Stock CPU overheating to 90 in 4 minutes
Printed From: ASRock.com
Category: Technical Support
Forum Name: AMD Motherboards
Forum Description: Question about ASRock AMD motherboards
URL: https://forum.asrock.com/forum_posts.asp?TID=1869
Printed Date: 27 Dec 2024 at 2:17am
Software Version: Web Wiz Forums 12.04 - http://www.webwizforums.com
Topic: Stock CPU overheating to 90 in 4 minutes
Posted By: Atomic
Subject: Stock CPU overheating to 90 in 4 minutes
Date Posted: 30 Jan 2016 at 6:30am
|
Hey Guys, I am using default settings in the BIOS with it running at 3.5GHz to avoid the need for any extra fans. However I've noticed that the temperature for the CPU spikes within 4 minutes of first turning on the PC. The temperature for the CPU looks to be spiking all the way up to 88/90 degress. I mean, Wow! :( It's a cold day today and there is no way it could be overheating so quickly. This causes the Computer to do an emergency shut down. After the computer shuts down I touch the CPU cooler and the radiator and both are cool to touch. This happens intermittently and after a few reboots the system starts running normally without any issues. The temperature then remains stable at "30 or 40" and no problems re-occur. I have been thinking it could be something to do with the water pump or a faulty temperature sensor. The temperature spikes both when inside the UEFI BIOS or the Windows 10 Interface. Picture Below:  |
Replies:
Posted By: parsec
Date Posted: 30 Jan 2016 at 1:06pm
|
Please describe your CPU cooler in more detail than, "... the CPU cooler and the radiator...". The CPU temperature, or CPUTIN, is coming from the CPU itself, not the mother board. If the CPU temperature sensor is faulty, then that is a CPU defect. The Package temperature shown might be the CPU socket temperature sensor. HWMonitor is showing the CPUTIN temperature as a minimum of 72C, and a maximum of 80C. The Package temperature shows a minimum of 73C, and a maximum of 85C. Given these are different sensors, they are quite close and in agreement. That tends to indicate no defect in either sensor. We can see core usage maximums in the high 80% - low 90%. Have you heard of the Windows 10 issue of 100% CPU loads for no apparent reason, that affects some system? I thought that might be an Intel only thing, but perhaps not. Your VCore values are all under 1.0V, the minimum to maximum being 0.888V - 0.904V. Very little difference between the two, and a low VCore for 80C+ CPU temperatures. Your core clock speeds of a maximum of 1419MHz, is either throttled or the CPU power saving options enabled core speed. No where near the 3.5GHz base clock or 4.1GHz Turbo clock speed. It looks like a throttled speed IMO. So I would say something wrong with the CPU coolers mounting, or its function, or both. CPU temperature can increase from 30C to 100C in a few seconds given a stress test load, high VCore, or poor CPU cooler function. Notice I said or poor CPU cooler function, which means the CPU temperature can climb to 90C - 100C in a few seconds with good CPU cooler function and a stress test load. Four minutes is not needed for the CPU temperature to reach 80C - 100C, that would indicate fantastic CPU cooling ability. Not being able to feel any heat in the CPU cooler system could indicate incorrect mounting, or the pump not working, since you mentioned a pump. ------------- http://valid.x86.fr/48rujh" rel="nofollow"> 
|
Posted By: Atomic
Date Posted: 30 Jan 2016 at 2:16pm
|
Hi Parsec, The CPU speed when it is NOT overheating and remaining stable and cool is running at 3.5GHz. However when I have the overheating issue. It does appear the CPU is throttling because it thinks it's overheating and therefore I get the 1.45GHz speed. Now, the CPU Cooler I have is called a "Deep Cool Maelstrom 120" Here is the link to it: http://www.deepcool.com/product/gamerstorm/2014-05/6_854.shtml" rel="nofollow - http://www.deepcool.com/product/gamerstorm/2014-05/6_854.shtml I have re-seated the CPU Cooler and it has screw holding it down. They seem tight and secure. The thermal paste was applied in factory and there is no thermal tube from which I can use to re-apply it onto the surface; if you want me to do a complete reset. The unit is still under warranty and I sent a Ticket to Deep Cool over 1 week ago. However I have received no word back from them. I don't know what to do next.
|
Posted By: wardog
Date Posted: 30 Jan 2016 at 2:53pm
|
If that's the MINIMUM RPM of the fan, then I'd say there's your problem. It's not raising RPMs during boot as expected. Whether that steady RPM reading is due to a fan setting in BIOS, a fan setting in F-Stream, a defective fan, or maybe too a defective fan header on the motherboard, I can't say from here. If it's a 3-pin fan move it to the other CPUFAN header and report back how it responds. But it shouldn't be staying steady as it shows. |
Posted By: wardog
Date Posted: 30 Jan 2016 at 3:01pm
|
I forgot. You're the one with the pump. What I notice right away concerning overheating is that the RPM of your CPUFAN(Pump!) stays constant. If that's the MINIMUM RPM of the pump, then I'd say there's your problem. It's not raising RPMs during boot as expected. Whether that steady RPM reading is due to a fan(Pump!) RPM setting in BIOS, a fan(Pump!) setting in F-Stream, a defective pump, or maybe too a defective fan header on the motherboard, I can't say from here. If the pump can be connected to another CPUFAN header, move it to the other CPUFAN header and report back how it responds. But it shouldn't be staying steady as it shows. |
Posted By: PetrolHead
Date Posted: 30 Jan 2016 at 8:28pm
A small correction: The CPUTIN temperature is listed under the motherboard, so it is based on data provided by a sensor on the motherboard. In other words, it is the temperature given by the sensor in the socket. The package temperature is from the CPU. ------------- Ryzen 5 1500X, ASRock AB350M Pro4, 2x8 GB G.Skill Trident Z 3466CL16, Sapphire Pulse RX Vega56 8G HBM2, Corsair RM550x, Samsung 960 EVO SSD (NVMe) 250GB, Samsung 850 EVO SSD 500 GB, Windows 10 64-bit |
parsec wrote:Posted By: PetrolHead
Date Posted: 30 Jan 2016 at 8:40pm
If you took the cooler of the CPU, then you should probably do a complete reset. The "rule" is that once you break that interface between the CPU, thermal paste and cooler, you need to redo it from the top. Reusing the old paste may lead to inefficient heat transfer from the CPU to the cooler. However, it probably won't lead to an overheating issue like the one you have, so you can probably wait until this issue is resolved. Btw, how about the air that the fans are blowing from the radiator of your liquid cooler? Have you tried if the air gets any warmer during the overheating? Or if it's warmer when the system is running normally? I think that would be a relatively easy way to tell if the pump is working as it should, unless the differences in temperatures between different scenarios are too small to notice. ------------- Ryzen 5 1500X, ASRock AB350M Pro4, 2x8 GB G.Skill Trident Z 3466CL16, Sapphire Pulse RX Vega56 8G HBM2, Corsair RM550x, Samsung 960 EVO SSD (NVMe) 250GB, Samsung 850 EVO SSD 500 GB, Windows 10 64-bit |
Posted By: Atomic
Date Posted: 09 Feb 2016 at 5:41pm
|
Hi Everyone, I think I have figured it out!!!! Looking at the BIOS closely when the system overheats. I can see the CPU FAN 2 doesn't show/register any RPM speed. This CPU FAN2 connector had the CPU heat sink/PUMP connected to it. It appears that the power wasn't being sufficiently given to the pump to operate/turn on. However if I leave the CPU FAN2 alone and do a reboot then the pump starts working. I then go back into the BIOS and I see the CPU FAN2 showing an RPM speed again. :) The water in the pipes does feel very hot when I touch it, when the CPU FAN2 header stops powering it. Looks like a motherboard glitch after all... Or could it be the CPU Cooler? Thoughts anyone? - ATOMIC
|
Posted By: Atomic
Date Posted: 09 Feb 2016 at 10:20pm
|
The pump is connected to the "CPU FAN2" connection. I have also tested this connection on another pin header on the motherboard and I get the same issue. Like I said earlier, on occasion...sporadically the RMP speed under the FAN Connector which has the pump connected stops showing an RPM speed. It show's N/A and the temperature climbs very rapidly and within a minute the computer shuts down from overheating. I can't figure out if this is a fault with the Motherboard FAN Connections or the CPU Cooler!!! I'm scratching my head on how to RMA or resolve this.
|
Posted By: PetrolHead
Date Posted: 27 Feb 2016 at 6:36am
|
Sorry for not responding earlier. At this point my guess is as good as yours. Have you tried connecting regular fans to those same fan connectors to see whether they also stop spinning? ------------- Ryzen 5 1500X, ASRock AB350M Pro4, 2x8 GB G.Skill Trident Z 3466CL16, Sapphire Pulse RX Vega56 8G HBM2, Corsair RM550x, Samsung 960 EVO SSD (NVMe) 250GB, Samsung 850 EVO SSD 500 GB, Windows 10 64-bit |
Posted By: parsec
Date Posted: 27 Feb 2016 at 11:51am
There is nothing wrong with your mother board, or the CPU Cooler. The fault is with how you have the pump connected to your mother board. The pump is in the part that you connect to the CPU. For small AIO liquid coolers like this one, the pump should be running at one speed, full speed all the time. The pump speed should NOT change like the fan speed changes. Your board has a fan connection called PWR FAN1, that is just below the 24 pin connector for the PSU. Connect the wire from the PUMP to the PWR FAN1 connector. That will run the pump at full speed all the time, as it should be. Or if you have a four pin molex to three pin fan adapter that comes with some PC fans, you can connect the pump directly to the PSU. That is how I used all of my AIO liquid coolers, I gave them a full 12V of power with no controls in the way to slow them down. When you have the pump connected to the CPU FAN2 connector, it is NOT running at full speed. If you can set the CPU FAN2 connector to Full On, then do that, if you can control the fan on the radiator by itself. If you see NA for the pump speed, no wonder your CPU is over heating. The pump speed should be 2200 RPM, +/- 10%, which means between 2420 and 1980 RPM. Or, anything less than 2000 RPM is not right. What fan speed profile did you select for the CPU_FAN2 connector? The default setting? Any setting except FULL ON is wrong for your pump. The best way to connect the pump to your PC is either of the first two methods I suggested, either use the PWR FAN1 connector, or connect it directly to the PSU. ------------- http://valid.x86.fr/48rujh" rel="nofollow">
|