Silicone-based conformal coating for preventive protection of computer parts?

[aBavarian Normie-Pleb]

Hi,

I'm about to change a few PC components from standard air to watercooling and was wondering if you can "just" apply protective silicone-based protective coating to all kinds of AICs/parts that can be found in computers to extend their potential life span.

Obviously sockets/slots/contacts/wires and silicon dice that get heatsinks would be masked for the application process.

Cooling components slows down the chemical aging of parts but protecting them from the oxygen in the air or water vapor/condensation should also have a positive impact...?

Or is this a lay person's ignorant thought?

Thanks for your input, I posted this question in the repair sub-forum since I guessed users here should have the most experience with this.

[dc101]

At $40 a can it seems rather expensive https://www.digikey.com/en/products/detail/techspray/2102-12S/10440570

While it might be worth it to protect your motherboard from a leak, I would recommend doing some research on watercooling vs air cooling and see if the entire effort is really worth it.

Linus https://www.youtube.com/c/LinusTechTips has done a number of videos comparing fan/heatpipe coolers to water cooling, and in some cases fan cooling came out on top. I think I would just strongly consider if all the expense of a custom water cooling system would really be worth the extra few degrees especially when you add in the cost of trying to apply conformal coating to your components.

This is a pretty good comparison of air vs water cooling

Cheers
-Tim

[Stray Electron]

  Absolutely the best thing that you can do to ensure long life of electrical devices is to keep them cool. Any kind of coating will only block heat dissipation and will shorten their life instead of increasing it. Use large heat sinks, a good fan and keep the dust out of your computer for the longest life. Do not use an air filter since it will decrease air flow. Instead be sure that your computer is picking up good clean room air and not dust laden air. Have you ever seen the rooms that serious industries keep their test equipment in?  They're so clean that they're like operating rooms! In many cases the humans in the rooms wear what look like space suits so that the room stays ultra clean. I have 30+ YO TE that came from military contractors and it is spotless, inside and out!

  I consider add on water cooling systems for computers in the same class of nonsense as the gimmicks that the audio phools are constantly touting. If XYZ computer needed a water cooled CPU then the manufacturer would have put one on it. Go read the manufacturer's data sheet for your CPU and see what it says about cooling. If you don't think that your computer has enough cooling then turn down the AC and use plenty of fans to move the air around. FWIW when I worked in the telephone industry, all of the TelCo central offices, where the main switching equipment was located, were kept at 64 degrees day and night for exactly that reason and I never saw a spec of dust. Even the remote, unmanned sites were kept at a constant 64 degrees. Military contractors are the same but they don't keep the rooms as cold as the TelCos did. Telcos have 140 years of experience in how to achieve the maximum longevity and reliability from their electrical equipment so they know what they're doing.

[dc101]

Yup, you basically summed up that video in the last part of your post. Linus basically said that Water cooling is a waste if you're just running a normal processor and normal speeds. And that the only real purpose for water cooling is if you're doing some extreme overclocking of your system components. Which I myself have never really been a big fan of, just never seemed like the amount of speed gained from overclocking was really worth the trouble.

  Absolutely the best thing that you can do to ensure long life of electrical devices is to keep them cool. Any kind of coating will only block heat dissipation and will shorten their life instead of increasing it. Use large heat sinks, a good fan and keep the dust out of your computer for the longest life. Do not use an air filter since it will decrease air flow. Instead be sure that your computer is picking up good clean room air and not dust laden air. Have you ever seen the rooms that serious industries keep their test equipment in?  They're so clean that they're like operating rooms! In many cases the humans in the rooms wear what look like space suits so that the room stays ultra clean. I have 30+ YO TE that came from military contractors and it is spotless, inside and out!

  I consider add on water cooling systems for computers in the same class of nonsense as the gimmicks that the audio phools are constantly touting. If XYZ computer needed a water cooled CPU then the manufacturer would have put one on it. Go read the manufacturer's data sheet for your CPU and see what it says about cooling. If you don't think that your computer has enough cooling then turn down the AC and use plenty of fans to move the air around. FWIW when I worked in the telephone industry, all of the TelCo central offices, where the main switching equipment was located, were kept at 64 degrees day and night for exactly that reason and I never saw a spec of dust. Even the remote, unmanned sites were kept at a constant 64 degrees. Military contractors are the same but they don't keep the rooms as cold as the TelCos did. Telcos have 140 years of experience in how to achieve the maximum longevity and reliability from their electrical equipment so they know what they're doing.

[aBavarian Normie-Pleb]

Thank you for your responses.

While I understand your points of view I maybe should have elaborated more on mine:

Have been watercooling stuff for over 20 years and while depending on the used configuration its nice to have more performance available my primary goal is silent operation under full combined CPU+GPU load.

So my question isn't if watercooling in general makes sense but if additional coating might offer some more protection, too - without obvious potential harm.

I'm not using AIO watercooling but external solutions like these:

https://shop.watercool.de/MO-RA3-Series

When using high-flow watercooling (above 100l/h) the heat gets transferred away from the components quite quickly and other "small stuff" also gets cooled rather well over the comparatively cool PCB's copper traces they are soldered to.

[Stray Electron]

  If you intend to use a completely passive cooling system, one that relies on just the thermal gradient of the coolant and that doesn't use a coolant pump or a fan to extract the heat from the coolant, then the cooling system is going to have to be massive.

  The only thing that a liquid coolant system is good for is to move heat from point A to point B. Such as moving heat from the engine of your car to the exterior of the car. But it's always at the expense of added costs, complexity, reduced reliability and additional energy needed to drive the coolant loop. You're far better off to avoid that situation if you possibly can. The simplicity and low cost of the original VW Beetle verses the traditional water cooled cars is a great example of that idea.

[YetAnotherTechie]

It's not fully clear to me what you're trying to protect from, so i'll put the cases i consider from your message:
1. you want to protect your board in case there is some splashing, in that case you need to consider that dripping might short circuit the connectors that you didn't coat, whether it's a slow leak or a blowout. In the best case scenario, that your board was protected from splash, if you're not running any air over the board, it will run significantly hotter, many many components use the pcb as dissipator, therefore your reliability will decrease.
2. you want to protect from moisture seeping into the components, and degrading them. In this case, you should know that unless they come in vacuum sealed bags with dissecant, your components have already absorved most of the humidity they will absorv from the enviroment, this is a continous cycle, and doesn't really affect the operation of the board. If you wanted to conformal coat to protect from this, the board would need to be baked in a controlled fascion, a process that would take a few days. All the previous negatives mentioned about heat would still apply.
3. I think it was on this forum, that i read about members from very humid countries like Philipines, that do experience failure due to corrosion from ambient humidity, in that case, it might very well be more advantagous to coat, and trade the loss of reliability due to corrosion against the hotter running.

[dc101]

Water vs air debate aside, would it be more practical to use coolant that is non-conductive and then do periodic inspects to check for any leaks that might cause corrosion?

If that's not possible/desirable then I would say you're back to buying a spray-on silicone-based conformal coating, and you'll just have to see if the reliability of your motherboard is affected. I don't think anyone other than the motherboard vendor would be qualified to say definitively that a specific motherboard would still operate normally with a conformal coating applied. I did a quick google search and the most common responses were concerns over the temps of components passively cooled, specifically the voltage regulators as well as a few others and possibly changing the impedance of the board.

So realistically, I think you're just going to have to do a google search for "conformal coating motherboard" and read through a bunch of the results and determine if it's something you're willing to try.

It would be nice if you could post any updates if you do end up deciding to conformally coat your motherboard, and might help out a future eevblog forum member who is interested in doing the same thing.

Cheers
-Tim

Thank you for your responses.

While I understand your points of view I maybe should have elaborated more on mine:

Have been watercooling stuff for over 20 years and while depending on the used configuration its nice to have more performance available my primary goal is silent operation under full combined CPU+GPU load.

So my question isn't if watercooling in general makes sense but if additional coating might offer some more protection, too - without obvious potential harm.

I'm not using AIO watercooling but external solutions like these:

https://shop.watercool.de/MO-RA3-Series

When using high-flow watercooling (above 100l/h) the heat gets transferred away from the components quite quickly and other "small stuff" also gets cooled rather well over the comparatively cool PCB's copper traces they are soldered to.

[TheMG]

If you do decide to go with water cooling and your case has a bottom-mounted power supply, make sure the power supply is fan side down!

I've seen computer PSUs fail quite spectacularly when a bit of water leaked and went down straight into the PSU! Fizzle! Snap! Pop! The bulk capacitor holds around 400VDC regardless of mains voltage, makes for some good fireworks!

[james_s]

While I haven't bothered to water cool a PC myself, I can see some advantages. The point of water cooling is that you can replace a number of separate heatsinks and smaller fans with a single larger (which is typically quieter) fan on a radiator that can be placed somewhere convenient for circulating a lot of cool air through. Water also has a tremendous thermal capacity so you have a lot of capacity for short term high loads without needing to have massive metal heatsinks hanging off of everything.