SEEKING: Thermal Guru | Metalurgist | Extreme Custom Modder | Scientific Opinions on Heatsinks/Pipes

Yeah since NBR moved to the new forum software (XenForo), any thread older than a certain age is automatically closed.

 

On a desktop CPU with an IHS and truly horrible stock thermal paste (the norm) the improvement is generally 15-20°C using liquid metal underneath and on top of the IHS. That was my experience with 4790K, 6700K, 7700K and 8700K processors. The improvement is sometimes not as pronounced on a lidless mobile CPU if the heat sink is already making good contact and an excellent non-metal thermal paste is getting the job done well enough. But, a 4°C temperature reduction also isn't anything to snub your nose at or dismiss as insignificant if it means the difference between experiencing a thermal throttle or thermal shutdown. In those circumstances, a 4°C temperature reduction is super important.

 

but 99% of the time mobile lidless cpu doesn't even have contact due to tripod design

 

Well... that's not going to benefit from liquid metal either. That's just an engineering abortion that needs to be returned for a refund, LOL. There are turdbooks and then there are jokebooks. That example is both.

 

dell refundware 17r4

better not buy in the first place, waste of time

 

If only everyone knew that before spending their money. Alienware was a great brand for a long time, and for good reasons. They are still riding on the coat tails of that reputation.

 

You forgot it's sister 15r3

 

perfection in practice


oh wait...

 

That's pretty crazy they use tripod designs in Alienware laptops but implemented a square heat spreader design in the lowly Dell 15" Inspiron Gaming 7577... unless I am mistaken here?

7577 Mainboard 1:
http://forum.notebookreview.com/attachments/screenshot_20171201-164801-png.152383/

7577 Mainboard 2:
http://forum.notebookreview.com/attachments/screenshot_20171201-165022-png.152385/

7577 Mainboard 3:
http://forum.notebookreview.com/attachments/screenshot_20171201-164950-png.152384/

 

Square 4 screw, but only 2 of the screw is exclusive to either components, 2 middle shares same screw hole.





In theory it could be done but the problem is that the height of cpu area is significantly thinner than gpu area. I'm talking about the die's Z-height and other supporting components.

 

Forgive my ignorance here, but what is so bad about the 2 center holes being shared between the two? I think I'd rather have that than a tripod design, no?

 

It's better but each section needs different pressure.

It COULD be done right and have everything balanced but with Dell's track record on alienwares I don't think that might be the case. It's always better to not mix up the pressure applied etc.


Does it work better than tripod? I can't say for sure, I haven't handled this laptop personally. But given the low heat output of the 1050ti it might be ok on the GPU side.

 

As you probably know... Alienware's ceo Azor is also VP General Manager for Dell's other gaming brand, the well known throttlemonster XPS.


Dell and Azor prefer the tragic TRIPOD Heatsink design... Alienware or XPS, it doesn't matter.

Spoiler: ITS'S ALL DELL NOW!!

 

The 7577 sports the GTX 1060MQ, but I see what you are getting at. The different pressure requirements could easily be adjusted with proper shimming I would think... or possibly even tab bending/tweaks. But the shims should yield the most square-fitting result I think.

 

Bending don't necessarily work because if heatsink is hot, metal is more malleable and will form back to whatever shape it originally is.

 

I think I am missing what you are getting at here... The additional screw is better than a straight-up tripod design no? I agree it looks like a cheap way of "squaring" it up but it seems better than not having it to me.

 

TRIPOD is a awful heatsink design. Smaller and smaller Motheboard due thinner chassis design (need the place for battery) means no space for all needed screws. And I have not mentioned savings on cost due less screws. Perhaps the lower cost is instead spent on more optimized software?

 

Better but not the best these companies used to do. While they charge more at the same time.

 

Yes I agree I don't see the absolute need for ultrathin ultrabooks and notebooks like everyone goes crazy for. I don't think tech needs to be forever miniaturized. I am pretty happy with the form factor of the 7577 that many people like to gripe about. I think it is a reasonable size to performance ratio myself. It has a better cooling system than most notebooks I have ever had my hands on. But then again my standards aren't super crazy high like many hardcore gamers.

 

ultrabook should also have 4 screw mounting design, no compromises


thinkpads are good example, though expensive

 

The question is, how much of the ****ty designs stems from greedy bottom lines and corporations saving money versus inexperienced rookie engineers fresh out of college that don't know any better and have no actual *practical hands-on experience*? lol

 

Dell and AW's ceo Azor talk nice about thinner laptops, but forget tell you that the footprint was increased for Alienware's due needed cooling for thinner design. But the expanded cooling grills in the rear end in the thin flimsy laptop design, doesn't necessary mean better cooling as you can see. It's only so big MB you can put in thin and flimsy. No space for screws anymore. And next Intel package will probaly increase in size and need it's space on the MB. Maybe we can see glued on HS or TWOPOD for next refreshed model?

More pipes or more HS metal doesn't mean a ****y, if it's attached with less than minimum amount needed screws. It's doomed to be uneven/ warped when it's finish attached.

 

While I've got everyone here, does anybody have any input to give on spot welding (AKA: resistance welding) copper to aluminum, copper to copper, or aluminum to aluminum and the pro's/con's on thermal conductivity? I follow some serious ebike groups online and have a lot of insight into high power drive systems which are also quite demanding on the hardware and rely heavily on electrical and thermal properties of all the parts and pieces in play.

It is well known that spot welding nickel battery pack tabs is the superior way of constructing a pack compared to soldering due to less electrical resistance than soldering. I am assuming that translates equally to thermal resistance as well. (If electricity can flow easier, heat should be able to flow easier too right?) I'm just curious why more spot welding doesn't happen with modding laptop parts. It seems to fit the bill perfectly for fabricating custom bits and pieces with superior electrical/heat transfer properties and eliminating the need for drilling/tapping/screws/etc depending on how and where it is being applied.

 

already been done before->
https://www.waresphere.com/store/he...1854-00hm903-at0sq001dt0-0c53568-0c53564.html

 

Maybe if they were not trying to cater to all of the children that are obsessed with thin and light trash, things would not be as pathetic as they are for the rest of us.

 

Put it the other way, Fox... If they changed to Thick and Heavy trash... This wouldn't help Dellienware or the other OEM's JukeToys. It doesn't matter what size it have!!

 

True. That's what happens when a company is no longer "in it to win it" so to speak. Seems like none them are now, since almost everything is a jokebook regardless of brand. I guess I am just more sensitive to the fact that Alienware went from being winners to losers almost overnight. One of the main decision-makers must have had an aneurysm, brain tumor or something. They had a good thing going, and then just threw it into the toilet. That's why I am harder on them. They knew what the right thing to do was and they chose not to do it. Most of their competitors have maintained status quo and never were doing the right things in the world of laptops. Maybe things were too lonely at the top. As they say, misery loves company.

 

Msibook has models still with MXM graphics, same for Asus (proprietary design) I think. Dellienware walked the extra mile and also throw in graphics as BGA over night. Maybe they will be the bell sheep for soldered Wifi as next high end. Yeah, they are worse than all the Big Jokebook makers.

 

ASUS never used standard MXM and they were the first ones who rode the BGA shaft when it was first offered by Intel.

ASUS is scum on the notebook side.











@Mr. Fox @Papusan @bloodhawk


4 screws but 1.5 screw on gpu: https://forum.51nb.com/data/attachment/forum/201712/01/161225h1yusz31y1730sqx.jpg

Thinkpad 25 Anniversary ed (CN): https://forum.51nb.com/forum.php?mod=viewthread&tid=1812699&extra=page=2

 

I mean... how much would it bloody cost to add a second on AT THE VERY LEAST?

 

too much for the investors

 

Yeah, Asus has always been trash even before the handchace with Intel.
Asus in fully flavor with Tripod (4 screws for the whole cpu/gpu HS package).
Missing Heatsink screws is one thing, but why throttle the processor although the 45w i7 BGA chips barely run with 15w Cpu Package Power Load?

 

Part of the problem is the people that build the trash. Part of the problem is the people that buy the trash. You cannot sell garbage unless people are dumb enough to buy it.

Example: ASUS NovaGo hands-on: Your laptop is now a smartphone

No wonder everything is such a horrible mess. Thanks, but no thanks.

 

man I could really use a glass or two

 

Use the same method, only with an inverted baggie; very precise and it has a clean surface. For the non-metal TIMs as well (in case of aluminium heatsinks).

It's really because soldering batteries the regular way is next to impossible. Only way to get a decent join is by sanding them first, but it's still not nearly as solid as when using a spot welder.

Not sure how this can work for situations that require heat transfer as well as electrical conductivity. My (dirt cheap*) spot welder is meant for batteries and only fuses two tiny pecks. This is more than sufficient to carry the amps of an 18650 cell, but is not a fully solid-on-solid connection and thus has minimal potential for heat conductivity. Another hurdle is that it works by (current-induced) resistive heating and in order to deliver those amps as effectively as possible it resorts to copper tips. These tips will heat up too, but by design not as much as the metals that are to be joined; steel and aluminium will have higher resistance to the current, so these can be fused together without the copper tips melting first. This means that spot welding copper is inherently impossible unless you use a tip with higher electrical conductivity than copper (silver, for example).

Spoiler: Cheapness ...

*) ... made possible by lacking inrush current limiters, meaning it charges its capacitors with a crazy peak current before tapering down. Always have to remember which socket to plug it in to because it will trip the breakers when used on the 'wrong' circuit.

 

Yes I was reading up on spot welding copper and aluminum in particular and saw that special composition electrode tips would be required in addition to massive capacitor banks.

The DIY example I saw used city bus electric brake regen capacitors to construct his bank and I think he mentioned tungsten electrodes.

As for the spot welder, I would start with an E-toE dual MOT configuration and the fattest wire that can be wrapped 0.5 to 1.0 times basically. Not 100% sure how the capacitors would need to be worked in just yet. I wonder if drill bits of a particular composition might be well-suited to the task... Or possibly diamond coating a tip similar to the way diamond and carbide cutting tips work.

Best case scenario for powering this would have to be from a 50amp electric oven plug for us residential dwellers. I would need to look into what it would take to have multiple transformers in series/parallel potentially as well.

I think it is at least fun to think about this stuff.

 

I want to reduce my battery by 66% (6 of 9 cells) and move my HDD in its place

Then place a fan in the 'then' unused HDD slot.

Would all be pointless though If I cant mold another thin but large area fin array, with attached heatpipes above the GPU to exhaust out the center rear exhaust area I would make.

Once I finish school I am going to try it anyways. I need to figure out a way to cool the VRMs on my laptop and personally I want to push 4.5Ghz+ daily overclock.

Probably will take some time regardless since I have to test measure everything and find a fin array wide enough and with enough fin density that I can cut down evenly

I dont need a large battery. Only enough to hibernate/sleep or not immediately shut off when unplugged accidentally.

 

That sounds like a great custom mod project. Which system is it happening on?

What is the construction of your current stock battery pack and what is the maximum amperage it needs to be able to push out at peak? Is it pouch cells or 18650 cells or custom from factory? I assume it's the standard 19.5V output? It might be hard finding a way to fit a proper BMS to make the battery mod work considering how locked down all the proprietary hardware is. Also your C rate might turn into an issue with such a low cell count on a high demand device.

Are you thinking bonded fin heat sink or extruded plate fin heat sink and do you plan on tying it into the rest of your heat sinking system via heat pipe or just letting it stand alone isolated?

As for cutting down a fin set, do you have a specific method in mind? My first two thoughts are Dremel with cutoff wheels and/or freezing the fins in a block of water then using something of your choice to do the fine cutting. A band saw would be ideal I think, or possibly some kind of diamond wire coping saw?

Keep us updated. Would love to see something like this come to life.

P.S.) This place has some good reading on the matter.

https://www.aavid.com/product-group/bonded-fin

 

This would be on my Alienware 17 Ranger.

All those questions are why I am waiting to begin with modding, I am pretty ignorant to a majority of those questions.

As for the battery I would try to prevent the high demand on them by having a battery profile to use when battery is for what ever reason in use. But that probably doesnt resolve configuration issues.

As for the fins I would probably have to look through a number of different ultrabooks and see if I can find a fin array thin enough, and dense enough. Then buy like 6-7(maybe more like 4-5 of them to run in parallel. With HVAC tape I would also be able to have a bit of flexibility instead of a strict line. Stacking would be much easier, but I dont know if this would displace heat well enough.

The other question however would be the contact plate and how I form the heatpipes. Would I just stack the additional pipes on the vanilla installation or split the pipes so that its 4 pipes on one set of fins and 4 pipes on another set of fins? This is all assuming I can mold pipes without breaking them of course.

I dont have tools and dont want to purchase any as I am always having to leave them behind when I leave the country, tired of replacing them.

Then there is also, should I try for 2 smaller fans, use one for the fin array and other for moving air off the components or just one larger fan for the additional fin array and leave the components to themselves as is.

Maybe put the left heatsink assembly in parallel

https://www.ebay.com/itm/Razer-Blad...NcAAOSwnTdaF9hl:sc:USPSFirstClass!76002!US!-1

 

https://www.ebay.com/usr/cicichen594?ul_noapp=true

I had a short conversation with him not long ago, English is not his first language but he communicates well enough to get the point across.

He does manufacture and offer a number of aftermarket products mostly for alienware that I saw (purchased both GPU and CPU iterations for my Ranger) but if the interest is high enough and of course if the money is right I wouldnt be surprised if he was able to manufacture better heatsinks for other laptops.

I couldnt say what his official position on the matter is, but he stated he designs and manufactures heatsinks, so at the very least, might be worth an inquiry...

Its anecdotal but my experience with him thus far has been positive, a number of other Ranger owners also purchased these heatsinks.

@Maleko48

After I wrap up school I may contact him again and see if he can fabricate a heatsink of a different design provided the price is right.