HP150 Touchscreen again

A couple of months ago I asked about diagnosing problems with the TOuchscreen board in an HP150. Well, I took some time off that project [1] but finally got back to it. I decided in the end to make a text unit that would plug into the touchscreen PCB, produce the clock signal, accept the sync and data signals and display the status of the 35 beams on LEDs. [1] Sort-of classic computer related. I've been writing a set of articles for HPCC on how to fix HP9800 machines. Being me, I built it from TTL (actually HC and HCT parts). It only took me a couple of afternoons do design and build it. It starts with a 4MHz master clock, divided down with a couple of '393 counters. A '30 adn '138 produce a paair of spaced clcok pulses from this, a '02 combines one ofthes with the sync signal from the tocuhscreen. These are all latched in a '175 to procude 3 clocks -- the clock to the touchscreen PCB, a clock to sample the data from the touchscreen PCB and an end-of-scan pulse to latch the received data and send it to the LEDs. The data is shifted into '4094 shift registers/latches whihc feed a couple of rows of LEDs. The powrr supply stats as 12V from my bench supply. That feedsthe +ve input of the touchscreen. A 7805 powes the logic and LEds inthe test box. And a 7660A provides the -12V supply for the tocuhscreen PCB. Anyway, after connecting it to the defective tocuhscreen PCB, I found that one of the beams appeared to be blocked all the time. Blocking other beams got the appropriate response from the test box, so I was pretty sure the logic was all working properly. Since I knew which beam was malfunctioning, I tested its IR LED in-circuit with an ohmmeter (system powered down, of course, It read differnetly of circuit. It's open. So the guy who told me that the emitter (LED) was the most likely failure was right. Now all I need to do is find a replacement. -tony