i was messing with my cpu fan last night (i think i may have plugged it in wrong partially) and suddenly i saw two tiny puffs of smoke directly above the header... and immediately my fan went up to full speed! i tried speedfan to slow it back down which didn't work, but moving to the sys fan header worked fine since they were both set at 50%, at this point i still had rpm readings from the sys header.
i then tried resetting but then i lost all rpm signals! thankfully the sys fan is still controllable but the cpu fan is still full speed only.

now to be honest i can live with one fan and no rpm read outs, but nevertheless it's a funny situation.

anyone know what happened? or has anyone experienced anything similar before?

How exactly were you 'messing' with it? Is it possible you shorted something out when plugging/unplugging a connector? This would blow a header most likely. As this is a PWM header you have probably blown the PWM signal so it just runs as a 12V DC fan now.

Try finding the RPM signals with different software. If you have an Asus motherboard try Asus PC Probe for example, I have previously found RPM signals with it that don't show up with other tools.

yeah just plugging/unplugging a fan... i think you may be right. it turns out i didn't lose rpm readings, just that the fan lost them, which lines up with the theory of blowing the pwm part of the header by inserting the fan incorrectly.
i'm just using the sys fan header now and i had to switch the cooler/fan anyway so all is well.

it's a little disconcerting though since plugging in a fan wrong seems like a pretty easy mistake to make.

Ah, but all of the manuals will tell you to do it with the system switched off and then will have something like 'verify that all connectors are correctly seated before powering on'. That way you don't damage things. But we have to mess with things and I've done similar things before. I have blown a fan before too but it was an oddly wired one and it didn't like 12V being sent down the RPM signal cable.

haha... fair enough.

i guess i could be viewed as lucky since i've always worked on my pcs while they're on, lesson learned though and i'll be much more careful in the future.

It's most likely you fried the control circuits in the chip that runs the header. The header itself, along with the connectors, cables and traces are very robust, although a direct short could evaporate a trace.

I had a similar failure some time ago when I was installing a home-made splitter cable. I had been too lazy to use proper shrink tubing around the splices, and paid the price when a couple of the wires shorted. The good news about an event like this is that the ground and 12V pins are almost impossible to damage, so the header still functions as a full-speed, unregulated 3-pin header.