cancel
Showing results for 
Search instead for 
Did you mean: 

noob question but VERY crucial question needs urgently answering!

Sjf2k17
Level 7
ok so after getting my cpu, motherboard and ram back i have bought some new thermal paste to obviously reassemble the computer... i bought - https://www.amazon.co.uk/d/a37/Fans-Be-Quiet-BZ001-DC1-Thermal-Grease/B00BLJGLAW/ref=sr_1_fkmr2_2?ie... - which im told is very good.

HOWEVER my main true question is, is this thermal methodr isky to use on the ryzen 1800x cpu? Before you instantly reject it please look at the product details.... it specifically says it can be used for heatsinks, cpu boards and many other cooling appliances ALSO it can withstand approximately 120-180 oC (or so ive been told).

https://www.amazon.co.uk/Thermal-Compound-Adhesive-Conductive-Cooling/dp/B072PQGKY1/ref=sr_1_1?ie=UT...

so the reason im asking is, i bought that tape because i would rather have a nice. clean. stress free time than to be messing about making sure levels are equal or no air pockets are in the thermal paste. has ANYONE tried using the thermal tape for high end cpu's instead of the patches or the paste? i cant find no articles on it nor can i find any youtube test experiments on its o im literally clueless as to whether its safe to try the tape or not? please please please read the full description of the tape before judging it due to traditional values.

*edited and added* so i found that even overclockers sell a variant of the tape with a pretty reassuring description = still im skeptic though as i still cant find no video tests using this - https://www.overclockers.co.uk/akasa-ak-tt12-80-thermal-adhesive-tape-th-001-ak.html

personally i dont think it is viable to use the tape otherwise EVERYONE would be using it, whod wanna do things messy and slow and stressfull when you can just cut a few strips and wham bam thankyou man, hey presto your done?
324 Views
18 REPLIES 18

Nate152
Moderator
The thermal compound you got is non conductive but you don't want to get it in the cpu socket.

I install the cpu, clamp it down with the latch then apply the thermal compound. I put a dab about the size of a small pea in the center of the cpu then spread it evenly over the entire cpu with an old credit card or drivers license. I do the spread method so I know the cpu is completely covered and won't smush out the sides when installing the cooler, it takes a little more time but to me it's worth it.

Take your time and you'll be rewarded with good results. I recall you said you're using the Corsair H110 GTX ? Just be sure to remove the pre applied compound from the waterblock with some rubbing alcohol, soft tissue or toilet paper, don't use paper towels as they're too abrasive.

If you have any questions before you begin please ask,

The X method is fine too, show us your results when you get your pc up and running. 🙂

Nate152 wrote:
The thermal compound you got is non conductive but you don't want to get it in the cpu socket.

I install the cpu, clamp it down with the latch then apply the thermal compound. I put a dab about the size of a small pea in the center of the cpu then spread it evenly over the entire cpu with an old credit card or drivers license. I do the spread method so I know the cpu is completely covered and won't smush out the sides when installing the cooler, it takes a little more time but to me it's worth it.

Take your time and you'll be rewarded with good results. I recall you said you're using the Corsair H110 GTX ? Just be sure to remove the pre applied compound from the waterblock with some rubbing alcohol, soft tissue or toilet paper, don't use paper towels as they're too abrasive.

If you have any questions before you begin please ask,

The X method is fine too, show us your results when you get your pc up and running. 🙂


Yeah I don't wanna mess it up again like I did last time, basically my story is did build the computer before hand so I had already gone through the process of cleaning off the paste off the cpu cooler ... yes I'm using the hydro H110i AIO extreme cpu cooler... anyways I sent my mate to get me some thermal paste from our local computer maintenance shop when I first built this rig as I didn't like the extremely thin layer I got with the hydro... so he returns with Arctic Silver 5 so I was pleasantly happy he chose a decent brand anyway I build my rig and start her up... NO issues so far then within 18 hours I started getting random system freezes going firstly to a blue screen with an error code saying something ike WATCHDOG_ERROR which lasted around 2 seconds then my monitor/tv would turn off, my system would still keep running with an error code 0 OR 8 (was hard to tell tbh) and a constant orange light lit up on the Vram part under the Q-Code LED... so I knew it was the ram that was faulty so I phoned scan and told them my issue and instead of just requesting the ram back they insisted on having the ram, mob and cpu sent back in so ok I wasn't guna argue that's when it all went to ****... as I was dismantling her I take the cpu cooler off first only to find that the thermal paste hadn't solidified at all not even remotely so the paste goes EVERYWHERE all over my motherboard, down the sides of my cpu (even when I wiped and cleaned it up before pulling the bar back and removing the cpu) it had already leaked onto my cpu sockets all over my cpu pins and in the socket itself ... I was horrorfied how it had gotten in everywhere, as I literally only used the two tiny dot method. Anyways I read the thermal tube an it says 100% none conductive so I think nothing of it clean it up as much as I can using isopropanol an ear bud and a new very fine children's paintbrush.. I completely cleaned the cpu pins and even the cpu socket was as good as new... sadly I left a few on the moherboard still when I sent them back and they could instantly see I had a leakage so then they revoked my cpu warranty revoked my motherboard warranty and sent them back to me they admitted they damaged the ram so they sent me new ram back also... so now I have to build it today and hope nothing ****s up with the paste again.. they said If I clean all the paste off they'll test them again for me.

Nate152 wrote:

Take your time and you'll be rewarded with good results. I recall you said you're using the Corsair H110 GTX ? Just be sure to remove the pre applied compound from the waterblock with some rubbing alcohol, soft tissue or toilet paper, don't use paper towels as they're too abrasive.


Paper products are less than ideal, a lint free cloth is better (read piece of cloth from an old t-shirt). You don't want to leave any residue behind when cleaning it.

On another note, you will want to note curing time. As the paste cures it improves its thermal conductance. It is best to wait to OC until it has cured.

andyliu
Level 9
huh, I never have that problem with arctic silver 5 when I build computer for my friends and they are fine.
I meant when you heat up, the paste should become sort of liquefy state and spread itself to further fill the gap between heatsink and and CPU
It shouldn't be that crazy to the point where surface tension can't keep the paste in between heatsink and CPU
at least that's how I understand it.
you sure you didn't get a knock-off thermal paste?
and ... make sure you dont apply too much thermal paste.
pea method with size of rice is enough if you are not sure how much you need.

Eleiyas
Level 10
Judging by your last post, I can only conclude 2 things (and I may be wrong here):

it sounds like you dismantled the PC while the components were still warm/hot - this might explain (why the paste was more liquid than it should have been. Also, if you only put the paste on, then the day after you dismantle it, then the paste hasn't really had time to ''get settled in'' (if you understand what I'm getting at).

gupsterg
Level 13
I have interchanged 3 R7 1700 on same C6H, ~12 times.

I use AS5, I apply a grain of rice amount and spread across IHS using a plastic card. So far contact has been good with HSF I have used. I also run a stress test to warm the CPU up prior to swapping CPU. Reason being that if the TIM is not warmed up the cooler can be difficult to remove. As AM4 is not LGA you can literally rip the CPU from socket.

I have had no TIM over spill on fitting or removal.

Not say others that have had an issue are wrong, just sharing what has gone on with me.

I have noted with Ryzen CPUs some are flatter than others, some have a slight dimple in the middle of IHS. My guess from solder process of die to IHS. I would advise at least one mount and removal of cooler to gauge TIM spread/contact and then user is aware how much TIM to use. Here is a post on OCN where I made a mistake. As seen in the lower images the sample CPU had a greater dimple IMO than another Ryzen I had. As the HSF base is catered more to Intel IHS IMO, which IIRC are generally concave, HS base convex near centre, it exaggerated the issue, so more TIM was needed in that case.
Intel Defector :eek: AMD Rebel


R9 5900X - Custom WC - ASUS Crosshair VII Hero WiFi - Ballistix Sport LT 2x16GB 3800MHz C16 - RX 6800 XT - WD SN770 2TB - 2x 870 EVO 4TB


24/7 OC: i5 4690K @ 4.9GHz CPU@1.255v 4.4GHz Cache@1.10v - Archon SB-E X2 - Asus Maximus VII Ranger
Sapphire Fury X (1145/545 ~17.7K GS 3DM FS)

:eek: CPU Validation 5.198GHz@1.314v with 4.4GHz cache + RAM 2400MHz@1T :eek:
Da Music video

Korth
Level 14
I also prefer Arctic Silver 5 for most CPU installations. Maybe not the very best of the best of the best but it's always counted among the top performers and it's cheap and common and - unlike most other TIMs - an application of AS5 seems to last at least a few years before it cooks off.

I also use Prolimatek PK-1 and PK-2 (I still have big jars full of the stuff) and Gelid GC Extreme (because it's so excellent for some things) and a Shin Etsu X23 (which we use for so many things at work). I sometimes even mix my TIMs together to "engineer" something more ideal for a particular application. I always mix my TIM up evenly (by vigorously shaking the tube, if nothing else) and heat it up a bit (stick the tube into warm water) before application. I scrape it across CPU parts with a razor blade, similar to the credit card method described above. Too little is far better than too much, if you have to wipe off any TIM that's leaked out of the edges then you'll find there's always "thick" spots and "thin" spots and bubbles/gaps hidden between the parts which can run too hot.

Phobya HeGrease is apparently quite good, too. And IC Diamond, although it contains abrasive microparticles which can apparently (at least over time and over repeated applications) actually scratch up the polished surfaces (and part markings) on processors.

Any premium branded TIM is going to outperform any cheap/generic TIM. And the differences between 99.5% or 99.55% or even 99.75% thermal efficiency aren't worth obsessing over unless you're going for some sort of world record overclock ... especially since there's so much controversy, bias, misinformation, disinformation, exaggeration, elaboration, and hype about which of the very best top performing brands is indeed the very best top performing brand. And one thing all those comprehensive lab benchmarks don't measure or compare is TIM longevity - I know that AS5 can last years and Gelid or X23 can last months, but I don't know how most TIMs compare in the long run - an important parameter for those of us who don't plan to remount CPUs with great frequency.

TIM coverage is far more critical than TIM conductivity. Low viscosity (thin and runny) TIMs are better used when the gap between mating part surfaces is extremely tight or interlocking part geometries makes TIM application inaccessible/difficult. High viscosity (thick and pasty) TIMs are better used when the gap between mating part surfaces needs a little bit of "filler" to maintain efficient thermal contact. Thermal pads and glues should only be used when a bulk material is needed to conduct heat across larger (air-filled) gaps and/or to provide some mechanical adhesion (structural stability) between parts. In fact, thermal pads are generally used only in situations where solid-state TIMs (like silica, mica, elastomer, HOPG, etc) cannot be used instead.

I haven't installed any Ryzen parts yet so I admit I'm unfamiliar with their specific surface features.

Other CPU parts often have slightly convex or concave surfaces. Designed to be paired with coolers which have perfectly matched slightly concave or convex surfaces - to provide greater cross-sectional area for heat transfer between the two surfaces. Although the majority of coolers are not exclusively designed to perfectly fit exactly one processor, they often have "perfectly flat" (and slightly-oversized) baseplates which are designed to accommodate a wider range of sockets and fittings and processor types. If the processor and the cooler don't "match" perfectly then a mid-viscosity TIM (like AS5) is an excellent choice, although some experimentation with different (or mixed) TIMs might be required to find an optimum heat transfer.
"All opinions are not equal. Some are a very great deal more robust, sophisticated and well supported in logic and argument than others." - Douglas Adams

[/Korth]

gupsterg
Level 13
+rep 🙂 . I likely 🙂 .

Besides IHS variation on flatness I have noted HS base variation as well between like HS. I now resort to getting say a metal ruler and placing it's edge on CPU / HS and getting an idea of their flatness. And also like said before at least one mount and dismount prior to final mount.

I have found AS5 good on longevity for application and storage. Easy to spread. Several times seen on forums peeps highlight separation within the tube, I have noted none. I have had 3 tubes so far IIRC, in 10 yrs IIRC. None have a production date stamp, so no idea how long they were in supply chain, etc. I store them upright.
Intel Defector :eek: AMD Rebel


R9 5900X - Custom WC - ASUS Crosshair VII Hero WiFi - Ballistix Sport LT 2x16GB 3800MHz C16 - RX 6800 XT - WD SN770 2TB - 2x 870 EVO 4TB


24/7 OC: i5 4690K @ 4.9GHz CPU@1.255v 4.4GHz Cache@1.10v - Archon SB-E X2 - Asus Maximus VII Ranger
Sapphire Fury X (1145/545 ~17.7K GS 3DM FS)

:eek: CPU Validation 5.198GHz@1.314v with 4.4GHz cache + RAM 2400MHz@1T :eek:
Da Music video

Korth
Level 14
IHS surface geometry is usually explicitly defined somewhere in the Intel (and presumably AMD) literature, "thermomechanical specifications" or somesuch. But (from what I've seen) there's somewhat wide margins and variations between individual parts.

HS surface geometry is basically never specified. Coolers advertise compatibility with LGA1151 and LGA1150 and AM3+ and whatever, but aside from whatever overlapping mechanical features the collective bunch of sockets share in common there's no guarantees. And who knows how stringent any specific company (Noctua or Bequiet or Cryorig) defines their tolerances, or if they even bother to examine every piece they outsource from their mysterious Made-In-Asia manufacturing partners.

"Flatness" to a machinist is a collection of very exacting technical metrics and tolerances. "Flatness" to mass-produced metal parts is far less consistent. "Flatness" to the marketing people is more of an approximation, lol.
"All opinions are not equal. Some are a very great deal more robust, sophisticated and well supported in logic and argument than others." - Douglas Adams

[/Korth]