Computox Input Latching

Given the initial look at the Haufe machine, it makes sense to try to understand how the Computox handles the X controller inputs a bit further.  So far we established that the controller input passes through a break switch on relay *OA, presumably an interlock to prevent the X player choosing a square already occupied by O, and then on to a switch on relay *XA, thought to only allow the player to make a move when it's their turn.  However, it wasn't clear how the fact that the X player had made a move was retained when the controller button was released.

With a bit more buzzing I think we have the answer to this, and it helps to understand the role of a few more switches (for those that are interested I've gone back to my ancient copy of Eagle, and started to build a specific Computox library of symbols necessary for these circuit diagrams).

Drof Computox X controller A position input latch

In this circuit fragment it becomes clear that the top 2 switches (T2) on the *XA relays are the latching mechanism.  If the relevant position isn't occupied by O, once the controller input is brought low, the *XA relay will be activated, this closes switch *XA T2 which then provides an alternate path to ground for the coil connection.  This keeps the *XA relay active, even when the controller button is released.  The second pin for all *XA T2 switches are connected together and fed through the normally closed RESET button to ground.  Once activated, the *XA relay state will persist until the RESET button is pressed, removing the path to ground and releasing all *XA and *OA relays.

Therefore, the  *XA and *OA relays act as registers, used to hold whether each position is occupied by the X or O player respectively.  The T4 break switches are used as an interlock, preventing both players occupying a square.  The T2 make switches are used to latch a square as being occupied during a game, and the T1 switches turn on the relevant lamp in the game display unit.  This mechanism is much simpler than that employed by the Haufe machine, without the cascading of inputs through subsequent position switches.  In the Haufe design I believe that this complexity might be an interlock, designed to prevent more than one input being accepted on each turn.  This would make sense, as the design in the Haufe patent was intended to be used commercially, and would be expected to be the subject of cheating attempts.

So is the X controller's common connection just connected to ground?  Well no, adding this information to the circuit gives the following schematic:

Drof Computox X controller A position input and Play lamp control

As we can see here, the X controller common runs through the changeover switch MDX T2 (this is wired as a break switch with pin 3 left not connected), through the MPX relay coil, from there through relay V switch B2 (which is also wired as a break switch) and then connects to the RESET button and ground.

So, in order for controller X to be active, relays MDX and V must be inactive.  If this is true, the common for the controller will be connected to ground and button presses will be registered.  The fact that relay V must be inactive is consistent with the finding that, for the X controller "Play" light to be illuminated, relay V must also be inactive (shown at the bottom of the diagram).  We now have two pieces of evidence suggesting that relay V controls which player's turn it is.  So the interlock that prevents player X making moves when it isn't their turn is likely to be V B2, and not *XA T2 as speculated in the previous post on inputs.

It's also interesting that when an X controller button is pressed on the player's turn, relay MPX will be activated as well as *XA, the purpose of this is not yet known.

So we now understand the purpose of 21 of the 47 relays, and 58 of the 257 switches.  Whilst it's slow going, progress is being made.