Gak

I've got news! I've found that with this 7600GS things do behave a little bit different... I've found this some minutes ago and i have done several times to be shure that it wasn't just a stroke of luck... With this card, the computer instead of just freezing (and having to reboot therefore), by pressing Ctrl+Alt+Del, or Alt+Tab allows me to return to the Desktop. Afterwards, with the TaskManager i shutdown the 3D application and with the aid of RivaTuner if i reload the display driver im able to reload the 3D application again (however with the same results, after a while (random i guess but no more than a minute, at max, sometimes a lot less) it crashes again. I've monitored religiously the temps now, and i've found that with 16Cº room temp, VF700 at full rpm's, there isn't even a "difference" between idle and fullload. 31Cº idle - 33Cº fullload However the card, as the other two, refuses to give anything more past 515/520Mhz at the core...

Hi again! About the temperatures they have always been around safe and fresh values. No problem shurely comes from them! About the psu rails, my systems is described in my signature... Do you think there are too much devices in the +12v line? I find it more easy to believe that "(...)it sounds as if you have just reached their limit on your board(...)". I don't have any friend with PCI-E boards around here. But i'll try to test this 7600GS later next week with a "friend" from MSN wich lives considerably near (far!). If it wasn't the third card with the same, exactly the same, Mhz limit, this wouldnt intrigue me. But it is repeating as it is some think that is limiting them below their own limit. Cause as i said, they don't even have a point where they start to give artifacts. They just crash. And btw, the core speed is 450Mhz... Not 400. I believe it's the "regular" version wich has 400Mhz. This is the "Sonic" version, more 50Mhz at the core and GDDR3 memory. I've already also put the card for sale hoping that someone buys it so that i can go and pick up another one... Just to be shure this isn't just a stroke of bad luck... Anyone has more ideias? Feel free to post them, please....

Hi! I don't know very well why i am writing this in the first place, but i guess that it is always because of that hope that i may be missing something. So i have an Ultra-D AB0. From what i know it isnt one of the best revisions. With this board i have had 3 graphic cards. Counting with this 7600GS. Sapphire X800GTO2 Connect3D X800GTO XpertVision 7600GS Now the odd part. All theese three cards, had their core Mhz limit around 515/520. No more, no less. In order to have a stable 3d environment i always have to find myself running them around 500Mhz core speed. Their memories never had problems. Always overclocked like hell. This is obviously a disappointment since every single one of those cards used to overclock like hell. Or at least in their vast majority, they all used to go up at least near 550Mhz at the core. Some of them even do 600Mhz. With the GTO2 it was rather easy to see, and the majority of the 7600GS are doing at least!, 550Mhz. Do you guys think this is a coincidence? A threesome coincide? Three graphic cards wich have precisely the same physical limits? Or should i be looking at the motherboard or some other bios configuration? They all present the same symptoms when past that 515/520 barrier. 2D mode is ok, i launch a game it is still all ok, as soon as i start to play, image freezes, sound enters a loop, then screen goes blank and im forced to reboot. With my old rig, before this Ultra-D, i was used to see some artifacts before reaching a point where the card would crash. Now, that doesn't even happen. It simply crashes. The three cards. Always behaving the same way. I have tried to pump a bit the PCI-E frequency, i've tried them in the second PCI-E slot with Sli enabled and disabled, iv'e tried with different coolers (VF700 and 900), i've tried different bioses (GTO2)... I just don't have any other ideia. And this is leaving me kind of sad.

I hadn't done it earlier cause i was assembling the rig back up again. But i assure you it's not my policy to leave unfinished businesses!

Done... You sounded, "angry", about my thread over at XS... :confused:

Nothing was plugged to usb/ps2 ports. I find it easier to believe in a corrupt bios... Thats why i imediatly reflashed it.

Don't ask me how, because i don't know the answer... This morning, i woke up, got past the computer and turned it on... Just for the kicks... Now, get surprised!, it booted! Like it would do normally... I didn't had a keyboard attached so, i turned it of, plugged the keyboard and turned it on again... And it failed to boot. I thought to myself that the board was somehow faulty. However i rebooted it once more, and it booted, went to bios and saved new settings... Did a Clear Cmos and flashed a new bios (623-3). Now it is working with no problem... Whattaf**k?????

All my friends only got AGP/300W systems. I'm in real trouble to test those... Testing the PSU with a DMM and all hardware plugged in even with this problem would be of any good?

So, i've done what RGONE said. What happened: After that the board booted and gave me just one led. As i though were better i installed one stick of ram and the gfx (didn't plugged the power cable to it). The board booted again, gave me one led and an intermitent beep. I hooked up the screen and saw there was a message in red (just that in the screen, no welcome/post whatever screen) sayin that i didn't connect the power cord to the vga. So i turned things of, plugged the cable, turned on and things went back to the same. 3 leds, no beeps, no reaction. After i tried one thing, removing all the extra power cables including that one near the ATX cable (the one with four pins). Also with the memory and gfx removed! I tried to boot, no go. 4 leds. So i plugged the 4pin cable again. Now the board does what i first mentioned, boots up and gives me one led. The rest happens just like i described above. I if repeat this step, i always get the same result. 3 leds, no boot, no beeps, nothing.

I'll try to do exactly as u suggested. But replacing the thermal compound of the chipset didn't voided my warranty?

Ok... Here i go... I bought my Ultra-D 27/09-2005. And she had been running flawlessly for the past couple months. Never had any of those weird problems that sometimes are described here. Recently, 02/08/2006, i bought a X2 3800+ and replaced my 3500+. Everything was going just fine... Despite my Windows installation was in need of a format! I went away for a week. And when i left i turned of my computer in the most normal way there is. Start>Turn Off Computer>Turn Off I turned of the PSU switch and unplugged the power cord from the outlet. I came back yesterday, and has an addict the very first thing i did when i came home was plugging the power cord, turn on the psu switch and power up the pc. To my surprise it had no reaction. 3 leds, chipset fan at full speed and nothing more, no boot screen, no beeps, nothing... I then thought to myself... "Hey... It must be some stupid thing that a clear cmos will resolve". I tried it (clearing cmos) and no go. So i tried even a longer one (overnight, +12h). For this clear cmos i unpluged everything from the mobo... Only things left were cpu and gfx. No go also. So, today, after lunch i took it of from the case and disassembled the rig. Blew some air to clear some minor dust and reassembled everything back up. Tested both sticks in all different slots. No go. Also tested them in a friends computer and they are ok, and the friends ram in my mobo gives the same result! Still no go. Same 3 leds, same behaviour. I guess this one has gone to meet the Creator, right? I must note that i was running my rig completely at default settings! I'm thinking in a RMA, however i've replaced the chipset's stock thermal compound for AS5, and now i have doubts if that doesnt void warranty... Should i try an hotflash? Despite i have no one around here with a similar board (i'll have to rely in some good soul in some other national forums where i hang around...) Any comments will be apreciatted! I'm totally in the dark here...

I even bothered to lap the cooler and polish it ... I even applied pressure to the cooler for some time... (Now imagine a guy 10 minutes with his hands sticked into the computer and staring at the screen (monitoring temps)...) I do believe it can produce temps 5Cº lower than stock cooler, but to do so it will have to be louder! I could manage to get good temps as i said, but it would do to much noise. At an acceptable level of noise it would give me higher temps... Now imagine... At full load (that is how the computer always is (crunching Rosetta) and with the case closed i get around 44/47 (depending on ambient temp (day/night)), with the Vantec i was getting 47 degrees (in the afternoon) idling... Full load it would easily go past 50Cº. And for me thats no good to a 24/7 rig... DFI bios does have options to control fans, however they are very limited... We can only manage to get the fans to spin at a certain speed or full load... For instace my Zalman 7000B. I could only manage (and i tried alot of configs) to get to work around the following rpm's: - the lowest possible - +/- 1600 rpms - full speed Same applies to chipset... It is very difficult to get to a half way in throttling the fans... Not to mention that my chip7 has been endangered today a couple of times... I really like Vantec's aftermarket look with my DFI however it simply doesn't do the job... Spent 14€ for nothing... :sad: Long time ago i remeber to have had a computer wich i couldn't hear three meters away... :sad: :sad: :sad: :sad: I wonder if anyone would buy me an Evercool (i would instantly pay trough paypal or wire transfer or something else) and ship it to Portugal... I would give a reward!

Seems to me then that it is not the motherboards fault rather the rpm sensor of the fan wich is unable to detect rpm's slower than 1470 (to be precise). Thanks for you info's!

To the various questions: The cooler is this: Vantec CCB-A1C Iceberq http://www.dfi-street.com/forum/showthread.php?t=58366 I also tried a variety of bios configurations for the chipset fan. With no result. When i was able to get the same temperature than the stock cooler for idle/full/gaming etc, the vantec was noticeable louder... I've reverted to the stock cooler for now...

Well... Just for the info! Did some lapping... Nothing too fancy... Used a dremel to polish and give some more clearance on the clip holes. (Although one side has now a bit too much space... But it works... It is easier to install and remove. However im still getting @ idle, temps 3Cº above stock cooler....