Heatsink for Opti 165 inside a SG01-E

R

RanceJustice

Supreme [H]ardness
2FA
Joined
Jun 9, 2003
Messages
7,110
Hey guys. Ever since I switched to my Foxconn mobo, I've been having heat issues, with boot temp around 40c or so. I've remounted my Zalman 7000-B CU , but so far it hasn't made too much difference. Looking at the bottom of my zalman, it doesn't look as... uniform? in color as it once did and I'm wondering if this is the problem. Even after cleaning it with ArcticClean, the irregularities still exist.

At this point it seems to me there are two possibilites - the Foxconn just reports high temps (as I didn't have this problem with the previous Biostar board) or something has gone wrong with my heatsink between/during installations. As such, I'm looking for the best heatsink+fan assembly that will fit inside a SG01-E setup. Another Zalman? a Scythe?

Any suggestions? Of course, height is a consideration here as the power supply is right above in this case setup. Thanks
 
It could also be that your Zalman is blowing hot air on to your temp sensor.
 
mattro said:
It could also be that your Zalman is blowing hot air on to your temp sensor.
Click to expand...

I was under the impression that all mobos had internal temp sensors in approximately the same place relative to the parts they monitor. Would the temp sensor of the Biostar have been in a different location than the current Foxconn?

Also, looking at the XP-120 and NT06 - would these work well enough fanless? I've always heard that the XP-120 had to have a fan with it, and that the NT06 would have to have the fan removed to fit in the case? I always heard the XP-120 was HUGE, but I'm guessing it would fit better than the SI-120 model?
 
Ever thought that the Foxconn is registering the correct temps and the Biostar was reading the incorrect ones? I get around 36-38c idle with a 146 @ 2.9GHz 1.5v with dual core opty hsf in a TG08. Have you tried the boxed cooler at all?
 
Firelord said:
Ever thought that the Foxconn is registering the correct temps and the Biostar was reading the incorrect ones? I get around 36-38c idle with a 146 @ 2.9GHz 1.5v with dual core opty hsf in a TG08. Have you tried the boxed cooler at all?
Click to expand...

Afraid I don't have an OEM cooler. I'd be fine if they stayed in the 40s idle, if they didn't go over 55c at load. I think that's the main problem, the load temps go up to 55-72 if I'm to believe Speedfan (Everest gives me odd readings). Maybe the biostar was faulty :confused:
 
Its well over 55/60 at stock, full load with 2 instances of SP2004. OCing up to 2.2 or so, with clockgen, it goes up to 71! Stock volts too. This is compared to the biostar that I could OC to 2.5+ at 1.450v with temps in the low 50s.
 
Not sure which one is right tbh. Have you tried contacting foxconn at all? Any chance of trying a thermometer at all?
 
I've used most of the SFF boards out for 939. You didn’t have that problem, because the Biostar lies. The Biostar reports absurdly low temps. It probably fudges by at least 10° under. The Jetway I had overreported by about 10°. I think the Foxconn was somewhere in the middle, probably the closest to actual. Then again, maybe it was the Foxconn that overreported...

Are you actually having heat issues or do you just not like the temps it reports? You're not suddenly running hotter, just because you are on a new board! If you are actually running @ 72° then your system would be hardcore crashing. Seriously, don’t worry about it. AMD’s new SSDOI process seems to prevent any real gains with added voltages and so most of are well within the safe range for volts. With lower volts, your chip can handle higher temps.

Seriously, how many people have you heard lost a chip in the last week, month, year from too hot temps? Zero, probably. The safety margins we use are so uber-anal it is not something you should even think about, unless you are doing something uber stupid like running without a heatsink.
 
I think sleepeeg is right the foxconn imo is about right for temps and wouldn't surprise me if it is the most accurate cos they are a huge company making for many OEMs so would expect it to be accurate
 
sleepeeg3 said:
I've used most of the SFF boards out for 939. You didn’t have that problem, because the Biostar lies. The Biostar reports absurdly low temps. It probably fudges by at least 10° under. The Jetway I had overreported by about 10°. I think the Foxconn was somewhere in the middle, probably the closest to actual. Then again, maybe it was the Foxconn that overreported...

Are you actually having heat issues or do you just not like the temps it reports? You're not suddenly running hotter, just because you are on a new board! If you are actually running @ 72° then your system would be hardcore crashing. Seriously, don’t worry about it. AMD’s new SSDOI process seems to prevent any real gains with added voltages and so most of are well within the safe range for volts. With lower volts, your chip can handle higher temps.

Seriously, how many people have you heard lost a chip in the last week, month, year from too hot temps? Zero, probably. The safety margins we use are so uber-anal it is not something you should even think about, unless you are doing something uber stupid like running without a heatsink.
Click to expand...


Well, I wasn't experiencing graphical corruption or failed Prime tests but the system alarm/bell started going off bigtime around 70c, which is when I shut it down. What do you have the warning threshold in BIOS set to? Oh - just for reference I'm running an Opti 165. When I was on the (lying) biostar, I had to up my voltage a little to get over 2.5ghz. Would it be safe to do that again, if neccessary? I've read bits and pieces about how the Dual Core Opterons/X2s are more heat and voltage sensitive etc...

Thanks for your comments Sleepeeg3 - I'll see if I can get the system stable without worrying about temps.
 
Okay, I've been ignoring temps, which are reported in the 60s, in order to test my CPU under Prime. It has been failing prime at 2.5ghz and 1.4v, so I upped the "Hammer FID" voltage to 1.45. Unfortunately, when I get into Speedfan it still says that Vcore1 is 1.39. Is that a quirk of the board? Is it really at 1.45? I'm thinking its not, as previously on 1.45 I could get up 2551-2601mhz , and now prime crashes at only 2501.

Also, what does the "Vcore" setting in the bios signify? It seems to have adjustments in mA (or was it mV) instead of voltage, and in small incriments. What I normally think of CPU voltage is listed as the Hammer FID Voltage and has the normal 1.35, 1.4, 1.45 etc.. voltage designations.
 
I had this issue with my Foxconn 6150 board, and it died today, so I'd beware. >_< Went to bed converting a DVD to AVI, and woke up with "DVI No Input Detected" on my monitor, and my computer refusing to POST.
 
BAH! Well, thanks for letting me know. I'll be on alert :rolleyes:

Anyone else have insight on changing the voltage? or what that BIOS vCore setting is? Chipset Voltage maybe?
 
Hmm.. Speedfan and CPU-Z both report the voltage at 1.39 still. I tried bumping it up to 1.475 in BIOS, which was the topmost voltage I used on my previous Biostar mobo. Still no go. Is there any reason for this to happen? For whatever reason, it seems "locked" at 1.4. Well, in other news temps have come down a little, probably from arctic silver kicking in.

Without upping the voltage, I don't think I can get my CPU past 2.4 :(
 
You must log in or register to reply here.
Back
Top