Jump to content
Corsair Community

H100i V2 Pump Failure Analysis


Charixfox

Recommended Posts

The first hint that something was terribly wrong was the slog of trying to do anything on the system.

 

"Why is it running so slow? This makes no sense."

 

*BEEP!*

 

"Oh really? Beep eh?" He glanced at the computer and noted the stark red glow of the Corsair logo on the H100i V2 pump.

 

A feeling of dread filled him as he started a monitoring program. Thermal throttling at 100C. CL4 reports: Pump at 0 RPM.

 

"Ah {censored}," he exclaimed, jabbing the power button, holding it down to force an instant lack of operation from the poor, abused system.

 

It's one thing to have a pump fail. Mechanical parts wear down. Parts eventually die. That's to be expected. Nothing lasts forever.

 

It's another thing entirely for this to be the third failed H100i V2 pump in a year.

 

Now, mind you, the computer is necessary for work. Downtime of more than a few hours is unacceptable, so RMA is out of the question. So every time this has happened, the reaction is, "Go get another H100i V2 at the store, swap it out, get back to work." The failed equipment gets tossed in the pile, forgotten, unloved, gathering dust until perhaps a future cleanup will send it to the trash.

 

Ah, but this time... this time it's different. A niggling recollection. A third addition to a pile that has a lot of similar parts. A A track of receipts: The first pump purchased in January of 2016. The second pump purchased in June of 2016. The third pump purchased in November of 2016. This fourth pump purchased in April of 2017.

 

As his eyes lingered on the pump pile, the corner of his lips twitched lightly. "Three. Three failed pumps." He half expected a crash of thunder and a desire to laugh maniacally, but thankfully that didn't happen. "Right then... let's get into these things."

 

About an hour later, he was shaking his head. "Well, the fins on the copper block are very nice, fine fins. I'm impressed at the quality of the block. Though it's interesting how much air is in the system, given that it's a closed system. And that color... that... that worries me. Then this." Tweezers held up a tiny, white speck, very much like a grain of sand. "They didn't. They couldn't've. No way."

 

The next system's radiator got a vicious shake and some tapping on the workbench before disassembly. Then the hose suffered the wrath of the cutting implements. Fluid poured from the system like the blood from a vein, laced with tiny, white, sand-like granules.

 

"And now for the moment of truth. Hi Radiator! DIE Radiator!" The resulting noise thereafter caused the dog next door to bark.

 

"Aluminum. They did. Corrode, and as soon as one unfortunate bit of aluminum oxide gets into the pump in the wrong place, it's game over."

 

"How often does this happen? Well, 3000+ reviews on Amazon. Let's see what we've got in 1-star reviews..."

 

---=====

The water pump died after 6 months of usage.

---

I purchased this unit in Aug of 2016 and installed it into a fully brand new PC build on an Intel i7 6700K 4GHz processor on an Asus z71 Sabertooth MoBo with Corsair Dominator 16GB (8x2) 3200MHz DRAM, EVGA SuperNOVA 850P2 PSU into a Corsair Vengence C70 series mid tower. A week ago on 3/27 the pump completely failed.

---

This worked fine for a month and then the pump died

---

Pump failed make sure u get a warranty. Such a waste of money having to buy another cooler after 5 months.

---

Pump died on me after 3 months

---

worked great for roughly 6 months, then the pump died

---

Dead in just ONE MONTH :(

 

I Left the computer overnight as always, in the morning I found it turned off, when I turn it on I got a CPU Fan error and the proccesor was 98 C :(

 

After many hours of checking the unit, even changing the thermal paste, I noticed that the pump was not working.

---

I got this in the mail and installed it and within a week the pump started grinding and gave out

---

It worked for one month and died.

---

DO NOT BUY!!! It has been 5 months and the pump failed.

---

Broke after 4 months.

---

5 month of use and the pump is dead.

---

Worked great for 3 months. Ordered back in August, it is now November and the pump has died. Had an h80i GT before this for about a year and that pump died as well.

---

I ordered it on August 30. On December 26 the pump completely died. Fans becomes too loud. I checked Corsair link software and found the pump was spinning at 0 RPM.

---

Experiencing a pump failure. I'm about 6 months in on the system, not particularly heavy use. Power system off when not in use. However pump failed after a few months usage. Seeking for a replacement from Corsair, customer service is atrocious, ridiculously slow, most importantly, I'm being told the item is backlogged.

---

failed 2 days later

---

Pretty akward to install, but seemed okay. Looked decent. Temps about the same as the the Gamer Storm single radiator liquid cooler I replaced with it. Big problem is: Installed Oct 1, 2016 pump dead as a doornail on Dec 17, 2016.

---

got it on july 19 2016 and its already having problems...it was working just fine then pump stopped working,

---

Pump dead after just 7 months

---

Piece of #%^* cooler worked great for 2 days then it stoped working contacted corsair and they said they would refund me as soon as i ship it back to them. I'm sure most of you know how that goes.

---

Pump died after 5 months.

---

Pump failed within only 4 months, CPU damaged. Costly product that does not deliver.

---======

 

"Well, huh. Yup, looks pretty common. Playing Russian Roulette with a $350 CPU is always fun, right?"

 

Three disassembled pump-failed units. Corroded block and aluminum oxide granules in all of them. It's so tremendously easy to find water cooling advice that makes it clear that mixing aluminum radiators with copper blocks can be a bad idea and absolutely requires anti-corrosives.

 

This makes me sad. This means I've spent about $400 on these parts in around a year and a half. I'm gonna go spend the same on a custom loop now. *sigh*

Link to comment
Share on other sites

Could you do a few RMAs? Given the Asetek:(: H100iV2s keep failing I would be inclined to ask for a CoolIT:sunglasse H110i, will a 240mm cooler fir in the case?

 

I think... that the highly-deconstructed items:

Block screws were evil (smashed some housings)

Block channel guides are a rubber or silicone

Many, many hoses were cut

The Hose to Pump is a one-way clip (broken)

The main board is soldered to the pump motor controller board which is hardwired to the magnet windings (cut)

The LED panel is actually nice and I saved those

 

They are likely VERY out of warranty now. XD

 

Anyway, the case will handle a 240mm or 280mm on the top, a 120mm on the back, and a 360mm or 280mm on the front as long as the port extensions aren't too long. But I happen to very much like my HD120 RGBs and so sticking to 120s, 240s, and 360s is on the agenda. ;)

 

I will say, the block is absolutely BEAUTIFUL though. I couldn't get a single one to unscrew successfully, so the fins got dinged (T^T) but where they're not bent fins... wow... this is a REALLY nice copper block on the inside (The IHS side is still sad and should be polished better). If I could figure out a way to safely extract it and make it into a custom block...

 

Where's my calipers? *Goes hunting* Ah!

 

0.09mm copper fins with a 0.13mm pitch. ^.^

Link to comment
Share on other sites

Don't you mean a 280 mm cooler?

 

If it was me, I would have bought a decent air cooler and RMA'd the H100i V2, then you would always have a spare cooler if/when problems arose.

 

But... but....

 

But custom loop... Bling... Blinkies... Computer water!

 

._.

 

 

(In all honesty though, for the first two failures, "MUST... GET... COMPUTER... WORKING!!!!" was the only thing on my mind. Same for this third, honestly. It was just "Three days later, saw the pile... MUST find out CAUSE!".

 

I mean, hey, if I'm writing Epic Stories™ about myself in the third person for dramatization... Why not add hammers to the mix?)

Link to comment
Share on other sites

Block screws were evil (smashed some housings)

 

Yeah, no joke there. I think Asetek do it on purpose. This prevented me from splicing together a new GPU cooler from multiple Asetek parts. I managed to get 6 screws out cleanly, but that was 18 short of the target. Plan fail.

 

 

Are you sure the specs are metallic? That would be interesting. I had theorized many of these were gummy, silicon-like deposits of adhesive combined with anti-corrosive salts. Neither one is particularly helpful.

Link to comment
Share on other sites

Could you do a few RMAs? Given the Asetek:(: H100iV2s keep failing I would be inclined to ask for a CoolIT:sunglasse H110i, will a 280mm cooler fit in the case?
Why would a coolit copper+aluminum system be any less likely corrode than an asetek cooper+aluminum system? Both are built based on specifications given by the same company: Corsair.

 

What I'd really like to see would be some feedback from Corsair on this. I'm still in my 30 days - perhaps I should return my 115i to amazon as defective and just buy something else completely.

 

As well, being that corsair doesn't seem to be taking any corrective action on these coolers and doesn't even give feedback on their own forums, perhaps I should consider also returning the corsair case and fans I purchased and take my business elsewhere.

 

Corsair?

Link to comment
Share on other sites

Yeah, no joke there. I think Asetek do it on purpose. This prevented me from splicing together a new GPU cooler from multiple Asetek parts. I managed to get 6 screws out cleanly, but that was 18 short of the target. Plan fail.

 

 

Are you sure the specs are metallic? That would be interesting. I had theorized many of these were gummy, silicon-like deposits of adhesive combined with anti-corrosive salts. Neither one is particularly helpful.

 

The things I do for Science...

 

Just confirmed that there is a electrical potential between the new block and the rad when disconnected from everything, which pretty solidly shows the possibility for corrosion. The way metal works, if it's "possible", it -will happen-, only the rate is in question.

 

Anyway, on the copper block image, the white specs are likely on the magnifying glass. Then you can see the rubberized stickdown section for the sheath. It makes both the gear-like pattern on the outside as well as the dark black bar across the center of the fins. The corrosion on the block is evident in the brown both on the fins as well as in the well around the fins. In a properly-running system, the block should be pretty, bright, shiny copper. Not this brown {censored}.

 

Inside the remaining not-opened radiator, even just in the barb fittings, there is substantial oxide, making the entirety of the inside of the barb white, including several nodules of growth.

 

Any anti-corrosive salts would need to be in solution, so any that have formed visible crystals are Very Bad. It could be a result of super-saturation due to evaporation of coolant, I admit, however that would not explain the pictured corrosion of the block.

Link to comment
Share on other sites

Why would a coolit copper+aluminum system be any less likely corrode than an asetek cooper+aluminum system? Both are built based on specifications given by the same company: Corsair.

 

Quality of construction and manufacture. There's a difference between saying that anti-corrosives are used in the fluid and actually using them.

 

This is speculation of course. I'm not going to purchase a 110 to analyze on my money.

Link to comment
Share on other sites

Well, this (along with my own issues and many, many, MANY other problem reports of Corsair water coolers) has scared me away from Corsair. Perhaps if Corsair monitored their own forums and responded to issues, things would be better. As it is right now, Corsair appears to be uncaring about customer relations.

 

I just ordered a EK based cooling system (360 slim radiator, supremacy EVO block, low power pump, etc.) Hopefully, the 360 radiator will work well in my case and with my HD120 RGB fans.

Link to comment
Share on other sites

I've owned my H110i GT for nearly 3 years and it has been flawless, both with and without Link. I previously had 3 H100i's which were mechanically excellent, until Link decided ti kill them with botched Firmware upgrades.

 

Whilst people can come on here and complain about problems, if they want anything done, they can contact Corsair via the Support Ticket System link at the top of the page, rather than hoping that someone from Corsair will monitor these forums.

Link to comment
Share on other sites

I've owned my H110i GT for nearly 3 years and it has been flawless, both with and without Link. I previously had 3 H100i's which were mechanically excellent, until Link decided ti kill them with botched Firmware upgrades.

 

Whilst people can come on here and complain about problems, if they want anything done, they can contact Corsair via the Support Ticket System link at the top of the page, rather than hoping that someone from Corsair will monitor these forums.

 

Of course I'd expect that many, many of these are just fine. It's absolutely possible to have a bad design that will not impact the vast majority of people in a manner that they are capable of discerning. Given that all three failed fans came from the same store, it's also very possible to have bad batches or shipment handling. Or case vibration from the fans could contribute to loosening the oxides.

 

"When you have a review score this high, don't worry about the fact that most of the negative reviews have exactly the same failure"

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...