- Joined
- Feb 13, 2001
- Location
- Twin Cities
I posted this late Saturday night, but it disappeared.
For those of you who have endured my whining about my previous KT7A not being able to run its FSB faster than 133 Mhz reliably and not even post above 136 Mhz.
This was despite trying:
All multipliers 5x-12.5x
Different core and i/o voltages
Different bios settings, both conservative and aggressive
Different CPUs
Purchasing PC150CL2 Ram
Purchasing new 400W PSU
Pulling out different cards
3 different video cards
3 different bios revisions, 2 different O/S and 3 different VIA 4in1 drivers
Cooling never was a problem (in the 30s)
Well, I convinced NewEgg to send me another one and RMA the old one. Their 30 day satisfaction guarantee along with reasonable prices and shipping have now garnered a loyal customer.
Right out of the box, it fired up at 7.5x150. I worked up to 10x150 2-2-2 and joined the ranks of the 1500 club. This is with a 1.0G/266 AXIA chip. A 50% OC !
I feel vindicated but burned out from weeks of troubleshooting, as time provided. Some people raised the question as to whether the old KT7A, being unable to OC the FSB past 133 constituted a valid RMA request. Your damn right it did. I expect a motherboard that I paid hard-earned money for to work with more than 2 FSB settings in the bios (100/33 and 133/33) !
Of interest to conspiracy theorists, the northbridge chip in the bad one was entirely different from the good one. In the old one, it had a metal top on it with the ID printing, or perhaps a metalized sticker (hiding what?), but in the new one it is the conventional all plastic type like the southbridge.
Of interest to bios revision trackers, the new board came with YH on it. The old board came with WW on it and overclocked the best with WW. The new one worked equally well with WW, WZ and YH, with this exception: Using WZ or WW, when I enabled fast command decode along with enhanced cpu, it would not even post. With those two revisions, having fast command decode disabled had no effect on Sandra 2001se benchmarks. With the YH revision, I could enable both with no problem.
I can't get the board to work at 155 FSB, but that is running the other busses pretty hot. As it is, with the FSB at 150, the GPU on my Gainward (read cheapo) GF2MX gets pretty warm running 3D Mark 2001. If I intend to run protracted amounts of 3D gaming, I will have to upgrade the cooling solution from a dinky little HS, held on with double-stick tape to something better. Luckily, I'm not a gaming nut.
Some benches:
Sandra 2001se MEM 627 701 / CPU 4177 2050 / MMX 8151 10,219
3D Mark 2001=3027 GF2MX (not overclocked internally)
So, the old board goes back Monday and I am very happy with my KT7A that meets my expectations. If you are using a KT7A with similar problems, don't discount the possibility that there is something wrong with it, though somewhat veiled by semi-operationality.
Hoot
For those of you who have endured my whining about my previous KT7A not being able to run its FSB faster than 133 Mhz reliably and not even post above 136 Mhz.
This was despite trying:
All multipliers 5x-12.5x
Different core and i/o voltages
Different bios settings, both conservative and aggressive
Different CPUs
Purchasing PC150CL2 Ram
Purchasing new 400W PSU
Pulling out different cards
3 different video cards
3 different bios revisions, 2 different O/S and 3 different VIA 4in1 drivers
Cooling never was a problem (in the 30s)
Well, I convinced NewEgg to send me another one and RMA the old one. Their 30 day satisfaction guarantee along with reasonable prices and shipping have now garnered a loyal customer.
Right out of the box, it fired up at 7.5x150. I worked up to 10x150 2-2-2 and joined the ranks of the 1500 club. This is with a 1.0G/266 AXIA chip. A 50% OC !
I feel vindicated but burned out from weeks of troubleshooting, as time provided. Some people raised the question as to whether the old KT7A, being unable to OC the FSB past 133 constituted a valid RMA request. Your damn right it did. I expect a motherboard that I paid hard-earned money for to work with more than 2 FSB settings in the bios (100/33 and 133/33) !
Of interest to conspiracy theorists, the northbridge chip in the bad one was entirely different from the good one. In the old one, it had a metal top on it with the ID printing, or perhaps a metalized sticker (hiding what?), but in the new one it is the conventional all plastic type like the southbridge.
Of interest to bios revision trackers, the new board came with YH on it. The old board came with WW on it and overclocked the best with WW. The new one worked equally well with WW, WZ and YH, with this exception: Using WZ or WW, when I enabled fast command decode along with enhanced cpu, it would not even post. With those two revisions, having fast command decode disabled had no effect on Sandra 2001se benchmarks. With the YH revision, I could enable both with no problem.
I can't get the board to work at 155 FSB, but that is running the other busses pretty hot. As it is, with the FSB at 150, the GPU on my Gainward (read cheapo) GF2MX gets pretty warm running 3D Mark 2001. If I intend to run protracted amounts of 3D gaming, I will have to upgrade the cooling solution from a dinky little HS, held on with double-stick tape to something better. Luckily, I'm not a gaming nut.
Some benches:
Sandra 2001se MEM 627 701 / CPU 4177 2050 / MMX 8151 10,219
3D Mark 2001=3027 GF2MX (not overclocked internally)
So, the old board goes back Monday and I am very happy with my KT7A that meets my expectations. If you are using a KT7A with similar problems, don't discount the possibility that there is something wrong with it, though somewhat veiled by semi-operationality.
Hoot
And last but certainly not least, WOW, nice OC!!!! Keep going and good luck.
