Often they impose additional limits in UEFI so that the thin heatsink with a small number of heat pipes can handle the heat transfer. Look at tests from Jarrod'sTech, for example, as he compares different laptop designs with the same processor what huge differences there are between them in performance as well as in noise.
@@goorthiss it is about the case if case is aluminum it dissepates heat not the paste itself if they use copper heatsinks like the used to be back in 90's it is a huge plus too but no they use plastic cases "magnesium alloy they say!" and using low grade heatsinks cause that difference ,mostly aluminum case if it pricey laptop helps to dissepate heat.
This eany "Ai" is good for idots, skamers and spamers.