12 Dec 2005, 19:37
In engineering terminology, the seven segments of an LED display are lettered "A" through "G", "A" being the topmost, and progressing clockwise, so the upper left vertical segment is "F" [with "G" being the central horizontal one]. So your problem is with "F". The associated segment drivers are the row of seven tiny rectangular blocks directly above the display [the two on the far left, and displaced slightly lower, are digit drivers]. From right to left, they provide power to the A, B, D, C, E, F, and G segments, respectively [don't know why they reversed the positions of the C and D drivers, but they did; might simply have been the most straightforeward way to lay out the circuit]. Thus, the sixth one from the right is the troublemaker. Gentle pressure might re-establish contact . It is unlikely the defect is in the display, since all three F segments are out. A quick way to check the display contact is to short out the transistor [with a fine tweezer, touch the bottom and upper right contacts; if the segment returns, there is certainly no problem with the display itself...it also tells you there is power available on that line]. Removel of the bad transistor is purely mechanical; gripping top and bottom with a pair of heavy, blunt-nosed tweezers, just give it a slight twist. It should break free without applying any great amount of force [especially if one contact is already broken]. A replacement transistor can be extracted from a donor module in the same way. Conductive epoxy is by far the best method of re-attachment [any heating of the substrate may loosen adjacent components, and so do more harm than good]. Note: Drivers can only be replaced with ones of like type; you cannot put a segment driver in a digit position, and vice-versa. I presume one set of drivers is PNP, and the other NPN, and are thus not interchangeable in the circuit. Also, this repair, while not very complex, should not be attempted without some form of visual aid; it is too easy otherwise to get conductive epoxy where it does not belong, and so create new problems. I do these at 30x under a binocular microscope; with care, the fix is almost invisible.