why 2x agp might not work...

[scotty]

in the december pc gamer magazine there is a question in the hard stuff (tech) column about ramndom shut-downs on a particular reader's machine. the interesting part comes in greg vederman's answer. he says that the design spec for the agp slot is that it runs at 3.3 volts, and that some motherboards are only running at about 2.8 volts. the system in question has lock-ups after it has been on for a while.

my question is for all the techie know-hows out there. is that voltage spec for the agp slot correct, and if so could that be why me and so many others in these forums cannot get their agp 2x to run stably? when i use the reg hack to force 2x, the system doesn't even get part way through boot-up before it hangs.

thx fer yer help
8¬)

------------------
p2 266@300 (4x75mhz) | qdi legend-v, lx chipset, bios 2.0sl | 96 mb pc100 | mill g200agp@99mhz 8mb sg, bios 2.6, pd 5.3 | win 98se

[WebDude]

That 3.3 volt spec is correct. The BF6 allows adjustment of the "IO" voltage. The manual explains that the "IO" voltage is the voltage supplied to the DRAM, chipset and AGP. Using the Windbond Hardware Doctor software that comes with the BF6 (it reports the mb voltages etc.), it sets the alarm low voltage for IO at 2.99, and the high alarm voltage at 3.79. I'm not enough of a techie to tell you what effect a 2.8 volt AGP voltage would have, but I would worry about it, because it's significantly below the 2.99 alarm voltage.

[scotty]

i guess if qdi was up to snuff, i'd have some way of setting or detecting i/o voltage. but i don't. poo.

i do have a multimeter tho, anyone know where i'd stick the probes to read the voltage at the agp slot?

8¬)

btw- Kruzin: the reason i put this thread here instead of general hardware is that i'm trying to find out if it is a problem specific to mill g200, or some motherboards in general.