Printable View
My vcore fluctuates too,seems to get worse the higher i raise it.
It seems to be my IDE that won't let me run my fsb higher.I have tried 4 diff hdds:
WD =214fsb
Seagate =214fsb
IBM =212fsb
Maxtor 80g ata133 8mb=217
All drives 7200rpm, ata100, 40g except the Maxtor
I tried turning off dma and going pio on a couple of them but it didn't make a diff.
omg my board sucks, your lucky as heck with 214fsb and stuff. i cant even do 180! LOLZ
Sorry about late delays in replying here guys...I've been out of town for the last 4 days.
Anyway-
sjohnson: Glad you sorted out that bandwidth issue:)
bigdawginva:
As far as voltage fluctuations are concerned, I also experience mild fluctuations. When set at 1.90v the range is 1.89 to 1.95v. These are the same type of fluctuations I had also experienced with my KD7-RAID. However, per my discussions with Abit (while conducting my KD7-RAID Mobo Review for Liquid Ninjas), I was told it was more the result of voltage sensor readings and not so much actual voltage changes. It was further explained to me that due to the very high voltage choices available on the mobo, the voltage sensor had to be calibrated differently and resulted in a "very sensitive" sensor:) Although I cannot confirm or disprove this, I've had no ill effects while overclocking at all. So I assume it is not a serious problem.
Thanks Mike. Great expalanation.
:(
My new KD7-E wants to go like the wind but I think, from what I have read here, that my 256mb of Crucial Registered PC2100 can't make the grade.
I can get my XP1600 close to 1600 going into WinME, good temps from my HS/CPU but page changes and slow mouse lead me to believe my memory is not making the grade.
I tested it on "memtest86 v 3.0" and it passed all the tests.
Current mem settings:
2.5.
no
3
7
3
2T
Enhanced performance
I will be adding a T-bred later this year.
Your opionions are appreciated.
Memory recommendations also.
Thanks,
weetoots
KD7-E v1.0 CY
XP1600 AGKGA
Radeon 7000 (don't laugh)
:confused:
I would get some Corsair XMS 3200 or 3500. I've built a good number of systems on the KD7-E now, and all worked great with the Corsair. I also tried some Kingston and Samsung which worked good too, but the XMS stuff just went a bit higher with less voltage.
Once you get good memory, just drop the multiplier and up the FSB. Get at least 3200 memory to give yourself some headroom to work with.
My manual for the KD7-E rev 1.0 on page 3-12 shows the pin header incorrectly. The pins (as shown on the page) are show from top to bottom as 1-2-3. However if you look at page 3-8 and fig. 3-7 the pin arrangement for CROM1 is actually, form the top of the page 3-2-1.
Maybe this is why some people are not getting this set up correctly. :rolleyes:
weetoots
Mine came from the vendor set wrong. I was simply stupid for not reading the manual lol...
sjohnson,
You can't be stupid if the manual is wrong! :toast:
No bud, I *was* stuupud - if I had read the manual I would have know that the setting enabled my "missing" BIOS setting for CPU Decode. Then, even if the manual showed the wrong setting I would have tried it the other way around, just to check if I had a bum board or something.
ABIT has a history of small manual booboo's. In any case, I've beat myself up for the last time (on this problem anyway :D )
received my XMS3200 today.. and guess what.. the kd7e doesnt even BOOT at 190mhz...well done abit :( i tried evrey bios and the Pio mode and..and..and..
this "new" kt333 suxx
anyone have this board an a SB Live card? i wanna know if the SB Live card can handle 40mhz
I do. Im at 209 mhz FSB for benching and 200 for daily use. never experienced a hitch.
thanks. thats what i wanted, 200fsb for 24/7
What kind of CPU voltage do you need for stability @200fsbQuote:
Originally posted by DuraN
I do. Im at 209 mhz FSB for benching and 200 for daily use. never experienced a hitch.
aaaargh wish i was lucky with that mobo :/
do you use a raid controller??
I can't seem to get to any bus speeds over 175 with the 1/5 divider and the 1/6 doesn't work.
Are any of you using the non functioning 1/6 ? There is no pci/agp lock for this board is there? I'm using the spd dram clock and spd ram timings so the memory shouldn't be the prob. I'm using the ocz 3500 el.
have u tried lowering the command decode?
I was using the comand that is used by default after clearing CMOS, which I thought would be of relaxed timings. Let me check though.
Mike,
No one is using the 1/6 divisor. It does not exist in the KT400 (or new KT333) chipset.
In fact, if you are seeing that 1/6 divisor choice in the BIOS, then you have the orginal BIOS release. I strongly recommend that you flash to the most recent BIOS. I find that the C8 works best for me. You will find the original release BIOS generally sucks and that you need to flash to a newer BIOS for improved overclockability.
You definitely want to be running the CPU Command Decode at Normal.
swissocer:
I'm currently running a Rocket RAID ATA 133 card with RAID 0 on the KD7-E. Works great all the way up to my max FSB at 206mhz. After 206mhz, the Radeon starts to complain.
Keep in mind that with only a 1/5 divisor, and PCI card, graphics card, or HD that does not like to be run out of spec may be holding you back. I've done a good handful of systems now based the the KD7-E, and have not had one that will not go at least high 190s low 200s. If you are sure about your hardware, you may have a bum board.
Thanx Mike,
Flashed the bios to C8, definite improvement....is there a fix for the data corruption issue? I was fine with 190x14=2660 but next try was 198x14= windows toast. The chip (xp2700+) is not modified...ie: no L3 mod.
Have you ever run the chip at 2.77GHz? That seems like an awful big jump, 190 - 198 just like that.
It depends on what's causing the corruption.Quote:
Originally posted by mike.elmes
Flashed the bios to C8, definite improvement....is there a fix for the data corruption issue? I was fine with 190x14=2660 but next try was 198x14= windows toast. The chip (xp2700+) is not modified...ie: no L3 mod.
The KT400 and "new" KT333 both suffer from HD corruption at high FSB speeds. There is no real fix, although the problem can be lessoned.
First, make sure you have adequate core and memory voltages going on. Unstable cpu and memory can cause the corruption to come on much sooner. Particularly with the memory. So see if rasing the voltages help. If no relief there, try dropping the memory timings a tad.
If no joy from any of that, than you are faced with setting your HDs to PIO mode if you want more FSB. That will help but it is your choice if you want to do that.
If none of that helps, it may be that your HDs are sensitive to running out of spec with the 1/5 divisor.
BTW, SPQQKY has a good point about bumping your FSB up in smaller increments. That's the best approach.
Thanx for the help on this....
I am going to try ramping up a bit slower and I'll add more voltage to both cpu, and ram.
I am now running windows 2k and 3dmark is finicky on that os...so I don't know if I can clock a bit higher. I'll give it a go and see if I do better under 2k.
I am still pretty happy with 190x14=2660....it sure is fast, 10,000 mips in Sandra:)
Cant wait to get back at her...I'll post my results.
mike
I have a Morgan 1G ,KD7-E, OCZ RAM PC2700 256m, they work well on FSB166. Indeed, I do benefit from your recommendation about KD7-E.Quote:
Originally posted by mdzcpa
It depends on what's causing the corruption.
The KT400 and "new" KT333 both suffer from HD corruption at high FSB speeds. There is no real fix, although the problem can be lessoned.
First, make sure you have adequate core and memory voltages going on. Unstable cpu and memory can cause the corruption to come on much sooner. Particularly with the memory. So see if rasing the voltages help. If no relief there, try dropping the memory timings a tad.
If no joy from any of that, than you are faced with setting your HDs to PIO mode if you want more FSB. That will help but it is your choice if you want to do that.
If none of that helps, it may be that your HDs are sensitive to running out of spec with the 1/5 divisor.
BTW, SPQQKY has a good point about bumping your FSB up in smaller increments. That's the best approach.
Now what confuses me is the booting problem. After unlocking my Morgan, I enjoy changing the multiple (sometimes higher sometimes lower), therefore, very often I can¡¯t boot when I turn on or reset my system. So I looked through the manual of KD7-E¡_¡_.
the solution seems to be offered by ABIT in the manual (page 4-5 ¡°solution in case of booting problem due to invalid clock setup¡±):
¡°Normally, if the cpu clock setup is wrong, you will not be able to boot. In this case, turn the system off then on again. The cpu will automatically use its standard parameters to boot. You can then enter the bios setup again and set up the cpu clock. If you can¡¯t enter the bios setup, you must try turning the system on a few times(3-4times) or press ¡°INSERT¡± key when turning on and system will automatically use its standard parameters to boot. You can then enter bios setup again and set up the new parameters.¡±
I wonder whether other users are suffering the same or I am unlucky. Although this solution can work (sometimes not 3-4 times but more ), I don¡¯t think it is a perfect one. It seems to means the responsibility lies in the users¡¯ overclocking (not standard CPU parameters). But ABIT always encourage us to oc our CPUs.