Dell Latitude E6410 Cooling Fan Replacement and New Thermal Paste!

[@dougfrommaine]

Hey. My fan was running constantly, so I figured I needed to get in there and blow out the dust. I couldn't remember how the case bottom came off, and found your video - thanks! And, wow, yeah, the cooling fins were pretty much 100% blocked! By the way, I'm the original owner of my e6410, and it's been running nearly 24/7/365 since 2009 !!

[@TheBrokenTech]

Awesome! I'm glad you found the video helpful! (posted from my E6410 😂)

[@dougfrommaine]

🤣@@TheBrokenTech

[@Danielle-zx3uo]

I have the same equipment as you, I just cleaned and replaced the thermal paste of the processor, I realized that the gpu only had the Thermal Pad, but no Thermal Paste, I put a little on it and the temperatures got worse, my problem was the gpu rolled at 99°C after I put the thermal paste on it, now it does not drop below 104°C.

[@TheBrokenTech]

That's way higher than I ever see mine when it's clean. Mine comes up to the high 80Cs and stays there, and that's streaming 1080/60 content. At 1080/30, it's more like 65C. I suspect you may have a thermal pad that isn't making good contact.

[@MrAbniel]

I took out my fan for cleaning pepposess as I experience extremely heat and I realized the thermal paste melt out and dry I re-apply that thermal paste back but it never work. i decided to clean it up and assemble it back now it just switch on then off within 5sec. Do u think the lack of thermal paste cause that?

[@TheBrokenTech]

Possibly, but probably not. There was probably some other failure first that caused the initial issue. I've replaced motherboards in E6410s before. It's not a fun job, but they're pretty affordable...

[@daniellopez-qj1ts]

Hi, What is the thermal pad sizie (thickness) that you use for different chips, CPU, GPU, etc. Could you help me?

[@TheBrokenTech]

For a long time I just reused the factory pads and added a little dot of thermal paste to them. I'm still experimenting with aftermarket pads to determine the thickness. The CPU doesn't use any pad.

[@thenoxy1549]

What temp does it run at with the new paste?

[@TheBrokenTech]

Right now (streaming HD video in the background as I reply) it's at 143F on the CPU and 155F on the GPU. Thermal throttling seems to happen around 200F on the GPU and that doesn't happen very often. I've never noticed it CPU throttle, but I don't do much that is CPU intensive. Laptop thermals are always a matter of how long before it hits thermal throttling and how often it happens. If pushed hard, pretty much every laptop will eventually hit throttling.

[@ZenDeividdo]

How well is it working after 10 years? Is this your day-job comp? BTW I kinda felt your pain trying to plug the fan in as my fingers are kinda stubby and I have small hands so if I had a laptop that I were servicing myself I'd probably be going through endless tools myself T_T This was an educational experience as I legit didn't know thermal paste went bad that quickly.

[@TheBrokenTech]

Still going strong right now. I don't use it for any business purposes, but it's almost always plugged into a TV and streaming something. It's doing that right this second, actually. 😁 I have the opposite problem from yours. I'm a giant so dealing with smaller things can sometimes be challenging. If not for tools, it sounds like we'd both just be smashing things with rocks. I didn't know thermal paste went bad that quickly either. I've never experienced that phenomenon outside of the IC Diamond. I'm happy with the Thermal Grizzly for the moment.

[@tarstarkusz]

The wrong oil is ALWAYS better than no oil. The cheapest thermal paste is better than no thermal paste. hell, 3m double sided foam tape is better than just resting the heat sink onto the chip. Generally speaking, getting "the best" nearly always means spending many multiples of good enough. A $20 thermal paste in virtually all circumstances is just nowhere near 20 times better than a $1 tube of thermal paste. That ratio just climbs the supposedly "better" you go up in the quality scale. You can easily pay 40 times the price on this low cost stuff for something that is 25% "better"

[@TheBrokenTech]

Strongly agreed on all points! If I were doing this stuff at an industrial scale the costs would absolutely come into play. With products like thermal paste, people at home seem to shop at the top of the market because even the top of the market is pretty cheap. I confess that's exactly where my mind is at with it too. For ~$10 I'll re-paste 3-4 PCs with that tube, which is fine with me. You can see that I did learn a lesson by not buying the gigantic tube though... That was like a $30 tube of IC Diamond that when in the trash that I used like 2 times and it used to be "the best". I'm not impressed...

[@tarstarkusz]

@@TheBrokenTech Problem is, you're not really getting better stuff, you're getting the most heavily advertised. I've had the same tube of heat sink paste (at home) for like 20 years and have done quite a few computers with it and I think it was $3.99 in radio shack when I bought it. This is especially true at the low end. Like the difference between cables. You're paying whole number multiples of the cables that make no discernible differences. All thermal paste really does is maintain a mechanical connection between the heat sink and the heat generating chip. While some paste may be marginally better than another, I highly question the alleged efficacy difference between the cheapest acceptable quality the most expensive one. OTOH, there is a bottom upon which any further down and you're just wasting money on stuff that doesn't work at all. When it comes to the reliability of a board or something, you may very well be buying the difference between under 500 hours and 10,000 hours of reliable life and the price is maybe 50-200% more. Thought that difference can be very large, way larger than radio shack paste and whatever paste is being pushed by stores and "influencers" this week. IOW, you get what you pay for, but only to a point. You VERY quickly hit diminishing returns. Like china tools are junk and you are throwing good money after bad. Then there's good tools. Then there are overpriced tools like Snap-on. The way Snap-On tools are sold leads to a lot of losses. They make up for those losses by charging a fortune for the tools.

[@TheBrokenTech]

@@tarstarkusz This sounds like an interesting topic for a deeper dive at some point in the future. Find "the best", the cheapest, and then something in the middle and run some tests on them. Some of the results are likely going to be from the paste simply aging and drying out though... I may have to think of a way to make that a long term test with a repeatable setup. I'm off to count how many E6410s I have laying around here...

[@tarstarkusz]

@@TheBrokenTech If you have a Peltier device, you can just hook it up to a heat sink and leave it sit there on a power supply. I really don't think you are going to get much of a difference between one paste and another other than possibly drying out. The cooling of a chip, once that connection is made is far more to do with the heat sink and fan than with what paste you are using (assuming the paste is properly applied). If the paste doesn't work at all, it will cause the computer to shutdown or the chip to fail. But once it is working, I just cannot see how a paste can cause the heat sink to give up energy to the surrounding air faster than it already does. This is probably especially true in a laptop where mass is limited. If you increase the thermal transfer rate into the heat sink without increasing the thermal transfer rate from the heatsink into the surrounding air, you will just decrease the amount of time it takes for the heatsink to come up to its max temperature. This is not to say it doesn't matter all. When I first bought this computer it had a dual core 2.66ghz chip. It was running pretty hot (like 50c at idle). I replaced the chip with a 3ghz chip dual core and it ran cooler (32c at idle) that the 2.66ghz chip did, but there wasn't much compound on it. Then I later replaced the 3ghz chip with a 2.x (I don't recall) quad core chip and it runs at the same exact temp it did when it had the 3ghz dual core. They were replaced about 4 years apart. The radio shack stuff had worked fine and was still plenty wet when I upgraded to the quad core. IIRC, the wattage was fairly close. Both will heat up to about 60c when under heavy load. I think what is happening is the heat sink has enough thermal mass and thermal transfer ability that it can maintain 32 at idle, but that nothing short of a larger heatsink or a larger fan can affect the higher temp under high load. IOW, the limit is the fan and heatsink, not the paste.

[@TheBrokenTech]

@@tarstarkusz I have some experience with Peltier elements and I wouldn't trust one unattended for very long. I'm thinking this is a test that may have to go for years. I'd say the one variable you may not be accounting for in your thoughts is mass flow rate, or how fast/often the fan runs, which until right this second I wasn't thinking of either. So, what I would need to do is monitor and data log temps as well as input power. The lowest of both should be the winner.