H100i failed, maybe?

I’ve been away for a month and when I tried to start up my PC I got a “CPU Fan Error, F1 for Setup”

I went into BIOS and told it to ignore the CPU Fan. The fans in the water cooler are working fine as are all the other fans. Now it’s giving me: “CPU over Temperature error.” And it is. It’s at about 60C on entering BIOS, goes up at about a degree a second to around 74C then creeps up over the next few minutes till the PC shuts down.

 

I can get the PC to boot up if I set the CUP Fan in BIOS to Ignore but it only runs for about 60 seconds before the CPU overheats and it shuts down.

 

How do I troubleshoot what’s wrong with the water cooler? How do I know if it is the water cooler? Or is it just a matter of throwing it away and having to buy another?

 

It’s a H100i,

 

Any help appreciated.

When you hit ignore, it lets you bypass but the point of the feature is to let you know there is problem. You pump is not running and the CPU fan error suggests it is electrical or mechanical and not a BIOS setting gone wrong. The slow and steady escalating temps are another classic sign of no flow. Sorry.

 

Contact Corsair support through the ticket system above to start the replacement process. You will need a copy of the purchase invoice to establish when the 5 year warranty started. Also keep in mind the turn around on this is not going to be instantaneous and you cannot run the PC (and should not try) in this current state. If you are like me and need to be operational all the time, you either need to get a new cooler of your choosing and save the replacement cooler as a back up, or find a cheap stand in for two weeks.

Thanks, knowing where I am is 90% of the battle. I don’t mind picking up a new cooler but, and this is a little off topic, at the mo I've got a 2011 socket and will at some point upgrade. Do you know if these Corsair coolers cover the newer chips as well or are they specific to the socket?

Thanks, knowing where I am is 90% of the battle. I don’t mind picking up a new cooler but, and this is a little off topic, at the mo I've got a 2011 socket and will at some point upgrade. Do you know if these Corsair coolers cover the newer chips as well or are they specific to the socket?

 

The current crop of coolers covers pretty much all of the chips. What, specifically, are you looking to jump to?

No Idea, I only noticed Nvidia had the 2080 out last week. Couldn't tell you what "lake" Intell is on but I'm on an ASUS P9X79 WS with a 2011 socket so when I do get round to it, probably later in 2019, I'm guessing it's going to mean a new motherboard. if Corsair is trying to cover the bases and not making them socket specific then that'll do for now. I need to get this done now and I can't upgrade the whole thing at the mo.

 

There's an "H100i GTX V2 240mm Extreme Performance" on eBay for £55 that looks like it would do the trick.

All of the coolers will work on 2011/3 socket and I used many of them on the X99 platform. Without getting into the merits of buying from ebay, be aware that even if the seller is absolutely 100% reputable, I would not wager much money on the long term viability of the H100i GTX. The name itself makes it about 3 years old and I am not sure that model ages well. You can evaluate your own cost vs risk assessment, but my over/under on the lifespan would be about 1 year more.

It was a "Corsair H100i GTX V2". The H100i GTX is an older model, circa 2013 but I just presumed the V2 was the latest version of it. After your comment, I got a bit suspicious and checked and there is no GTX V2, it's just an old H100i GTX, so I canceled it.

 

That said I do buy stuff used off eBay more and more these days. Very occasionally it's crap but on the whole, if you don't mind not having the latest stuff, it's worked out well and saved me a lot of money.

The latest versions in the 240mm size are the H100i PRO (Asetek) and H100i Platinum (CoolIT). The Pro series has been pretty reliable thus far. The Platinum series just came out and offers included RGB fans, more complex pump LEDs, and self-contained lighting aspects. That one is going to come in a the highest price with the most "in the box extras". However, since it was just released, that may cause some price decline on the H100i Pro, even though it is less than a year old. There are some threads in the cooling section discussing the merits. I would avoid the GTX/v2 line, unless you are comfortable with needing to do this again soon in exchange for a bargain now.

I uninstalled Corsair Link years ago because it was so buggy but I thought it might tell me more about what's going on.

 

I put the CPU fan in BIOS as ignore to avoid the CPU Fan Error and left it a few hours so it was cold to avoid the CPU Temp Error and it booted up. C-Link is a small program so I've managed to quickly install it from a USB before the PC shut down.

 

I waited an hour or so again then booted it up. I ran Open Hardware Monitor and Corsair Link side by side and took a screen grab every minute till it shut down.

 

https://postimg.cc/gallery/2tfszpc0g/

 

C-Link read the temps as about 10C higher than OHM, the temps out of OHM looked about the same as the temps in BIOS. C-Link also recorded the Pump as working at about 2440 rpm up until the PC shut down.

 

I'm not so sure about what's going on now that C-Link is claiming the pump is working. I can't think of any other troubleshooting apart from buying an old cooler and swapping it out to see if that's the problem.

I should have stated "your pump is not functioning as it should" rather than the colloquial "not running". Your cooler is not functioning and needs to be replaced. Whether it is a electrical failing or a mechanical one doesn't matter from the user end. Neither is serviceable and you can't use it in this condition. Your very first shot has the coolant temp at double the expected peak value (68C). That alone tells me the fluid isn't going anywhere. It is not uncommon to have a zero flow state, but a positive pump reading. You could have a very solid blockage or the pump may be spinning ineffectively for mechanical reasons. Either way, you need to move toward a replacement. No further testing is needed. There is very clear evidence the cooler has failed.

I'm guessing from your, and everyone else's posts that the cooler is the only thing that can give these results. I'll go ahead and pick up another one today.

I'm afraid your case is fairly clear, although that is in fact better. Partial blockages where things are a little off but not quite catastrophic are much more difficult to discern and you have to rule out every other possible factor. For you, there is a clear chain of evidence.

 

1) CPU boot error - this is always a bad sign and rarely comes out on the positive end. With some coolers powered by the motherboard headers, it may be possible to set some super low fan profile (should be 100% always) that prevents the pump from starting. We can rule that out since you have a full registered pump speed. Same thing with a SATA powered cooler -- you have pump and fans and thus you have power. Lack of power to the cooler is not the source.

 

2) Coolant temp is very high - Normal maximum coolant temps are usually in the 40-45C range, depending on your room/case environment. People in tropical locales may bang up against that number all the time or even creep toward 50C with added GPU heat in small cases. But with a normal expected variances of +10C, anything in the 60s is way out of range and can only be caused by a zero flow state. You described that state perfectly with the 45-50-55-60....80-85-90-shutdown in 60 seconds. That is the classic sign of a zero flow state.

 

3) Fan speeds and pump settings alter coolant temps by a handful of degrees. No amount of cooler tinkering can replicate your temps, even if you turn the fans off. Now turning the pump off would look much like this, but we covered that in #1.

 

I am very certain yours is not going to have any future use and you should move toward a replacement of your choosing or RMA it through Corsair.