High temperatures on cpu and mobo

First, make sure you have the latest bios installed.

Second, sounds to me like you have a bad paste job on the CPU cooler. If the CPU is 80°C the heatsink SHOULD be hot enough to be uncomfortable to the touch, although it will probably never actually get hot, which is the point of it, it's getting rid of the heat.

It's also possible you are getting bogus sensor readings. Try installing HWinfo (NOT HWmonitor or Openhardware monitor) and run "Sensors only". Compare your thermal readings to those being reported in whatever other monitoring program you are using.

Also make sure the CPU is operating at the stock settings and that it is being reported as the correct CPU model. I've seen a few cases where this was not the case and max turbo core speeds were exceeding the factory specs as well as applied core voltage being wrong.

Probably also worth mentioning that the board you chose really isn't a very good choice for a FX8 processor. It would definitely not be on the short list of appropriate boards for any of the 8 core FX chips.

http://www.tomshardware.com/forum/id-2384024/motherboard-tier-list-970-chipset.html

I would say no. It could be 1 of two things, Your heat sink isn't sitting right on the processor or you didn't put enough pressure on the cpu with the heat sink and thermal paste to get a good bond between them.

Other things you could look at is updating the bios to the latest version might fix the issue. If you have overclocked, reducing them.

That is where I would start.

Ok, let's clarify a few things. There is no radiator on an air cooled system. There is a heatsink, which is vastly different from a radiator, which is what is used on water/liquid cooled configurations.

They don't "wear out". Unless they are damaged, ie, fins bent all to hell or the heat pipes going through the fin stack are pierced in such a way that the water or ammonia, or whatever transport fluid they've used inside the heat pipe, has been vented, then they simply do not "go bad". The fin stack can become filled with dust and dirt, needing to be blown out, or the fan that is installed ON the heatsink can fail, but the heatsink itself does not unless it's been damaged or altered.

Also, "professional" is a pretty vague term that is HIGHLY subjective when it comes to computer hardware maintenance and repair. I've met a lot of "professionals" that I wouldn't trust to blow the dust out of a heatsink. And besides that, even the best PC technician can occasionally get a bad paste job. Just because they know what they are doing, or at least somewhat know, doesn't mean there isn't still a problem with the paste job.

Putting an "AC" fan, does nothing really. It might change things a degree or two, but overall, it's a waste of time unless it is capable of dropping the ambient temperature in the entire room by 5 degrees or more, and no fan can do that without being part of a complete air conditioning system. Fans do not, and cannot, cause temperature drops below ambient temperature so putting one up to the side of the case is pointless.

Having only one fan in the case however IS a problem. You need more than one fan to even begin to create a thermal balance using the intake and exhaust configuration necessary on most modern systems. There are very few systems anymore than can get by with only a single case fan and so far as I know, there are NONE that include the 8 core FX chips. You really need two intake and two exhaust for that configuration. Minimum. Plus the CPU cooler.

Intake fans should be located at the front and bottom case locations. Top and rear fan locations should always be exhaust unless you have a special configuration OR, some water cooled configurations use the top locations as intake through the radiator that is mounted there.

Updating the BIOS is not risky if you are even remotely qualified to be working on a PC in any aspect. This is not 1996. BIOS updating is not only common now, it's often mandatory. Honestly, if you are uncomfortable with updating the BIOS, then you should not even be working on your system at all. No offense, that's just the way it is.

Yes, case IS a part. So, let's start by you providing a complete list of all parts including model number or best description you can provide.

CPU
Motherboard
Power supply
Memory
Graphics card
Case
Case fans
CPU cooler
OS system version

And to the best of your knowledge, how old each part has been in service.

So, let me get something straight. Your previous motherboard was only about ten months old when it died? If so, that was almost certainly heat related VRM or cap failure due to the lack of airflow in the previous case.

All of the FX processors, especially the 8xxx and 9xxx are what's known as high leakage processors. They have a very high TDP or "Thermal design power". Not only do these higher core count chips need very good motherboards to sustain them but they also need very good cooling across the board and the CPU cooler does not supply that. This whole configuration is skewed from the start without at least four 120mm case fans. Two in and two out. Again, this is a minimum requirement in addition to a good CPU cooler.

I'm going to assume that the ONE fan you currently have is the 92mm fan that came with the case, and that is not only not sufficient, it's almost insulting to the new motherboard, especially in light of the old motherboard that likely died as a result of the lack of cooling.

Since I was not there and could not see first hand what was going on, this is only a guess, but it is likely a very accurate one based on many, many hours and occasions of working with this series of processor.

Case in point, my megathread.

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0ahUKEwic8LC-3aPXAhWMVbwKHcpVDv8QFggnMAA&url=http%3A%2F%2Fwww.tomshardware.com%2Fforum%2Fid-2464608%2Famd-series-piledriver-architecture-cpus-megathread.html&usg=AOvVaw1G-UKIoq9IbIvhmtapMUWJ


That CPU cooler, while not great, and a rather old model, is sufficient for a stock FX8 configuration WITHOUT any overclocking. Couple of things here though.

While what I said about the heatsink no "going bad" is basically true, there are a couple of ways this could prove to be false. One way is if the unit was subjected to an unusually extreme and rapid change in temperature, there have been one or two instances where I've heard of (Never seen myself) the heatsink base becoming warped to the point where even when installed correctly it no longer sits flush against the top of the CPU and therefore cannot properly transfer heat. This is very unlikely unless you were overclocking and using some form of liquid nitrogen or a similar extreme configuration.

As mentioned earlier, if the fluid inside the heat pipes was to be released because somebody snipped the ends off the pipes where they terminate at the top of the fin stack, such as I've seen a few times when somebody thought it was ok to do that in order to make it "fit" inside a case without hitting the side panel. Once this has been done, or any other accidental or intentional reason why the integrity of any of the heat pipes has been compromised, the cooler is no longer fit for use and should be discarded.

It's also very possible that the CPU cooler simply did not get correctly secured in place or that one corner of the base assembly has somehow worked itself loose or slipped out. Something, anything, causing the cooler base to not be fully in contact with the CPU lid, OR, an improper paste job. These are the ONLY things that could be causing abnormal CPU temperatures aside from there simply being a lack of cool OUTSIDE ambient air being brought into the case due to there not being enough case cooling (fans).

One additional factor may be the region in which you live. I may be wrong but I'm going to assume you are in India or another far eastern country where there is generally a much higher average temperature than what is common in most parts of the world. This by itself is not a problem, but it certainly may ADD to the existing problem. Much harder to cool your system if the average temperature is 100 degrees versus if it were 80°F.

What country ARE you in, out of curiosity?

That's definitely not a great case, but I also know that prices and case choices are fairly limited in basically all Flipkart and Snapdeal regions. If you live in an area where those are two of the main computer hardware online vendors, I understand fully the complications of simply being not able to obtain good hardware at anything like a reasonable price.


I really think the very first thing you need to do is populate ALL case fan locations with as big of fans as will fit in those locations. This is going to mean either getting rid of that 92mm fan that came with the case altogether, or perhaps even rigging up a homemade mount to cool the motherboard VRMs. There are plenty of online guides on how to make homemade VRM coolers that I'm not going to waste any time going over that.

So before we go any further, what country are you in?

Well, that's a trip, I was just helping somebody from Latvia yesterday. I don't think I've helped anybody from either of those two countries, ever, that I know of, on here. Always nice to see members from different places reaching out here. Do you speak English or are you using translation software? Just wondering because if you are using software then it's very effective. I can't tell, and usually I can tell if it's being translated.

Anyhow, enough of that.

You're sure the problem with the old board wasn't actually the RAM itself?

65-75 degrees under full load is perfectly acceptable on AMD 8 core Piledriver and Bulldozer processors.

What EXACTLY are you using to monitor the CPU core temps?

As far as the fans go, take the time to reroute the wiring out of the way and add some fans. I don't care if it was "fine" before, believe me when I say that even the best systems, with the best cooling, only do an adequate job with FX8 processors, without plenty of airflow through the case. The CPU cooler can't even do ANYTHING at all, if it is using hot internal case air to try and cool the CPU with.

One thing you can do temporarily is simply take the side panel off. If temperatures drop more than 5°C then you know for CERTAIN that you need better case airflow. If temps do not, then there is another primary problem but that does not mean you do not still need some additional airflow.

Your case supports

Front : 1 x 80mm OR 120mm Fan

Rear : 1 x 80 / 92mm Fan (Included 1 x 92mm Black Fan)

Left Side : 1 x 120mm Fan


And I would HIGHLY suggest that you add both the front and side fans if at all possible, or run the system with the side panel OFF until you can afford to purchase those fans. That is an extremely old style case design, not well optimized for the thermal needs of your AMD FX series 8 core CPU, and you might want to consider another case going forward.

It's not JUST the CPU that is endangered by these thermal conditions. The memory (RAM), motherboard and graphics card, not to mention with this style of old design that still has the power supply up at the top, the power supply itself, ALL rely on there being good, cool, outside airflow into the case if you want them to last very long.

Your board almost certainly died due to thermal fatigue, not age. I have, and have clients with, many AMD 970 and 990fx chipset motherboards, and even older models, that have been running for 6-10 years or more with no failures. Age is usually not what kills computer hardware. Heat is. Even when you "think" something is working fine, often (Especially for users that don't even know that they should or can monitor these things) there are things like dust filled CPU coolers, lack of airflow, faulty fans or blocked intake and exhaust vents that are the real culprits of failure.

Also, even when the CPU core temps are ok, because you have a decent enough CPU cooler keeping it just within tolerance, doesn't mean that the VRMs and chokes on the motherboard, or the memory, or the north and south bridges, or the power supply, or anything else, is getting enough cool air for them to continue operating at peak performance for a long time. This is irrefutable and I can show you hundreds of examples proving this.

So, anyhow, let's make sure you actually even have an immediate problem

Download CoreTemp: http://www.alcpu.com/CoreTemp/

Install CoreTemp and then open the program. Go into the options >settings and on the advanced tab, check the box next to "display the distance to tjmax in temperature fields".

Close the settings dialogue, move the CoreTemp box over out of the way and then run Prime95 version 26.6 (32 or 64bit, depending on which bit depth your operating system is) which you can download here:

http://windows-downloads-center.blogspot.com/2011/04/prime95-266.html

Watch your distance to tjmax readings and if any core gets to less than 20°C distance to tjmax, there is a problem. It would be preferred to keep about a 30°C distance to tjmax for normal operation.

Are you using the same CPU cooler you were using before? If so, there is probably a problem with the CPU cooler. Usually when they start reporting inaccurate RPM readings it's because there is either a problem with the wiring, the fan header, the motherboard or the cooler. Since you have a different motherboard it clearly cannot be the motherboard or fan header that is to blame. Too unlikely to have the same exact problem with two consecutive motherboards.

Most likely it is the fan on your CPU cooler, especially if it's the same one you were using before. We are talking about the CPU cooler fan RPMs, right?

No, it's not weird

Core temperatures increase and decrease instantly with changes in load.

Click to expand...

Intel’s specification for DTS response time is 256 milliseconds, or about 1/4th of a second. AMD is either the same or so close as to be indistinguishable from the same.

Basically, this is because:

The temperature measured at the heat sources near the transistor “Junctions” inside each Core by individual Digital Thermal Sensors (DTS). As such, “Core” temperature is a specific term.Since Windows has dozens of Processes and Services running in the background, it’s normal to see rapid and random Core temperature “spikes” or fluctuations, especially during the first few minutes after startup, which should eventually settle.

Click to expand...

This information is taken directly from the Intel temperature guide, but the architectures of processors are so similar that there virtually can be little to no difference in how this works in theory and in operation. Package temperatures may rise and fall more slowly, as do thermal readings for the air temperature inside the case, but GPU and CPU CORES react to loads and basically almost instantly rise or fall in temperature so long as there is normal heat transfer away from the CPU by way of a heatsink or similar device.

If there was not a heatsink to transfer heat away from the CPU, then core temperatures would drop far more slowly, and the entire package would gradually increase in temperature to the point where it became difficult for the cores to drop in temperature at all if there is any kind of load present.

If your temperatures went to within 1°C of TJmax then there is only two possibilities, the way I see it.

1. There is an overclock setup and it is setup too high for the cooling to be capable of handling.

2. There is a problem with either the CPU cooler or the CPU, or possibly the motherboard, but that is unlikely since it DOES seem to be accurately reporting the thermal margins according to load conditions.

I would immediately stop using this computer until the cooling problem is resolved.

The first thing, as I said before, is to make absolutely certain that the CPU cooler heat sink is completely and fully seated down onto the CPU by way of the mounting hardware.

If you can easily move or twist the CPU cooler, it is not mounted correctly.
If you can see that one corner or one side is not equal with all other sides, it is not mounted correctly.
If ANY kind of mounting hardware other than what was intended to be used with that CPU cooler from the factory is being used, it is not mounted correctly.

If the CPU cooler is FIRMLY and EVENLY mounted to the motherboard, then it's possible that either too much or too little thermal paste was used, or that an air bubble was trapped between the CPU lid and heatsink base.

If the cooler is mounted correctly and the paste is 100% correctly applied, then it's possible something else is going on like the CPU cooler fan is not actually spinning or the fan blades are not spinning as fast as they should be because the center shaft has actually broken free from the hub of the fan blades. In this case the fan blades may still spin, but will not spin as fast as they are supposed to be spinning which means that very little air will be passing through the fin stack to aid in cooling the CPU.

This MIGHT also account for the problem with the skewed sensor readings saying 60000rpm. Might also just be a failing fan on there.

None of that really addresses the fact that within a minute or so it is reaching TJmax. Even without a fan on the CPU cooler heatsink, if it is mounted correctly it should take longer than that for the CPU to reach TJmax because some coolers don't even use fans on the heatsink.

I think you might want to start with getting a very good CPU cooler, installing it and seeing where you are at that point.

Alternatively I think no matter what you do, you will probably need to get a new CPU cooler fan, at the least.

Given the age of your CPU cooler, it's probably not too surprising that the fan is showing it's age, if that's what it is.

You might try connecting a different fan to the CPU_FAN header, turn the system on and see if it is still reporting crazy numbers.

So, to begin with, it wouldn't matter if that fan was a 50 dollar Delta industrial fan, which would mean it's a pretty damn good fan, it could still be bad especially when it's 4+ years old. If you tend to leave the system running when it's not in use this will DRAMATICALLY decrease the lifespan of any and all fans in the system.

Even if you don't tend to do that, 4-6 years is a pretty good lifespan for any fan installed in a system that sees regular, frequent use. Bearings and small electronics just don't last indefinitely and that's without even factoring in the metal whiskering that occurs over time.

That being said, there is nothing wrong with replacing the fan you have on there, with another just like it, if that's what you want to do. I'm sure if you contact Arctic they'd be glad to sell you one. You can get one here:

https://www.arctic.ac/us_en/spare-fan-freezer-13-pro.html

However, it's NOT a really great fan. THIS would be about 10x better, and is probably one of the best you can get, bar none.:

https://www.amazon.de/Noctua-NF-F12-PWM-chromax-black-swap-Premium-L%C3%BCfter/dp/B07654PNFQ/ref=sr_1_1?ie=UTF8&qid=1509912508&sr=8-1&keywords=NF-F12+PWM+chromax.black.swap


But this would be a good option too. I'm currently using four of the DS140 Aerocool fans in my case as I installed them a while back just to test sound levels and airflow, and never bothered to put my Noctuas back in because they worked so well. I got lazy, then went away for work for a while, came back, had shoulder surgery and never got back to it. I will do it soon, but just saying that for a company not well known for making great fans, they hit it out of the park on this design.

http://www.aerocool.us/accessory/images/dsfan_12bk.html

You will encounter a problem though, and that's the fact that Arctic decided to make it hard on anybody trying to replace the fan on their cooler with any different kind of fan, because they made the setup basically proprietary and integrated the fan shroud to also be the mounting system used to attach the fan to the heatsink. If you use another fan you will need to find clips (Maybe from another brand of cooler) or use zip ties to attach the fan to the heatsink. Otherwise, maybe you could somehow cut the sides off the shroud on the original fan and attach them to the frame of the new fan using tiny screws, I dunno, you might be best off using the original style fan if you are not comfortable with any other method.

I don't think it's absolutely necessary to replace your current cooler, but replacing the fan might be a foregone conclusion if you are unable or unwilling to test it. You COULD also plug the CPU cooler fan into a different header and see if THAT header sensor reports crazy RMP readings too.

I don't think you have done as I had suggested so let me tell you again. It would be wise to unplug the CPU cooler fan from the CPU_FAN header and plug it into one of the sys_fan 4 pin headers. None of them are very close to the CPU so you may have to obtain a 4 pin fan cable extension in order to reach them OR not plug the CPU cooler in for this test, which means you need to be rather quick about what you are doing as it's never good to run the system very long with a CPU cooler fan running.

I guess I need to ask here, is the 92mm fan that came with your case a 3 or a 4 pin fan? If it's a 4 pin fan then HOPEFULLY it's a PWM controlled model and you can plug it into the CPU_FAN header and power up the system to see if you are still getting the crazy RPM readings. If you don't wish to do any of this, ok then, just replace the fan.

Avoid the Hyper 212 coolers, ALL of them. They suck, I know, I've used it many times in the past on client builds and even written tutorials on installing them.

http://www.tomshardware.com/faq/id-2520482/solving-temperature-issues-hyper-212-evo.html


If you buy a whole new cooler, I'm sorry, but I'm going to advise you to avoid the entire Arctic cooling product line. It simply is not that good compared to other brands and coolers. It's not terrible, certainly better than most comparable coolers by Cooler Master and similar budget lineups, but it's not great.

You are severely limited in your choice of CPU cooler because your case only supports coolers up to 150mm in height. I have to ask here, how does that cooler even fit inside your case? That cooler is 159mm in height and your case only supports coolers 150mm in height according to the specs on both the cooler and case product pages.

I'd recommend any one of these:

PCPartPicker part list / Price breakdown by merchant

CPU Cooler: CRYORIG - H7 49.0 CFM CPU Cooler (€34.90 @ Caseking)
Total: €34.90
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2017-11-05 20:58 CET+0100



PCPartPicker part list / Price breakdown by merchant

CPU Cooler: Thermalright - Macho-120 73.6 CFM CPU Cooler (€55.29 @ Amazon Deutschland)
Total: €55.29
Prices include shipping, taxes, and discounts when available
Generated by PCPartPicker 2017-11-05 21:06 CET+0100


And actually there are a lot more that I'd put on the recommended list, but I need to know HOW it is that you have a CPU cooler that is taller than what your case supports, installed? Is it because there is no side fan installed and the specs on the height of coolers for that case assumes you will have a fan installed there? What gives?

I think you're going to possibly need fan adapters. I only see TWO fan headers on that motherboard that are suitable for use with voltage or PWM controlled fans. The other fan header, closer to the CPU fan header, is a three pin power fan header, which probably only allows for full speed fan operation and will not allow any kind of voltage or pwm power control. I could be wrong, and I'll look into it, but usually when they are labeled pwr_fan they are full speed 12v headers.

We need to clarify. Is the rear exhaust fan, or the CPU fan, the one that gives the funky rpm reading?

Can we make sure also that we are looking at the exact same case? This one here?: https://www.aerocool.com.tw/en/chssis/pgs-v/v2x

Then I would get a 3 pin fan extension and use that to connect your rear exhaust fan up to that 3 pin fan header that's up near the front of the board, so you can use the other four pin headers for the fans you're getting.

Something like this. There are probably less expensive ones out there:

https://www.amazon.com/3-Pin-Extension-Sleeving-Connectors-Length/dp/B00C46K88M/ref=sr_1_3?ie=UTF8&qid=1509922577&sr=8-3&keywords=3+pin+fan+extension+cable

I've only been looking at the DS3P FX rev2.1 specs.

You can use up to two (Three can be done but is not recommended because it draws too much power from a single header circuit that way) PWM or voltage controlled fans, via fan splitter cable, per fan header.

If you install only one fan, it should be in front as an intake.

I think I know now why you are getting funky fan readings. If you are using a 3 pin voltage controlled fan on a four pin PWM header, it's going to not only run at the wrong speeds, usually full speed, depends on the implementation of the pin arrangement and not all are the same, it cannot send accurate RPM readings. This might even screw up other sensor readings depending on how the circuit is implemented.

Did you have the CPU fan connected to the CPU_FAN header, or to another fan header? If it's connected to the CPU fan header then the readings should be correct, unless of course on your previous board you have it connected to a different non-PWM fan header and caused something to be damaged OR had it installed backwards on the fan header which could cause any number of problems.

I would highly recommend the Cryorig H5 if it will fit. The other options I think are even better, but also more costly, are the Noctua NH-U12S, NH-U14S (This is what I run on my main system), Thermalright Macho X2, Thermalright True spirit 140, Phanteks PH-TC12DX and Scythe Mugen 4 or Mugen 5.

I'm hopeful that with a new cooler, which means a new fan, installed in the right place, and if you get a fan extension cable and connect your rear 3 pin fan to the 3 pin fan header located here, that it will work correctly and temps will be much lower especially with the cooler mounted correctly and pasted according to best practices.

Sounds good. Let me know. If not, we'll see what else we can figure out.

Not to mention, like I said before, there are plenty of "professionals" out there that have been making a living doing things wrong for a very long time.

And, more importantly, there are a lot of "IT" people who are knowledgeable when it comes to networking and communications equipment setup and configuration, that surprisingly are not very knowledgeable or experienced when it comes to enthusiast or general hardware questions or configurations.

One example that is highly relevant to your situation. I worked for Estes rockets, a hobby rocket manufacturer, in their warehouse for a number of years. Our company had several warehouses and various offices used for data entry etcetera. They had an IT engineer that they sent to reconfigure our network and upgrade a few of the older systems. This guy was sharp as a whip when it came to the network hardware and the software configuration for the network but he no idea, AT ALL, how to install the new motherboards, CPUs and CPU coolers. I ended up having to do it after he bungled the first cooler installation and damaged some of the hardware.

There are IT engineers and there are computer hardware engineers. Just because you are one doesn't necessarily mean you know jack about the other. You have to keep in mind that many of these guys/gals get into this as a profession and they focus on that. They are not necessarily enthusiasts who strive to learn as much as possible about a broad range of things, instead focusing only on what is needed for their jobs.

The other side of that is that for as much as I know about hardware and software, installation and configuration, network engineering is an area where I am particularly weak. I know my way around a home network and can configure a small business network, but at some point I have to admit that I'm in over my head and call in somebody with more experience if the network is too complex.

Anyhow, point being, dude might not have REALLY known what he was doing just because he's been working in IT for a long time.