Close
0%
0%

Laptop Cooling Enhancement Project

A Simple Project Utilizing common and novel technology and techniques for solving laptop overheating issues without Active Cooling systems.

Public Chat
Similar projects worth following
As a fan of the Dell XFR E6400 and similar laptops the most common issue for squeezing every ounce of processor power out of one is cooling. I rapidly found that compiling and multitasking on ANY laptop with or without a VPS adjutant is a little outside of their wheelhouse. Not being one to appreciate the nonstop cycle of upgrading and throwing away a perfectly good laptop after I like one, I utilized a series of Intel T9900X processors in my Dell XFR E6400's which I happen to like very much... just one problem with running 3-4Ghz on a laptop... The HEAT.... (and the battery power, but that's another issue altogether) Having a great deal of experience with thermal transfer, I set out to utilize simple techniques to achieve maximum results to increase the thermal transfer of existing heatsinks without making major modifications or utilizing major active cooling solutions that I use in more sedentary PC and Server ventures.

In thermal management there is a simple principle...  moving heat.


First up is the heatsink itself.  The heatsink is a rather common design and includes an oil filled pure copper heat pipe permanently mounted to an aluminum and fill metal attachment grid that is mounted under spring tension coupled to both the cpu and gpu with thermal paste and/or a silicone thermal pad.  Usually a thermal pad for the gpu and thermal paste for the cpu.  The heat dissipation matrix is comprised of an aluminum hollow core box type radiator that the fan blows through to remove the heat from the assembly.

Lets get a basic understanding: Copper has a higher thermal conductivity, and therefore is superior to aluminum in processor cooling.  Copper is better at conducting heat than aluminum, but aluminum is able to radiate the heat into the air better than copper because of its lower density

Lets take a look at the numbers:   

Thermal Conductivity: λ / W·m−1·K−1  (Pure Metal)

Copper‎: ‎386

Brass‎: ‎120

Aluminium‎: ‎237

But we must remember densities: 

metalg/cm3lb/in3
aluminum2.700.098
zinc7.130.258
iron7.870.284
copper8.960.324

Now let us think of the "skin effect" of a 3-6 um coat of copper atop the aluminum substrate to allow the heat to transfer through the fins and "amass" at the very edge to be removed by the incoming cold air blown through the fins by the fan.  In short, a tiny layer of copper would increase the heat dissipation.... this leaves a conundrum because it is difficult to plate aluminum with copper due to a number of reasons.  The usual process is to first plate it with Nickel, and then plate copper atop that harder Nickel surface, however in this instance the Nickel would reduce the "skin effect" and heat transfer that we are looking for.

Method:  There are multiple methods of electroplating, but taking the aforementioned factors into account the option was quite clear:  A novel and highly effective electroplating method that utilizes peracetic acid to slowly electroplate a pure copper "skin" utilizing the distance between anode and cathode to slow the plating process to facilitate strong bond between the aluminum substrate and the deposited (electroplated) copper.  

Materials:

First, Source a PURE COPPER scrubbing brush; Make absolutely sure that it is not a copper plated metal brush but truly a PURE, SOLID COPPER "brillo" style pad...  you can use other forms of copper, but the hassle of reducing them to a surface area that will be useable for electroplating is not worth it.  -(I recommend ebay) **Remember**- Any contamination with another metal will effect the outcome negatively... even the short time the metal to be plated comes in contact with the solution before plating occurs can poison the results.  

Secondly, source white distilled vinegar and standard 3% H2O2 (Hydrogen Peroxide) -(I recommend Wal-Mart or equivalent)

Thirdly, source a large enough container made of plastic, nalgene, or glass (preferably with a top to protect the remaining solution for reuse)

Fourthly, You'll need a 6 volt battery and wires with alligator clips for your anode and cathode.

And Obviously - your heatsink (or other item to be electroplated)

Instructions: 

Common sense applies.. here are the important details:

First carefully remove the heatsink and clean with soap and water fully before cleaning every inch with IPA (Isopropyl Alcohol) **70% or higher is best.  Pay close attention to the area that contacted the gpu and cpu to remove any trace of oil, adhesive, or thermal compound which is generally a mixture with a silicone oil or mineral oil base depending on the brand / type... even silicone pads are impregnated with silicone oil to aid in contact transfer so clean this area perfectly...

Read more »

  • 1 × Container Container with lid made of Glass, Plastic, or Nalgene
  • 1 × Battery 6 Volt Battery
  • 2 × Wires Copper wire with Alligator clips on one end
  • 1 × Vinegar 1 Gallon of distilled white vinegar ( ~8% acetic acid)
  • 4 × Hydrogen Peroxide Pint of Hydrogen Peroxide ( ~3% H2O2)

View all 8 components

  • Initial Results

    zackf6911/28/2018 at 05:15 0 comments

    UPDATE:  Average temperature charts show  a 5-10 degree Celsius decrease in CPU Temp.     Prior to modifications the CPU would hit the 80 degree Celsius emergency shutdown point multiple times a day...  now average temps yield 58-65 degree Celsius and up to 72 degree C when compiling with both cores overclocked.   Though I should publish charts et al, the whole purpose of this simple project was to K.I.S.S. my way to an upgraded cooling system for laptops...  Success.

View project log

Enjoy this project?

Share

Discussions

KD wrote 06/27/2019 at 15:14 point

Would this work for desktop coolers as well? I can see this being a huge boon for the SFF community.

  Are you sure? yes | no

matthewkleinmann wrote 11/28/2018 at 04:40 point

Are you sure the heat pipe in notebook heat sinks is filed with oil?  I had thought that they were filled with ammonia or alcohol.  The electroplating sounds interesting.  I may give that a shot.  I have a couple dead laptop heatsinks to play with.  If you mix up another batch of your heat sink goo, I would be interested in trying it.

  Are you sure? yes | no

Martin wrote 11/28/2018 at 11:35 point

Ammonia would be very unlikely for standard heat pipes as most heat pipes are made of copper. Copper is not compatible with ammonia. AFAIK alcohol or water are most common, no oil. The filling has to be able to vaporise at the given temperature and pressure.

  Are you sure? yes | no

zackf69 wrote 11/28/2018 at 15:42 point

You are correct Martin.  Here is a fine example of new tech that is utilizing top tier tech standards... (sintered core and channel wicks) https://bit.ly/2RhkYd6    Most laptop heat pipes still utilize copper braided wicks and alcohol/water mixes... As I mentioned in my other post I may add another list on experimental coolant fills.  I believe an alcohol/water and ethyl chloride blend would work much better, but good old fashioned oil/water fills and mineral oil fills are not uncommon...  most manufacturers "cheat" on the fill amounts and substances.  Even when they are made correctly you'd be lucky to see them damp if you cut it open... not an oozing fluid fill unless it has been properly filled with an oil / solvent differential fluid (either mineral oil/silicone/ or perfluorintaed hydrocarbons)... particularly for medium temp 60-80C applications like cpu/gpu cooling.

  Are you sure? yes | no

zackf69 wrote 11/28/2018 at 15:32 point

Good Question.  It depends... usually modern heat pipes are a derivative of NASA heat pipe technology.  At first it was a braided copper core utilizing various fluid such as ethyl chloride / alcohol water mix for temp stabilization in satellites... now they use copper wicks or sintered copper channels filled with powdered copper wicks and filed with water, water/oil or eutectic mix wicks and various fills depending on the prime temp gradiant.. I postulate that a better wick with an Ethyl Alcohol and ethyl chloride mi might increase efficiency, but ammonia and copper is a no no and stainless steel or internal stainless/nickel blends would hurt thermal transfer.  As for the nano thermal paste I may very well do so and add the results to this project.  PM me and I'll share. (Not Shilling here), but I lost my butt selling a few test runs on ebay...  I sold WAY too cheap and 30 or so got "lost in the mail" which completely ate up my 12 cent each profit margin after shipping snce it cost a minimum of $1.30 for each piece in shipping alone.   I still believe that it is worth it to do if only for specialists and satellite use / etc...  The current products on the market super suck and there's no excuse to not be able to get top tier tech products at a reasonable cost.  Right now if you don't make them yourself, you can't get them at ANY price.  

  Are you sure? yes | no

equinitry wrote 11/24/2018 at 09:35 point

i doubt that copperfying metal parts will increase the thermal conductivity significantly...besides u need to polish the contact plates or maybe all.

would love to see a micro liquid system xD

  Are you sure? yes | no

zackf69 wrote 11/25/2018 at 04:47 point

Actually I found that by applying a thin plate of copper on the heat transfer fins you get a "skin effect" of heat transfer that significantly increases the heat dissipation without other major modifications.  (I always polish my contact points with nano diamond thermal paste and microfiber brush prior to applying my own thermal paste upon reassembly) - I agree however...  my next step up would be a semi liquid capillary cooling system with Plasma Vapor Deposition metal coatings for passive cooling and a micro TEM / micro FreePistonStirlingCooler hybrid system with CVD enhanced contact plates for active cooling... but those are expensive and not very friendly to consumer budgets...  This is something that you can do that works well and won't break the bank.

  Are you sure? yes | no

Jan wrote 11/24/2018 at 08:47 point

You could go the Apple route as well: Set the fan-on limits so high, the components will fail after a few years doing more than writing texts at Starbucks :)

Typing this on a Dell Laptop (and having opened a few models over the last few years) I'm looking forward to your techniques and findings!

  Are you sure? yes | no

zackf69 wrote 11/25/2018 at 04:48 point

: )   Yes but it's QUIETLY frying the chipsets....  you sure you're not holding it wrong?  LoL.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates