Malfunctioning VT240 - help please

OK. the keyboard is working properly as far as I can tell, data is going in and out, and I even swapped it for the keyboard on my VT220 and the same symptoms persisted. I just verified all four ROMs on the T11, and the ROM for the 8085, against the images I found on the MAME site. So far so good. One interesting finding - two of the lines (DAL3 and DAL1) on the T11 do change states several times, but once the self-test has crashed, they stay high with almost one volt of "wiggle". All the other data/address lines are either high, low or switching between a good 1 and 0. There are several places that the bus connects, including the ROMs, 1-bit dynamic RAMs and various octal latches & bidirectional buffers. I connected a 10 ma VOM between each line and ground (to make sure a low-resistance path (such as in the 'LS245 at E55) wasn't forcing it high somehow. All of the DAL15-0 lines requires more than 1.9 ma to bring it to ground (well, 50 mv burden at 250 mv full scale, anyway). That leaves the unlikely possibility that one of the octal TTL devices, or ROMs. has developed a weird internal pathway that only interferes with DAL3 & 1 on some bit patterns, but not all the time. Seems like a zebra rather than a horse. The only part that drives multiple low-order DAL lines at once besides the E19-22 ROMs is the E55 LS245. The T11 spec sheet says that a good logic 0 (<0.4 volt) should be possible with up to 3.2 ma sink... So I suspect the T11 has a couple of bad output pull-down transistors on those lines. Anyone got a spare T11 chip I can buy or borrow? Or send you mine to plug into your board and see if it fails the same way? :) thanks.