[SOLVED] AIO Failing

[dande691]

Hi, i just wanted to ask here too, im pretty sure my aio is dying but i wanna be sure so this is the situtation, my fans turned to full speed and temps were at 80C while playing osrs, i shutted down my pc and cleaned the aio, changed thermal paste and checked all wires, after starting the pc temps are beetween 70-80 at idle, after some time the pump stops making noise and temps get to around 50 but after starting alsa its goes to thermal throttle pretty quick, i have 6 fans on my pc, 3 out and 3 in and i tried this with case open too, could it just be a big air bubble or is it dying?.

Edit: my cpu is I7 4770 and aio is Silverstone PF240-ARGB, the radiator is placed at top of the case and i have tried to rotate my pc to every side about 45 degrees and wiggled pipes.

 

[kerberos_20]

aio has lifetime about 2-3years
if you get air bubble (big one) in cpu socket, then pump will die within seconds
post picture of your pc insides

 

[dande691]

aio has lifetime about 2-3years
if you get air bubble (big one) in cpu socket, then pump will die within seconds
post picture of your pc insides

Hey, thanks for the answer, i have already removed it from the case and dont really have a good photo but check SilentiumPc Armis AR6X EVO TG you get an idea how it was placed, pump was at the lowest point so i suspect corrosion, also its 1,5years old and been sitting for 6 months before using it, also that air bubble/pump dying in seconds is wrong in many ways but i wont get on that for now.

 

[Phaaze88]

The corrosion inhibitor in the fluid should be good for a few years, but there's other ways for a pump to fail:
Mechanical - pump dies on its own. Can vary based on the parts the manufacturer used, and how you use the unit.
Thermal - friction from being exposed to air for extended periods of time.
Chemical - the already mentioned corrosion. Even when the inhibitor expires, corrosion doesn't kill it out right. It takes time.
Biological - gunk/algae forming in the loop. There's biocides in the premixed fluid, but that also has an expiration date like the corrosion inhibitor. Also doesn't kill the pump out right; the organic material needs to build up first.


How's the mounting hardware and backplate for that cooler?
Some coolers come with crappy mounting hardware or very flimsy backplates, making mounting the cooler more difficult, or backplates easily having wiggle room when it shouldn't.

 

[dande691]

The corrosion inhibitor in the fluid should be good for a few years, but there's other ways for a pump to fail:
Mechanical - pump dies on its own. Can vary based on the parts the manufacturer used, and how you use the unit.
Thermal - friction from being exposed to air for extended periods of time.
Chemical - the already mentioned corrosion. Even when the inhibitor expires, corrosion doesn't kill it out right. It takes time.
Biological - gunk/algae forming in the loop. There's biocides in the premixed fluid, but that also has an expiration date like the corrosion inhibitor. Also doesn't kill the pump out right; the organic material needs to build up first.


How's the mounting hardware and backplate for that cooler?
Some coolers come with crappy mounting hardware or very flimsy backplates, making mounting the cooler more difficult, or backplates easily having wiggle room when it shouldn't.

Its quite good, solid, easy to adjust, does not wiggle or move, never had any problems with it, also been performing really well w/fx8300 @5ghz, now ive been using it with I7 4770 and never had any problems until now, it does pump some fluid and i can feel the pump doing its job but the radiator wont heat up so something must be slowing the flow, thats why i suspect corrosion, few have had some problems with corrosion when it comes to Silverstone aio's.

 

[Phaaze88]

Radiator not heating up can mean:
-fluid flow is blocked/clogged.
-pump is dead.
-heat not actually making it to the cooler cold plate.
-impeller has come loose from the shaft.

Most AIOs are going to have problems with corrosion at some point, because most of them are mixed metal(copper cold plate + aluminum radiator).
Then there's Alphacool's Eisbaers which is all copper.

I think it's too early to write in corrosion. You've only had it 2 years at most, and it takes time for the damage to come in even after the inhibitors have expired.
Perhaps Silverstone's using some 'dirty water' like Enermax did with their Liqtech I and II models?

 

[dande691]

Its little too early but it does feel like the impeller is on and it does pump a littlebit, maybe the water ain't the problem but anti-corrosion part of the fluid or mixture, kinda sad to see Silverstone fail like this because i have never had any problems with any Silverstone part, anyways im gonna rma it tomorrow and in few days we'll know, ill be back, also thanks for your answers.

Edit: What do you think about ek-aio basic 24? i could get one for replacement.

 

[Phaaze88]

What do you think about ek-aio basic 24?

~Ehh, not much to say about it at first glance. It's 'another AIO'. There's not much separating them these days - the biggest differences being the fans used and any aesthetic accents.

The pump's an in-house design.
It's a mixed metal AIO.
Those Vardar S fans are on the cheap side. Sleeve bearing fans, thus will last longer if they're installed vertically - in other words, mount the AIO in the front.


My personal opinion though: 240mm AIOs don't have enough going for them. If your chassis can support either a 280mm AIO, or an air cooler that sports 140mm fans, 240mm AIOs sorta fall off a cliff.
Those other coolers end up having better performance to noise profiles because of the 140mm fans.
But if the chassis doesn't support either, then I suppose a 240mm will have to do.

 

[dande691]

~Ehh, not much to say about it at first glance. It's 'another AIO'. There's not much separating them these days - the biggest differences being the fans used and any aesthetic accents.

The pump's an in-house design.
It's a mixed metal AIO.
Those Vardar S fans are on the cheap side. Sleeve bearing fans, thus will last longer if they're installed vertically - in other words, mount the AIO in the front.


My personal opinion though: 240mm AIOs don't have enough going for them. If your chassis can support either a 280mm AIO, or an air cooler that sports 140mm fans, 240mm AIOs sorta fall off a cliff.
Those other coolers end up having better performance to noise profiles because of the 140mm fans.
But if the chassis doesn't support either, then I suppose a 240mm will have to do.

was looking at it because EK is pretty hyped maker, also my case supports 280mm and i could get one but never really had any need for it, Silverstone kept my 4770 at ~50C when gaming and fx8300 @5ghz around 60, always quiet and looking good at the same time, also the fans on it are amazing so i never saw any need for that, also i don't like the looks of air coolers, i like clear case and "non messy coolers".
Im gonna check for some 280 but don't know yet if im gonna make the jump.
im gonna keep you updates on that aio's problem when i know what happened to it(if you wanna know).

Edit: was checking tests from 2019 and it does seem really good so i might just get a new one from rma, seems to outperform many air coolers and 280 https://www.techpowerup.com/review/silverstone-pf240-argb-aio-closed-loop-cpu-cooler/6.html

 

[Phaaze88]

Edit: was checking tests from 2019 and it does seem really good so i might just get a new one from rma, seems to outperform many air coolers and 280 https://www.techpowerup.com/review/silverstone-pf240-argb-aio-closed-loop-cpu-cooler/6.html

1)That was an 8700K. There's a lot at play here, but at the minimum, that cpu's die has a greater thermal density than the cpu's you're used to(32nm Vs 22nm Vs 14nm).
2)The test chassis was a Thermaltake Core P3. Practically an open bench system.

3)The biggest point, that I believe you overlooked, was the fans being run at 100% to get those results. That's where much of the performance comes from.
These products are advertised around 100%. Anything else may as well be icing on the cake.
Unfortunately, few can tolerate running fans like that regularly - users are more likely to run the fans slower, and hamper the performance seen in that link.
A fan's performance does not scale linearly either. Manufacturers don't really post fan curves, so how a fan performs at a given speed below 100% is anyone's guess.
This is where coolers equipped with 140mm fans can pull ahead, because at the same rpm they can move more air than 120mm.

4)You then might be persuaded to link, or point out the fan noise level and rpm chart.
It does not display fan sound profile. Many reviewers don't cover it, because what's tolerable varies individually.
Some fans can have annoying hums and whirrs, have them at specific rpm ranges, or when used in specific positions(push V pull, or horizontal and vertical).
140mm generally tend to have more tolerable sound profiles than 120mm.
This combined with noise level, allow the user to more comfortably run 140mm coolers at higher rpms, thus gain more performance.

5)Air resistance. While both an air cooler's finstack and an AIO's radiator present resistance, it's typically higher on an AIO radiator.
This means the penalty from running lower fan rpms is worse on an AIO.
This isn't even including the additional air resistance that may be present by whichever panel of the chassis you install the radiator against.
Your Armis AR6X EVO TG presents less resistance at the top than it does at the front.


I did mention that there wasn't much separating AIOs from one another besides fans, and there really isn't.


EDIT: I see this cooler only has a 2 year warranty - it used to be 1 year. I guess Silverstone doesn't have much confidence in this product...

 

[dande691]

The test chassis was a Thermaltake Core P3. Practically an open bench system

Hey, i know, i usually read the whole test, also this case is fantastic when it comes to airflow(my own words ofc) having about 2x32cm openings on both sides for intake and "escape holes" as i call them all over the case, i run my case fans at 40% at all time and cpu fans beetween 20-40% before 65C, i got a new one from warranty, they didn't open the old one but said the pump works like it has to, no side noises from the pump and normal current usage(no surprises there) so it leaves the corrosion as my main guess, stress tested the new one with AIDA64 and it stabilized at 54C with bundled paste so all good , silent and good looking, thanks for the chat mate👍

Edit: also for your older question, the back cover is solid metal with thin rubber insulators on it, metal bolts, finger screws with springs and metal "fitment plate" on pump and i know that my mobo is a bottleneck for that 970 EVO but it was nearly free for me so i decided to use it, gonna use pcie x16 adapter someday in the future.
.

 

[Phaaze88]

They didn't bother to open it? Likely because they would've broken it once they did. Most aren't designed to be serviced.
"Not worth it to try and examine it, so we'll just [e-waste] this one and just send 'em a new one. If we're lucky, this one will start becoming a problem outside the warranty."

You haven't really piled on the power consumption yet. The higher system power consumption is, the more air that is needed to flow in and out.

i know that my mobo is a bottleneck for that 970 EVO

I did not say that. Don't put words in my mouth, please XD

 

[dande691]

yep, maybe it just easier for them because its Silverstone's warranty, no reason to ship a part that cheap from 1 country to other, also its not that bad if it breaks, i could get a new one after warranty ends because we have a really good consumer protection in Finland

The higher system power consumption is, the more air that is needed to flow in and out.

something i already know as automation technician, nearly all energy(in form of electricity) turns into heat inside a pc, i was talking about the pump current usage tho, you can determine your pump by measuring its power usage.

I did not say that. Don't put words in my mouth, please XD

sorry about that, now you ain't got a change to do it XD.

 

[Phaaze88]

something i already know as automation technician, nearly all energy(in form of electricity) turns into heat inside a pc, i was talking about the pump current usage tho, you can determine your pump by measuring its power usage.

I see.
Not everyone knows about the power consumption though - that's why there's the occasional 'poor thermals in a hot box chassis'. They want a solution, and sometimes don't even want to change the chassis...

sorry about that, now you ain't got a change to do it XD.

Fiddlesticks...

 

[dande691]

I see.
Not everyone knows about the power consumption though - that's why there's the occasional 'poor thermals in a hot box chassis'. They want a solution, and sometimes don't even want to change the chassis...

best one's being those who have FX8150 and HD6990 in oem case with one 80mm fan and stock cooler back in the older days "i just builded this pc and it shuts down, not my fault"

also sorry about editing my older post so much, you already answered it before i edited it 😀

 

[Phaaze88]

It's all good, no harm done.

best one's being those who have FX8150 and HD6990 in oem case with one 80mm fan and stock cooler back in the older days "i just builded this pc and it shuts down, not my fault"

I find it can be more amusing when they assembled the parts themselves. At least with DIY, you have control over that.
Prebuilts are more of a compromise.

 

[dande691]

I find it can be more amusing when they assembled the parts themselves. At least with DIY, you have control over that.
Prebuilts are more of a compromise.

sometimes you get some parts for cheap and don't have a case so oem/prebuild case has to works i guess, funny part being that used cases are cheap.

 

[kerberos_20]

sometimes you get some parts for cheap and don't have a case so oem/prebuild case has to works i guess, funny part being that used cases are cheap.

ive had some time ago PC without case, mainboard was screwed inside drawer with drilled holes for some "air venting"...oh boy so much dust after just few months 😀