[mchain]

Refurbished business unit bought from Micro Center 1 1/2 years ago.  Pentium D 2.8, 3.0 GB RAM, Win 7 Home Premium 64-bit SP1.  No discrete video card, onboard video only.  Two HDD's and Sandisk ReadyCache installed.

 

Issue:  CPUID Hardware Monitor is showing unexplained spikes in both CPU and AUXTIN temperatures.  Appear to be so briefly transient that these do not ever show in current value column. 

 

I'm thinking this is reason to be concerned?

 

CPU runs to 72 C -74 C and then CPU and case fans spin up, holding around 2500 rpm until load is removed or process stops running.  Running a streaming video in a browser will eventually cause these peak values to show.

 

Used HeavyLoad run of 10 minutes + to get the screeshot attached below.

 

Thank you.

Attached Thumbnails

• 

[Advertisements]

Hello mchain,

 

Possible causes may include a flaky/underpowered PSU and an overheating MB.

 

Streaming videos is demanding on system resources and unfortunately older integrated video chips are not best suited for such usage and they will heat up when put under load, more up to date integrated video chips cope better due to improved technology and better support from the CPU and Ram but you would still be better off with an add on video card if streaming videos on a regular basis or gaming.

 

Your screenshot is showing that the +12V rail on the PSU is very low and needs to be checked out;

 

Download Speedfan and install it.  Once it's installed, run the program and post here the information it shows.  The information I want you to post is the stuff that is circled in the example picture I have attached.

If you are running on a vista machine, please go to where you installed the program and run the program as administrator.

 

 (this is a screenshot from a vista machine)

 

 

The HP Compaq dx2300 Microtower Business PC was shipped with a 250W PSU as standard and that would be the bare minimum for the original specifications, an SSD has since been fitted but was the PSU also upgraded to allow for the extra power demand.

 

JAFYI: CPUTIN is the MB CPU temp sensor and AUXTIN is the power supply temp sensor if there is one, the max safe operating temp for your CPU is 62.3°C and the computer would/should shut down if CoreTemp (the sensor on the processor) reaches or exceeds this temperature, this unless the setting has been changed in the BIOS.

[phillpower2]

Hello mchain,

 

Possible causes may include a flaky/underpowered PSU and an overheating MB.

 

Streaming videos is demanding on system resources and unfortunately older integrated video chips are not best suited for such usage and they will heat up when put under load, more up to date integrated video chips cope better due to improved technology and better support from the CPU and Ram but you would still be better off with an add on video card if streaming videos on a regular basis or gaming.

 

Your screenshot is showing that the +12V rail on the PSU is very low and needs to be checked out;

 

Download Speedfan and install it.  Once it's installed, run the program and post here the information it shows.  The information I want you to post is the stuff that is circled in the example picture I have attached.

If you are running on a vista machine, please go to where you installed the program and run the program as administrator.

 

 (this is a screenshot from a vista machine)

 

 

The HP Compaq dx2300 Microtower Business PC was shipped with a 250W PSU as standard and that would be the bare minimum for the original specifications, an SSD has since been fitted but was the PSU also upgraded to allow for the extra power demand.

 

JAFYI: CPUTIN is the MB CPU temp sensor and AUXTIN is the power supply temp sensor if there is one, the max safe operating temp for your CPU is 62.3°C and the computer would/should shut down if CoreTemp (the sensor on the processor) reaches or exceeds this temperature, this unless the setting has been changed in the BIOS.

[mchain]

Hi philpower2,

 

Thank you.  Below find screenshots of Speedfan below.  First Speedfan screenshot is system at idle.  The rest are when the system was stressed for five minutes using HeavyLoad.

 

PSU is a Corsair CS550M installed May of this year.  OEM CPU was working fine, but these things get towards end of life.

 

BIOS shows exact same temps and voltages as Speedfan, and no, no settings changes there.  Default settings in use.

 

What is done to a system when sold as refurbished?  CPU paste replaced? 

 

System has never shut down unexpectedly.  Routine maintenance done every 2-3 months.  Could AUXTIN be the Northbridge chip temp?

Attached Thumbnails

• 
• 
• 
• 

[phillpower2]

Thanks for the detailed update mchain 

 

Refurbished could mean anything I'm afraid, cleaning out the dust for example, the thermal paste would not be replaced unless the HS and cooling fan are replaced.

 

TMPIN 2 = MCH = Northbridge.

 

If you have a small desktop fan and it is safe to do so (no children or pets around) take off the side of the case and direct the desktop fan so that it is blowing cool clean air into the case and then check your temps.

 

You have a good quality PSU there, consider fitting a low profile add on video card that has its own cooling fan as this will reduce the internal case temps and provide better graphics.

 

Above all getting rid of the hot air inside the case is a must, micro tower and other SFF computers are in the main business machines suited to office work, putting them under continuous load produces excess heat that cannot be expelled quickly enough from the cramped internal of the chassis which results in everything in there getting overly hot.

[mchain]

Hello philpower2,

 

Ran system idle for 17 minutes, cold boot.  

• First screenshot was several hours ago and is the BIOS Health screen.
• Second screenshot is @ 17 minutes; note the spikes in temp still there.  Lower minimum power consumption, tho.
• Ran HeavyLoad for 5 minutes 12 seconds, but third screenshot shows differing kernel load times between cores.  This one was taken @ approximately 4 minutes in.
• CPUID does show lower temps and no change in fan speed; also lower power consumption.

Thoughts?

 

Thank you.

Attached Thumbnails

• 
• 
• 
• 

[Plastic Nev]

Hi, till Phill can get back to us, my own gut feeling here is that there may be a poor sensor connection, a badly soldered joint on the motherboard or similar, rather than any real overheating problem as those temperature spikes are too short lived and too high for it to be a true temperature reading.

 

As for the temperature shown of 108C and then later at 133C, for CPUTIN, I would expect to find a burning smell if anything was really that hot.

 

Lets see what Phill thinks, though I do suspect a motherboard connection fault, and it would be near impossible to find exactly which connection without a full diagram of the motherboard circuitry. (Impossible to get hold of usually anyway). If I am right, a replacement motherboard is the only practical answer.

 

Nev.

[phillpower2]

Agree that a MB issue looks likely whether or not a faulty sensor is the cause I couldn't say for sure, I still feel that adding a fan cooled video card would lighten the load and drop the temps.

 

You are going very in depth into this mchain so it makes sense for me to ask what you see your options as, the PSU, CPU and memory are all fine so would you consider a replacement compatible MB.

 

FWIW: Integrated video chips are fine as a back up and light office type usage but they will soon heat up when put under load, this weakens them so please be careful and not overdo the testing.

 

Thanks for the additional input btw Nev 

[mchain]

@ Plastic Nev and philpower2,

 

Thank you both for your answers.

 

From what both of you say, it looks like a motherboard issue?  And no, no burning smell from indicated high temps.  Don't know what the timing intervals for CPUID and Speedfan are, but think Speedfan is much slower than CPUID and so never sees these temp spikes.  System has been doing this since day one and often wondered about the low specs of the +12 v rail.  This same value was present on both the OEM and Corsair PSU's in CPUID but not in Speedfan or BIOS, and so tended to disregard that reading.

 

BTW, caps look brand new.  No swelling, leaning off vertical, or leakage anywhere.

 

As this is an older motherboard, PCIe Express is version 1.1, and outputs only 25 watts at the slot max.  HP anomaly, I know, specifications call for 75 watts, but here only 25.  Obviously, is going to need a card that the system can boot with.  

 

Interestingly, BIOS default settings is set to boot to the PCIe slot, but the Intel chip still runs.

 

Thanks for the warning re running HeavyLoad.  Since the system never crashed, was wondering if it could cause damage to the system anyway.  Now I know that is a possibility.  Will heed that warning.

 

Reason for going into deeper detail is because what I do not know could be fatal to a system, so thanks for that.  I'll consider looking into an exact replacement motherboard in the next month or so, maybe?

 

[phillpower2]

The MBs own software can sometimes produce anomalies between itself and 3rd party monitoring software and in particular Speedfan, uninstall any testing software that you no longer require and then download and run MSI Afterburner and see if that helps with your fan speeds and voltages etc , details here disregard the overclocking information etc and only use the monitoring part of the software.

 

Interestingly, BIOS default settings is set to boot to the PCIe slot, but the Intel chip still runs.

 

 

That is how it should be and the reason is so that if an add on card is fitted it is detected on boot and the BIOS chip disables the MBs integrated video chip.

 

Going offline shortly but will check out add on video cards tomorrow, Nev may have ideas as to where and what is available meanwhile.

 

That is one of the most odd things with this, those sort of temps being reported would normally trigger a thermal shutdown either by the MBs thermal sensor/s or the PSUs thermal sensor, the present Corsair PSU for example is set to cut out at 40C.

 

You are welcome btw 

[mchain]

Ran MSIAfterburner only as monitoring software.

 

Intel GPU is set to max 256 MB RAM, i think, in BIOS.

 

Screenshots below:

 

A tech @ Micro Center recommended AMD Radeon 7770, but that requires a 6-pin connector (80w TDP) and there's the fact the PCIe slot is natively spec'd to 25w, not 75w, so didn't go that way because of that.  Not that there is no 6-pin available; in fact, there are two of them I could use but don't want to risk blowing the slot out.

 

Re the motherboard anomalies,  BIOS is one version behind the latest.  Latest version is recommended only if user sees a hang just before BIOS/MBR boot transfer, which I see from time to time, but executable is designed for XP/Vista and may not be compatible for Win 7 64-bit, as it is to be launched when Windows is running.  So I've left that alone.  Chipset drivers are the latest versions available.

Attached Thumbnails

• 
• 

[phillpower2]

I never do and never suggest updating the BIOS unless it is to address a specific/known problem.

 

Regarding add on video cards, one such thread that I found What graphic card's will fit my DX2300