Please help identify component

[trebla]

While installing Coolermaster Hyper 6 cooler on EPoX 8KDA3J (AMD Athlon 64, Socket 754) motherboard I broke a component which interfered with cooler support frame.

This much I know about it.
Marking: WW1 / 46 (WW1 along the longer side and 46 rotated 90 degrees, most likely 46 is a lot number)
Package: SOT23-3
ID on board: D39 (also D6 is same)
Location: Between CPU and ATX 20-pin connector)
Connections: Lead 3 goes to ATX 20-pin connector lead 8 (PWR-OK, aka POWER GOOD), lead 2 sinks into any of the middle layers, lead 1 can be followed on the picture.

Measured resistance on leads of the another, non-broken one near the CLEAR CMOS jumpers (position D6), any combinations between leads gave readings of 600K or more up to infinity. Measurements done with Fluke 73 multimeter.
No lead connected with ground.

Have been searching on the net for a week but still no luck.
Based on electrical measurements cannot be a diode, transistors are marked with Q on schematics, colleagues have suggested Zeners, field transistors, JFETs, voltage supervisors.

WWx markings seem to be exceptionally rare, only real match with marking is http://www.smd.ru/files/upload/1245/ru/bat54.pdf ,see p2, marking section.
Near matches are http://www.marsport.demon.co.uk/smd/smdw.htm.

E-mail to EPoX support has given no reply.

Could anybody help to figure out the component, either original or if knowing exact purpose of it, suggest a replacement.
I'd rather like to save the board that cost me 100 Euros than throw it away.

Thanks in advance.

[ante]

Hi Trebla,

Most likely it is a diode (single or double) since the ID is D39!

[trebla]

the results of onboard diode function test for the known good one are the following:
(Fluke 79)

red-black
1-2->OL
1-3->286 ohm
2-3->beep

black-red
1-2->beep
1-3->OL
2-3->OL

[ante]

Trebla,

Does

[trebla]

Trebla,

Does

[trebla]

OK, here are the results of off-circuit diode function measurements for the known good one.
done with a cheapo Mastech M-830B tester

component leads 1 and 2 on the double lead side, 1 left and 2 right, 3 on the other, single lead side. (Philips style of marking)

pos-neg
1-2 -> OL
1-3 -> 200
2-3 -> 200

neg-pos
1-2 -> OL
1-3 -> OL
2-3 -> OL

these results to some extent confirm the component to be a Philips (manufactured in China) BAT54C

[ante]

Trebla,

Yes, I think it

[trebla]

OK, my PC is running again. :) Posting this message from it.
ordered a couple of BAT54C, changed the broken one and the one removed for measurement and the PC fired up like nothing had happened.