[SOLVED] How to turn off the overtemp alarm?????

Status
Not open for further replies.
Feb 13, 2020
18
2
15
You the alarm: it sounds like the house is on fire. Speedfan shows that "AUXTIN0" and "AUXTIN3" are "on fire" at 128C. I don't know what these are, but after going through the system with a thermal gun, I can guarantee nothing is that hot but these things are just screaming.

Turning off system notifications doesn't work. And nothing in UEFI seems to correspond to whatever these sensors are. Any thoughts on how to turn this off without turning off all sound? This is my daughters homework computer so there is some urgency here.
 
  • Like
Reactions: ColossalMoon
Solution
First thing I suggest is that you boot directly into BIOS Setup and go to the main Monitor Menu - see manual p. 2-34. Check there for the temperatures of your CPU and Motherboard. These numbers are reliable. IF they agree with your Speedfan indications of some temp in the 125 degree range, then you know you do have something wrong. On the other hand, if they do NOT show such problems, then Speedfan may be wrong. In that latter case, you might need to seek advice from Speedfan or from others here about how to correct Speedfan's actions.

If the BIOS screen does show a very high temp somewhere, you certainly need to investigate what that item is, and why it is hot. But if not, then we look elsewhere.

Interestingly, the manual indicates...
Feb 13, 2020
18
2
15
You can go into bios and shut down the alarm system power>hardware>cpu

Thanks for getting back to me. There doesn't seem to be this option in UEFI. Maybe it's hidden as something else. There I've turned off everything I could find related to fan and temperature monitoring, but "it" still continues to alarm.

Is there some other setting somewhere that would cause this alarming?
 

Paperdoc

Polypheme
Ambassador
First, tell us the maker and exact model number of your motherboard so we can look up its abilities.

Next, consider that this MAY be real, or it may be an error by Speedfan, a third-party utility that might need some configuration adjustments. I'm guessing Speedfan is configured now to always load and start up when you boot. Can you try to disable it temporarily and see if that problem goes away?
 

Paperdoc

Polypheme
Ambassador
First thing I suggest is that you boot directly into BIOS Setup and go to the main Monitor Menu - see manual p. 2-34. Check there for the temperatures of your CPU and Motherboard. These numbers are reliable. IF they agree with your Speedfan indications of some temp in the 125 degree range, then you know you do have something wrong. On the other hand, if they do NOT show such problems, then Speedfan may be wrong. In that latter case, you might need to seek advice from Speedfan or from others here about how to correct Speedfan's actions.

If the BIOS screen does show a very high temp somewhere, you certainly need to investigate what that item is, and why it is hot. But if not, then we look elsewhere.

Interestingly, the manual indicates that the mobo itself has only two temperature sensors - the one inside the CPU chip, and one installed on the mobo. There are no other sensors available for your BIOS to show, so what Speedfan is reporting as "AUXTIN0 and ...3 is not clear. It MAY be reporting non-existent readings that do NOT pertain to your mobo and should be ignored and maybe Disabled if they can be.

Now, the warning noise comes even when Speedfan is not running, you say, so we can conclude that the sounds are being generated by your mobo BIOS. In the manual on p. 2-35 I see four items you can set that generate warning alarms. Each is related to the minimum speed of a fan - one for the CPU fan, and one for each of the three CHA_FAN headers. There are few other possibilities, too, I'll note later. The CPU Fan Speed Low Limit is set by default to 600 RPM, but many current fans do operate properly at lower speeds. If possible, see if you can find the specs for your CPU cooling fan and see what its minimum speed should be. Also observe what speed it really is running shortly after you start up while the CPU is cool. If that is lower than the current limit setting, that is what is causing your warning sound. You can change that limit setting according to what speed the fan really DOES run in the cool system, or according to its min speed spec. The function of this Limit setting is to detect FAILURE of that fan and send out a warning you cannot ignore (It's working now!), but it is not part of fan speed control. So it should be set to below any normal speed for that fan.

Very similarly, look at the items for each of your CHA_FAN headers. They each have their own minimum fan limit that can cause an alarm, and again the default 600 RPM setting may not suit your actual fans. In these CHA_FAN headers, also note in the min speed limit setting that there is an option to IGNORE which you should choose for any fan header that is NOT being used for a fan. (Don't choose that for the CPU_FAN header, though.)

Now, for the CPU_FAN and all of the CHA_FAN headers one choice of Profile is "Manual". In that mode, you get to specify your own fan response settings. That is, what is the correct minimum fan speed at a minimum temperature you select, and what is the max speed at what max temperature? You also get to specify the slowest fan speed control setting (in terms of "Duty Cycle"), which is not exactly a specific speed. IF you have NOT chosen this Manual Profile, then the options for these items will not be shown and you can ignore them.

So, check into those fan speed limit settings. THEY may be what is causing your alarm sounds, and they may be wrong and require re-configuration. But all that depends FIRST on checking to verify that actual important temperatures are NOT too high and you can stop worrying about the temps.
 
  • Like
Reactions: Squids4daddy
Solution
Feb 13, 2020
18
2
15
First thing I suggest is that you boot directly into BIOS Setup and go to the main Monitor Menu - see manual p. 2-34. Check there for the temperatures of your CPU and Motherboard. These numbers are reliable. IF they agree with your Speedfan indications of some temp in the 125 degree range, then you know you do have something wrong. On the other hand, if they do NOT show such problems, then Speedfan may be wrong. In that latter case, you might need to seek advice from Speedfan or from others here about how to correct Speedfan's actions.

If the BIOS screen does show a very high temp somewhere, you certainly need to investigate what that item is, and why it is hot. But if not, then we look elsewhere.

Interestingly, the manual indicates that the mobo itself has only two temperature sensors - the one inside the CPU chip, and one installed on the mobo. There are no other sensors available for your BIOS to show, so what Speedfan is reporting as "AUXTIN0 and ...3 is not clear. It MAY be reporting non-existent readings that do NOT pertain to your mobo and should be ignored and maybe Disabled if they can be.

Now, the warning noise comes even when Speedfan is not running, you say, so we can conclude that the sounds are being generated by your mobo BIOS. In the manual on p. 2-35 I see four items you can set that generate warning alarms. Each is related to the minimum speed of a fan - one for the CPU fan, and one for each of the three CHA_FAN headers. There are few other possibilities, too, I'll note later. The CPU Fan Speed Low Limit is set by default to 600 RPM, but many current fans do operate properly at lower speeds. If possible, see if you can find the specs for your CPU cooling fan and see what its minimum speed should be. Also observe what speed it really is running shortly after you start up while the CPU is cool. If that is lower than the current limit setting, that is what is causing your warning sound. You can change that limit setting according to what speed the fan really DOES run in the cool system, or according to its min speed spec. The function of this Limit setting is to detect FAILURE of that fan and send out a warning you cannot ignore (It's working now!), but it is not part of fan speed control. So it should be set to below any normal speed for that fan.

Very similarly, look at the items for each of your CHA_FAN headers. They each have their own minimum fan limit that can cause an alarm, and again the default 600 RPM setting may not suit your actual fans. In these CHA_FAN headers, also note in the min speed limit setting that there is an option to IGNORE which you should choose for any fan header that is NOT being used for a fan. (Don't choose that for the CPU_FAN header, though.)

Now, for the CPU_FAN and all of the CHA_FAN headers one choice of Profile is "Manual". In that mode, you get to specify your own fan response settings. That is, what is the correct minimum fan speed at a minimum temperature you select, and what is the max speed at what max temperature? You also get to specify the slowest fan speed control setting (in terms of "Duty Cycle"), which is not exactly a specific speed. IF you have NOT chosen this Manual Profile, then the options for these items will not be shown and you can ignore them.

So, check into those fan speed limit settings. THEY may be what is causing your alarm sounds, and they may be wrong and require re-configuration. But all that depends FIRST on checking to verify that actual important temperatures are NOT too high and you can stop worrying about the temps.

Thanks for the super detailed answer! A couple of points, and I'll do this in bullet points so it's easier for you to read. The CPU and MOBO temps are just fine: it's definitely neither a real nor reported problem with CPU or MOBO temp.
  1. I stuck a VERY reliable temperature gun all through the case all around and there are NO temperatures that are high.
  2. I found a reference here--> https://linustechtips.com/main/topic/416216-what-is-auxtin-and-why-is-it-so-hot/ to what this is. It's a PSU thermocouple booboo of some kind. Unfortunately, I can't find any resource that tells how to turn the damn thing off.
  3. And here's where it gets troubling. I 'fixed' it in the second worse possible way. I pulled out the PSU far enough to get to the make/model, then put it back in. And it 'fixed' it. Speed fan now shows the AUXTIN0 in a 'normal' range. AUXTIN3 is still 'flame' hot, but it's not alarming. So what this means is that the next time 'something' 'jiggles' 'something' it's going to happen again. I did go through a push on all the connections.
So while the problem seems to be 'solved', it's not. Aargh.. There are resources for the PSU (THermaltake TR2 Bronze) that talk about which which is the 'data' wire from the PSU to the MOBO nor is there any software allowing me to turn off that feature.

I think this worth pursuing further because I can't be the only one and a real solution may become a service to the next 5 poor bastard out there that have this problem.
 
Feb 13, 2020
18
2
15
Thanks for the super detailed answer! A couple of points, and I'll do this in bullet points so it's easier for you to read. The CPU and MOBO temps are just fine: it's definitely neither a real nor reported problem with CPU or MOBO temp.
  1. I stuck a VERY reliable temperature gun all through the case all around and there are NO temperatures that are high.
  2. I found a reference here--> https://linustechtips.com/main/topic/416216-what-is-auxtin-and-why-is-it-so-hot/ to what this is. It's a PSU thermocouple booboo of some kind. Unfortunately, I can't find any resource that tells how to turn the damn thing off.
  3. And here's where it gets troubling. I 'fixed' it in the second worse possible way. I pulled out the PSU far enough to get to the make/model, then put it back in. And it 'fixed' it. Speed fan now shows the AUXTIN0 in a 'normal' range. AUXTIN3 is still 'flame' hot, but it's not alarming. So what this means is that the next time 'something' 'jiggles' 'something' it's going to happen again. I did go through a push on all the connections.
So while the problem seems to be 'solved', it's not. Aargh.. There are resources for the PSU (THermaltake TR2 Bronze) that talk about which which is the 'data' wire from the PSU to the MOBO nor is there any software allowing me to turn off that feature.

I think this worth pursuing further because I can't be the only one and a real solution may become a service to the next 5 poor bastard out there that have this problem.

Edit: Okay, buttoned it all back up and slid it back under the desk and then to the dumpster dive of hooking all the cable back up. THis was enough jiggling to re-fuckify the whole thing and I'm back to the 'waaahwaaahwaaahwaah". Now I can definitively say that this is alarm is coming from AUXTIN0--which is currently reading at 375 deg F! Back out with the fluke and there is NOTHING in the box anywhere above 110 deg F: it's a false reading.
 
I don't know what these are, but after going through the system with a thermal gun, I can guarantee nothing is that hot but these things are just screaming.
An Infrared thermometer is telling you the surface temperatures of objects, not what they are inside. The surface temperature of a heat sink, for example, should be a lot lower than the temperature of the chip underneath. Even more so if the heat sink were not properly mounted to the component in question.

375F (190C) does seem a bit unlikely though. I would expect things to be burning up at that temperature. Perhaps there's a failed temperature sensor on the board, or some software is attempting to read a temperature sensor that isn't there.

Does the sound come from your speakers/headphones or from a tiny speaker on the motherboard itself?
 

Paperdoc

Polypheme
Ambassador
HMMM. Are you SURE that this alarm DOES come on even when Speedfan is NOT running? To be sure, get your machine going and ensure Speedfan is not loaded and running. Then use the old CTRL-ALT-DEL "three-finger salute" to call up Windows Task Manager. Check both the Aplications and the Processes lists and look for anything that might actually be background operations for Speedfan. If there is something there, then Speedfan may be set to run at least a part of its monitoring operations even if you are NOT trying to use it. In that case you may need to start up Speedfan and see if you can adjust its configuration to never start up anything.

What I'm thinking here is that I do not know of any standard means for a mobo BIOS to monitor alarm signals from the PSU and send you a message or warning sound. So IF I'm right about that, the alarm is NOT being generated by the BIOS, and some other software must be doing that. Hey, maybe even some monitoring utility supplied with the PSU. Did you have to load anything like a "PSU driver" at some time long ago?

MAYBE you could start up another thread posting under a heading like "Help for Speedfan Alarms" and see if anyon ehere has dealt with this in the past. You might also contact Tech Support for your PSU and ask them for details of this issue.
 
Feb 13, 2020
18
2
15
An Infrared thermometer is telling you the surface temperatures of objects, not what they are inside. The surface temperature of a heat sink, for example, should be a lot lower than the temperature of the chip underneath. Even more so if the heat sink were not properly mounted to the component in question.

375F (190C) does seem a bit unlikely though. I would expect things to be burning up at that temperature. Perhaps there's a failed temperature sensor on the board, or some software is attempting to read a temperature sensor that isn't there.

Does the sound come from your speakers/headphones or from a tiny speaker on the motherboard itself?
Cryo, it is coming from the speakers.
 
Feb 13, 2020
18
2
15
HMMM. Are you SURE that this alarm DOES come on even when Speedfan is NOT running? To be sure, get your machine going and ensure Speedfan is not loaded and running. Then use the old CTRL-ALT-DEL "three-finger salute" to call up Windows Task Manager. Check both the Aplications and the Processes lists and look for anything that might actually be background operations for Speedfan. If there is something there, then Speedfan may be set to run at least a part of its monitoring operations even if you are NOT trying to use it. In that case you may need to start up Speedfan and see if you can adjust its configuration to never start up anything.

What I'm thinking here is that I do not know of any standard means for a mobo BIOS to monitor alarm signals from the PSU and send you a message or warning sound. So IF I'm right about that, the alarm is NOT being generated by the BIOS, and some other software must be doing that. Hey, maybe even some monitoring utility supplied with the PSU. Did you have to load anything like a "PSU driver" at some time long ago?

MAYBE you could start up another thread posting under a heading like "Help for Speedfan Alarms" and see if anyon ehere has dealt with this in the past. You might also contact Tech Support for your PSU and ask them for details of this issue.

Okay, so... a couple more things. I'm really really really sure this is not coming from speed fan. As you said, I was VERY VERY careful through a couple of starts to assure no speed fan was running. The alarm continued.

In parallel to this I was rebuilding a raid array. So I went to bed last night - speed fan on, this thing still screaming, speaker off, raid rebuilding. I get up this morning. The raid rebuild is done. So I cycle the machine and....hey oh!!! no klaxon blaring. Note that speed fan has not loaded.

I load up speed fan, fuly expecting the alarm to come back on. The alarm is NOT on: quiet as mouse. But AUXINT0 and AUXINT3 are 375 deg F and 187 degF respectively.
 
Feb 13, 2020
18
2
15
Interesting development, but I'm afraid I have no explanation for this. We will await further stories.

It's back on. I set speedfan so that on the auxtin0 it wouldn't alarm until it reaches 500deg f. It's still screaming even though speed fan isn't showing an over temp. I hadn't been in that screen before and it's showing the alarm is coming from a chipset called Nuvoton NCT679ID.

More interesting: I downloaded "Coretemp". Coretemp shows the CPU cores running between 47 and 51 C. Speedfan shows the cores running at 66C!
 

popatim

Titan
Moderator
Its not speedfan in alarm, its the motherboard.
AuxTin's are not wired to your PSU because PSU's do not normally have a Temperature or fan speed output wire in desktops.

As their name implies, they are Auxiliary Temperature inputs and the motherboard manufacturer can wire then to wherever they want or not use them at all. I would think they would be grounded or tied to high for this condition rather then just left floating.

You probably need to contact AsRock and see what the reading is for.
 
Feb 13, 2020
18
2
15
Yes...ASUS is making me enter a serial number which I'm having trouble retrieving because it's on the backside of the mobo which I can't get to without dissambly: which of course has it's own risks.
 

Karadjgne

Titan
Ambassador
For starters I'd set SpeedFan up in °C. It's much easier to associate numbers that are similar. And familiar.

Every component that's accessible has an address. In the early years, most of those addresses were similar, even the same range, and thats when things like SpeedFan and Hwmonitor were born. Today is different, and different vendors and different motherboards have dissimilar addressing. Even my old Z77 mobo isn't listed in SpeedFan, so it takes a generic approach to those addresses.

This means you can end up with false readings as the cpu tries to interpret a temp sensor from readings on Sata controllers, pcie hub, VRM's, Southbridge chipset, even the audio chipset. None of which have an actual temp sensor.

In both Hwmonitor and SpeedFan tmpin4 is 255°C and tmpin6 is -125°C (yes, a negative number). Both are physically impossible and have the pc not freeze/melt.

As far as temps go, you aren't hitting things with you IR. it's quite easy for the Sata chipset (usually down under the back end of the gpu) to hit 108°C if pushed. Pcie hub (small heatsink by the bottom of the ram) can also hit those temps, as can the Southbridge (small heatsink under the socket) or the VRM's mosfets to the left/above the socket) if overloaded.

None of which might be anywhere near the motherboard probe, which could be in an airflow pathway.

That Nuvoton chipset is a Super i/o chipset. It's hooked upto low bandwidth stuff like floppy controllers, mice, keyboards or other basic plug in stuff, and is linked to the Southbridge chipset which goes to the cpu.

So what's plugged in that's not USB? Is there anything like a ps2 keyboard or mouse, speakers or add-in cards not using the USB buss, but the serial or parallel bus? Are any of SpeedFan's fans set on either of those 2 buss that could be providing feedback and overheating/shorting/playing Hell with the chipset? If you do, start unplugging them individually, and checking the alarm status afterwards.
 
Feb 13, 2020
18
2
15
For starters I'd set SpeedFan up in °C. It's much easier to associate numbers that are similar. And familiar.

Every component that's accessible has an address. In the early years, most of those addresses were similar, even the same range, and thats when things like SpeedFan and Hwmonitor were born. Today is different, and different vendors and different motherboards have dissimilar addressing. Even my old Z77 mobo isn't listed in SpeedFan, so it takes a generic approach to those addresses.

This means you can end up with false readings as the cpu tries to interpret a temp sensor from readings on Sata controllers, pcie hub, VRM's, Southbridge chipset, even the audio chipset. None of which have an actual temp sensor.

In both Hwmonitor and SpeedFan tmpin4 is 255°C and tmpin6 is -125°C (yes, a negative number). Both are physically impossible and have the pc not freeze/melt.

As far as temps go, you aren't hitting things with you IR. it's quite easy for the Sata chipset (usually down under the back end of the gpu) to hit 108°C if pushed. Pcie hub (small heatsink by the bottom of the ram) can also hit those temps, as can the Southbridge (small heatsink under the socket) or the VRM's mosfets to the left/above the socket) if overloaded.

None of which might be anywhere near the motherboard probe, which could be in an airflow pathway.

That Nuvoton chipset is a Super i/o chipset. It's hooked upto low bandwidth stuff like floppy controllers, mice, keyboards or other basic plug in stuff, and is linked to the Southbridge chipset which goes to the cpu.

So what's plugged in that's not USB? Is there anything like a ps2 keyboard or mouse, speakers or add-in cards not using the USB buss, but the serial or parallel bus? Are any of SpeedFan's fans set on either of those 2 buss that could be providing feedback and overheating/shorting/playing Hell with the chipset? If you do, start unplugging them individually, and checking the alarm status afterwards.
I figured it out, and thought I'd circle back in case you ever see this again. It was Marvell Tray. No idea why it's alarming-but deleting it solved the issue. Thanks for your help!
 

Paperdoc

Polypheme
Ambassador
Glad you found the culprit and solution. I had to look it up, and I see Marvel Tray is a utility used to configure a RAID array using Marvel hardware, but it is not required for normal operations. Some comments I saw on the web suggest it has very poor documentation. I guess it falls into group of third-party utilities that need some adjustment before they can tell you the truth.
 
Status
Not open for further replies.