As above, the jumper won't damage anything, it is there to be used.  This is a bit of an odd symptom.  Did it start before or after you removed the heatsink?  The reason I ask is, what you describe can happen if the mechanical joint between CPU and sink is corrupted.  Heatsink reinstallation is a bit of an art.  All the residue from the original compound should be removed, and a moderate amount of new compound applied.

I doubt you're looking at static damage, as that wouldn't produce variable or transitory symptoms.  You can pretty much rule out BIOS corruption, as you've replaced that without changing anything, and it's astronomically unlikely the flashrom has a defect that affects only the fan.

Redo the heatsink installation and see if it clears up.

x_lab rat wrote

---

" it's astronomically unlikely the flashrom has a defect that affects only the fan."

Have you ever worked with computer other than your casual use of AOL mail? I've seen it a million times where some hotshot tech will only replace the fan and heatsink for an issue like this where the problem truly lies within the systemboard. The only way to find out is to put it on a bench and swap parts around , but you should never rule out what your experience considers an unlikely resolution.

---

"Have you ever worked with computer other than your casual use of AOL"

Beg pardon?  I was a C2 engineer at Dell for 2 years, entire systems, desktop and portable.  What I said was specific to the condition of BIOS fan-control profile, not the rest of the chipset, which still may require replacement.  But considering it is an expensive solution, one rules out the simpler ones first.  This is why fan and heatsink reinstallation is always dispatched first, though that may not ultimately turn out to be where the problem resides.

While it is possible for a flashrom to have a single-address, row or column error, in two BIOS revisions the exact physical address of the fan control profile can change, which would put a different parameter in the errored address and cause something different to fail.  Row and column errors would also very likely span more code than just the fan control.  It is also entirely possible for two BIOS revisions to have the same code error in them, but the enduser has no way to manipulate that other than flashing to another revision, which was already tried.  It is further possible for the CPU thermal diode to be defective (apx 2 per M), but that is the most expensive possible solution.

Forum protocol is to add to other contributions and expand upon understanding of issues.

Well, here's the latest:

First, I reset the heatsink. That didn't make any difference, so I removed the CMOS battery overnight.

Next day, I replaced it, rebooted to the BIOS, and restored the default settings.

I even uninstalled and removed the USB card, since this all started immediately after installing it. Even though I didn't think it was related, I thought I'd at least be thorough. Couldn't hurt to rule it out.

Still, the fan is running fast. It exhibits exactly the same behavior at each boot-up, as described above. And it also does the same thing while in the BIOS setup, or even booting from a boot CD.

Even though I don't use an ESD strap, I'm pretty careful about discharging any static electricity by touching the case before handling any components. I know it's not foolproof, so the possibility of static damage can't be ruled out.

I tend to think that booting the computer with the CMOS jumper improperly mounted is the cause of my problem.

Since the fan passed diagnostic testing, the fan controller would seem to be the culprit. My guess is that it's not replaceable and I need to replace the motherboard.

I have, in the meantime, discovered a way to decrease the fan noise.

Ear plugs.

If you're conscientious about having a hand or forearm on the chassis when handling unmounted parts, the odds of static go way down.  It performs the same function as the strap as long as you're consistent with it.

Jumper misuse does not damage anything.

"Since the fan passed diagnostic testing, the fan controller would seem to be the culprit."

You have that backwards.  If the fan diagnostic passes, the fan controller hardware IS working, from the chipset to the fan anyway.  The only things left out of that test loop are the heatsink/CPU junction and the CPU itself.  The sensor-diode fault was characteristically evident from day-one, but anything can happen.

Right, if a change (the card) causes misoperation, undo the change.  Everything in a computer is related to everything else.

Since fan control does work (the diagnostic), the only remaining common fault is heatsink mount integrity.  You redid it, but you could have made the same mistake both times.  You know, like proofreading your own stuff, if the mistake looked right the first time it will most likely look right the second time too.

Absent further evidence, I don't have any other ideas to offer, sorry.

---

 

 This is why fan and heatsink reinstallation is always dispatched first, though that may not ultimately turn out to be where the problem resides.

 

---

You read what you wanted to read.  I said 'flashrom' and nothing else.  You said 'main board'.  Not that the resolution of either argument addresses the issue, the user can't replace the flashrom alone, and several thousand other parts besides flashrom reside on the main board and can cause anything to happen, or not happen.

But we're supposed to be addressing the problem, not each other.