Nforce2 No boot after overclock? Look here!

Question

MSI follows the rules with nForce2.
Found at [URL]www.theinquirer.net[/URL]
Quote
Nforce 2 BIOS problems confirmed and fixed
The perils of ignoring the reference design
By Arron Rouse: Tuesday 11 March 2003, 19:10
A SHORT WHILE AGO WE REPORTED that there could be a serious problem with Nforce 2 chipset motherboards. No sooner had the article gone up than Nvidia got in contact wanting to know all of the details.
Users of a new Shuttle Nforce 2 based machine had reported problems with the BIOS. They had been trying to overclock their machines only to find that, once they went past a certain point, the machine would stop working.
At first it looked like a problem with the Shuttle but then Asus motherboard owners reported the same problem. Then Abit and Epox. Not only would they stop working, it was proving difficult to get them working again.
Nvidia stepped into the breach and admitted that there was a fault with the Shuttle and Abit boards. The fault lay in the manufacturers missing off a jumper that Nvidia had specified on the Nforce 2 reference design.
Hard Tecs 4U, a German site, unwittingly uncovered the missing jumper. They had reviewed six Nforce 2 motherboards from different manufacturers and managed to kill all of them. The only one that was easy to recover was an MSI K7N2-L board which has the jumper. Using the jumper sets the BIOS back to a 100MHz FSB safe mode.
Other people had discovered that you could use a 100MHz FSB processor, for example a Duron, in a motherboard that had stopped working to get it going again.
And there lies the problem. Setting the FSB too high can stop Nforce 2 motherboards from POSTing. Once the motherboard has stopped POSTing, the only way to get it going again is to reset the FSB back down to 100MHz. Only most motherboards don't have the jumper so you can't do that.
Nvidia obviously knew about this problem because they built the reference Nforce 2 motherboard with that jumper. The fact that many manufacturers didn't put the jumper onto their motherboards means that someone deserves a slapped wrist at the very least. Whether that's Nvidia for not emphasising the need for the jumper or the manufacturers for not including it is something we may never know.
However, what we do know is that a fix will soon be at hand for everyone. Nvidia has managed to create a fix for the BIOS on Nforce 2 motherboards that will solve the issue. That fix has already gone to several manufacturers who should be able to supply an updated version of the BIOS for their specific motherboards soon.
Nvidia has said that it managed to get that BIOS fix to Shuttle the next working day after our initial article. Hats off to Nvidia for getting the problem sorted so quickly. µ

Answer 1

Thanks for the tip.  I had been clearing the CMOS to recover from a no-boot overclock, but this sounds better.
This topic is covered in the K7N2F Series manual (English) on page 2-32.  J10 is the jumper referred to in your post, and there is a "MSI reminds you..." note on that page to explain what to do in case of boot failure.
There is a second jumper on the board that is also mentioned on page 2-32.  This is J11 which has two options--open for 100MHz, and shorted for 133/166MHz.  Does it seem like J10 and J11 might do the same thing?  
Three observations:
1.  On my K7N2G-ILSR, with J10 and J11 in the default positions--User mode and 133/166MHZ--I was able to change the FSB on a 2500+ Barton (166MHz FSB spec) from 100 to 200, and get the CPU to run at the chosen settings.  Well, not quite--see #3 below.  So, J11--whatever it does--does not appear to limit the FSB settings.
2.  Another oddity is that, after clearing CMOS, the 2500+ always boots up at 100MHz FSB, even though its default is 166 MHz.  Same thing happens if the "Load BIOS Setup Defaults" is used.  Shouldn't both of these actions result in the processor starting up at its rated settings?
3.  (Related, but may be off topic.  I will start a new thread on this.)  The 2500+ Barton arrived unlocked.  So I set the FSB to 100 (for safety and ease of math) and tried every multiplier.  Here are the results (multiplier and CPU clock)--
07.0    700
07.5    750
08.0    800
08.5    850
09.0    1100
09.5    950
10.0    1000
10.5    1050
11.0    1100
11.5    1150
12.0    1200
12.5    1250
13.0    495
Observations;
A.  The 2500+ will start with a multiplier of 11 (correct) after clearing CMOS or selecting BIOS Defaults.
B.  What happened at 9.5 and at 13.0?  I thought this bios (V 1.2 February 14, 2003) was supposed to support the 3000+ Barton, which requires a 13 multiplier.  FWIW, the CPU clock on the 2500+ goes to 833MHz with 166FSB X13.0.  Is my Barton less than fully unlocked?  Or is there a bios bug?
C.  Multipliers above 13 would be helpful--just like FSB above 200.
--wcb