I would like to build a working unit for the Hackaday Prize 2018 but other projects have taken precedence... #YGREC8  has progressed a lot :-) But the Numitron display is almost done, there is not much left to do, before I get a functional pre-series module, so why not do it now ?

I stopped just before sending the boards to fab because an astute observer noticed that I forgot (again) to include the mounting holes. And I had already finished routing the board...

Another consideration is directly derived from 5. BLANK : the modules should be cascadable and the most-significant 0s should be blanked. This means that the circuit must have a blank input and blank output signal.

Blanking the digit is easy. Just insert a relay in series with the 0V common rail (normally open or normally closed, as you like). This could be done on a backplane where several digits are connected.

Cascading is different because the coils require so much current and the circuit has not been designed for this dual purpose. The cascade needs to do a OR (or AND) of 4 inputs (detect the input=0000) and it's already using many relays.

One approach is to extract the corresponding signal at the 0/8 node, at the leaf of the binary tree, but it also drives 6 segments (15mA each) so the overall current is already at the limit of the tree relay's handling capacity. I don't want to fry the contacts.

Should I just leave the proto as is and send it to fab so I can test the additional circuit during its fabrication ?

Oh and let's not forget @Fredrik Högberg  who started # Lets make OSHW more successful!  He'll handle the PCB fab part. Thanks !