• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Albatron PX845PEV pro and case temp

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

bob63

Registered
Joined
Nov 4, 2001
Where is the case temp taken on this mobo. Is it taken at the Winbond W83627hf chip, a sensor on the mobo, or is it taken off the north bridge much as the CPU temp is taken? Since I've watercooled my case temps actually went up 2 degrees C. The only reason I can see is I now have all these hoses that could be blocking some air flow. I did install a Thermaltake Tiger 1 hs with fan inplace of the stock passive hs on the Northbridge, but it didn't lower my temps at all. This leads me to believe that the case temp is not taken off of the north bridge, but by another sensor on the mobo. I could be that prior to WC I had more airflow across this sensor then I do now. That would acount for my higher readings now, but I'd like to be certian of where this temp is taken.
 
i don't know if this helps but i use motherboard monitor to check my temps while in windows and the cpu is set up as winbond 1, the case in winbond 2.

As for case temps, if your pump is inside your case, like mine, and is not in a resoivor, it is likely disappating the heat into your case, thus increasing the case temperature, those things can get hot, I ended up having a fan to push fresh air onto it to get rid of some of the heat faster.
 
Last edited:
No my pump is external of the case, so I can't see this being the problem. I am also using MoboMon5, however I'm using Winbond 1 for case temp, and winbond 3 for CPU temp. These are the setting that are suggested on the MoboMon site, and they also most closely match the temps given in the bios. I've E-mailed Albatron to inquire as to the correct settings. I'm also beging to suspect that Winbond 1 "case temp" is read internally on the Winbond chip, not the North Bridge. Thanks for the help though.
 
I'm still leaning twords the idea that winbond1 reads the internal sensor on the winbondchip and is not a reading of the northbridge as some have posted. This would explain why improved cooling of the northbridge did not show a decrease in case temp. I'm going to try and rig up a fan that blows directly on the Winbond chip and see if this lowers case temps. I tried to get the info directly from Albatron but was not succesful. They responded right away but did not really answer the question.

Dr Pro,
I am trying to get a better idea of where my system is at temperature wise, but I'm getting some conflicting answers from various sources and was hoping that you could clear things up. My understanding is that case temp is measured on Winbond 1, CPU temp on Winbond 3, and that Winbond 2 is not used, nor are any of the diode channels used. I should mention that I am using Motherboard Monitor 5 to monitor temps and these are the settings they recomend. These setting also most acurately match the readings in the bios. Another point of confusion is as to what is meant by "case temp", is this the internal sensor in the Winbond W83627HF-AW chip, a temp reading from the Intel northbridge, or another sensor. One last question. There was a rumor a while ago that Albatron was preparing a Windows based temp monitoring program for their boards, can we expect to see this any time soon.

Regards and thanks,


Dear Customer,


Thank you for using our product, and sorry for the inconvenience.
Actually,we haven't send our product for testing with hardware monitor utility,and it will detect some error information.We are developing our hardware monitor utility.We will release it soon.
Please check it.
If you have further questions,please do feel free to contact us.
Thanks!


Best Regards.
 
Hmm.. I never really checked where it measures case temp...



But, you say you 've gained 2c since going to watercooling? You have to remember, you've now lost airflow from the CPU heatsink fan. This usually results in higher temps of components near CPU.
 
Here's my reply...

Dear Customer,

Thank you for using our product, and sorry for the inconvenience. Due the we put the system sensor nearby the power solution, and it should be higher then cpu, and it's normal. Don't worry about it. Please check it. If you have further questions,please do feel free to contact us. Thanks!

Best Regards.


My question was this...

>I recently installed a PX845PEV Pro and updated the BIOS to 1.08. The System temperature and the CPU temperature appear to be switched on the HW Monitor page in the BIOS. My System temperature is listed as 40C and the CPU temperature is listed as 32C. Is this correct? I believe the cpu temperature should be 40C and the system temperature 32C.
 
Well you certainly got a better explanation then I did. Not exactly sure what they mean by "nearby the power solution". Do they mean the top right corner of the mob, or near the ATX connection, or possibly even the 12v connection. I looked over the board and did not see any obvious sensor, but thats not saying there isn't, just that nothing stood out to me as being a sensor. The Windbond chip is located to the left and below the north bridge, but that may or may not be where the actual case sensor is. Anyway it is good to know that the case temp should be higher then the cpu temp. This also confused me, but so many others were reporting the same thing, I just accepted it.
 
I was a little confused by the email as well. I guess the system temp is not being reported by a onchip northbridge sensor.

I thought I had killed my Abit IT7 Max, which is why I installed the Albatron. The IT7 is notorious for reporting incorrect temps, and when I hooked up the Albatron... same temps but they were flip flopped. I just assumed that the BIOS readings were switched. If what this rep says is correct, then my IT7 was reporting temps incorrectly. I guess I'll believe Albatron since they've had 3-4 BIOS releases and they haven't changed anything.
 
For anyone who's suspicious of their cpu temp in MBM, the cpu temp should be read from Winbond 2 Diode. I know Winbond 3 looks like it's the cpu temp, but it's not. Winbond 2 diode still under-reads by 3-5C though.

As for the case/system temp, I'm pretty sure it's read from the VRM location on the motherboard, which is usually hot.
 
NookieN said:
For anyone who's suspicious of their cpu temp in MBM, the cpu temp should be read from Winbond 2 Diode. I know Winbond 3 looks like it's the cpu temp, but it's not. Winbond 2 diode still under-reads by 3-5C though. ]


I just tried the Windbond 2 diode setting in MBM and it reads 11°C. Now don't get me wrong, I'd be ecstatic even if it was reading 5°C low, but some how I think that that 32°C I'm getting on Winbond 3 is closer to the truth.

NookieN said:
As for the case/system temp, I'm pretty sure it's read from the VRM location on the motherboard, which is usually hot.

What is the VRM, and where is it on the mobo? I agree that it would have to be a hot area as my case temp is about 10°C above my CPU temp.

Just out of curiosity, are you using a PX845PEV pro? If not you may be correct that Winbond 2 diode reads CPU temp on your motherboard, but not on the 845pev. My understanding is that the mobo engineers have some options in how they configure these chips and that varies not only between manufacturers, but even between different models from the same manufacturer.
 
I'm using the 845PEV Pro. I thought Winbond 3 was the right one too, but when I tried stopping the fan it just stayed around 50C. When I stop the fan on Winbond 2 diode the temp goes up to ~62C (and the chips starts throttling, which makes me think it under-reads).

But yeah 11C is too low to be believable. I suppose the sensor placement can vary from board to board. I actually switched to W2 diode because that's what the MBM website said was the correct sensor for that board.

The VRM is the voltage regulation module. It's the collection of MOSFETs and capacitors near the cpu that keep the voltage and current supply stable. The MOSFETs can get pretty hot during full load, so I suspect the sensor is near one of them.
 
NookieN said:
I actually switched to W2 diode because that's what the MBM website said was the correct sensor for that board.

I looked at the site again because I thought it had said to use Winbond 3. Winbond 2 diode is listed as CPU for the PX845PE (not the PEV) and Windbond 3 is listed for the 845PE pro 2 and the PX 845G (again neither of these is the 845PEV either). Unfortunately the 845PEV is not listed. Are the temps you get on Winbond 2 diode, plus correction, close to the temp seen in the Bios? As noted mine are way off which is why I originally dismissed the Windbond 2 diode. Wouldn't the 'diode' imply that there was a temp diode in the middle of the ZIF socket? I thought the PIV's read off of an internal sensor. Your comments about the fan are interesting and I need to do some experiments of my own. I recently switched to water cooling (CPU and GPU). My CPU temp did not change but my case temp went up a couple of degrees. I attributed this to the hoses blocking air flow to the sensor. I have since removed a couple of fans, and lowered the voltage on the others. This caused an increase to both temps. Unfortunately I was not monitoring W2diode at the time so I would not be able to say if this one changed or not. Interesting...
 
bob63 said:

Winbond 2 diode is listed as CPU for the PX845PE (not the PEV) and Windbond 3 is listed for the 845PE pro 2 and the PX 845G (again neither of these is the 845PEV either). Unfortunately the 845PEV is not listed.

Yeah I noticed that too. But as much as I can tell from Albatron's website, they don't make a PX845PE. There's just a PX845PE II and PX845PEV. So I figured either the MBM website has a typo or the PX845PE was on the market for all of a week.

Are the temps you get on Winbond 2 diode, plus correction, close to the temp seen in the Bios? As noted mine are way off which is why I originally dismissed the Windbond 2 diode. Wouldn't the 'diode' imply that there was a temp diode in the middle of the ZIF socket? I thought the PIV's read off of an internal sensor.

Well, they're close-r to what the bios says. I wouldn't say they're dead-on though. The bios actually seems to read a little higher than W2 diode in MBM. Winbond 3's idle temp was close to the bios's I think, but I grew suspecious of it when my board was resetting itself with the fan off and the sensor only read 50C.

I think the P4 has both a thermal sensor and a thermal diode on-chip. The thermal diode is what the motherboard should be reading.
 
I always used # 1 & #3 due to those being almost dead-on to what was in the bios. (PRO)

I never really bothered to verify temps, but I might have to get out my meter & check for myself...
 
why dont everyone post what sensor or winbond there using on there 845PEV Pro. this is getting me confused and curious too see if im using the wrong winbond.

Sensor 1: ?
Sensor 2: ?
Sensor 3: ?

me i just wrote down what the temps are in the bios and tried to match it.
 
I have bios 1.08 and the reason i read the temperature for the core at winbound 1 is that when prime 95 or burnin wizard or someprogram like that runs this is the sensor that gradually moves up and down like my cpu cores temperature would. Winbound 2 reads 69 C and I have checked my bios and this is overcompensated by 30 C I consider this to be the case and the last temperature that I see is Winbound 3 which is at 127 never changing.

If what you guys are saying and winbound 2 is the sensor, what is the compensation that I should run it at?
 
I use
Sensor 1 = Winbond 1
Sensor 2 = Winbond 3
Sensor 3 = Winbond 2 Diode + 15C compensation.

There is nothing I can see that is case temp. Sensor 3 with the comp is CPU temp. I'm not sure if 15C is enough. I think it should be more like 18 - 20C. I'm kind of judging by the EPoX mobo I had before this one. Idle CPU remp was ~ 36 - 40C, load ~ 55C.
 
sensor 1 - winbond 1 - 35 idle - 44 load
sensor 2 - winbond 2 - 14 idle - ??? load
sensor 3 - winbond 3 - 30 idle - 39 load

I have thought of getting one of those external thermometers that you attach between your hsf and the processor but I havn't had the extra money to get one yet. Maybe that would help but I don't know. All that I know is albatron made one hell of a board and I love it =D
 
I've been using Winbond 1 for System and Winbond 3 for CPU myself. I also have a sensor set to Winbond 2 diode + 15 but that number always seems a bit screwy. The 1 and 3 temps agree with the BIOS temps pretty closely.

The big problem I have is that the motherboard list at the MM5 site doesn't have an entry for our Winbond I/O chip. That sensor is W83627HF I believe. They only refer to 37HF. A bunch of Abit boards use this chip and most of them refer to Winbond 1 as case and Winbond 2 diode as CPU.

And of course Albatron isn't much help in this matter. I've emailed them a few times and they respond that they are writing a Hardware Monitor and to use the BIOS temps. Maybe they don't know which sensors either...
 
Back