On a Gottlieb system 1 CPU board, the 10788 (top right spider chip) is easy to diagnose. The section for Xi strobes, plus DBS, is relatively autonomous. The generated patterns are easy to track. For example, DBS is a square signal running at about 130Hz. You should observe it on pin 35 of the device, also available on pin 9 of buffer Z14 (Z14 is the third TTL device counting from the top right of the board). If you can observe the 9 signals in the right frequencies and phases, the 10788 is likely OK. If you still have display problems, the cause of your worries comes probably from the logic glue around the 10788. 7448s are often defectuous and hard to find. The 7421 is also frequently dead and even harder to find. The problem with the 7421 is that it controls the 7448 enable pins. So a defectuous 7421 might disable 1 or 2 7448s. And the consequence is that everything is turned off in such a situation, or you have 1's displayed in many places.
Purple : DBS (130Hz)
Yellow : X1 (258Hz)
Blue : X7 (258Hz)
Before incriminating a 10788 in a display failure, you need to make sure that the PPS-4 system is OK. data/address/control must be OK (no short or open on these). The Processor+both A17s must be running as well. If at least 1 typical pattern of the switch matrix can be observe, then you can be 99.9% sure that the PPS-4 system is working. Then, if you are in this situation, and if DBS signal is absent, you can be sure that the 10788 is dead.
Before deducing anything, of course you will have checked that the reset SPO is deasserted and that clkA and clkB are present and correctly synchronized (on the 10788 itself).
Another classical problem with the display you may experience is the dip switch #13, which inhibit the display of credits. So, if dip switch #13 is not set correctly, or defectuous, you may wonder why there is nothing on the credit window... We have had the problem during our first clone test. We stayed puzzled for 4 days because of this!
The keyboard section is not used on the Gottlieb system. We have implemented something but we cannot guarantee that what we've implemented is identical to what is actually done in the original 10788s.
But it does not matter, since it is don't care on the system 1 (all Yi are tied to 5V).