Is the Computox a Haufe Machine?

We've already established that the Computox isn't a clone of the machine described in the Haufe patent.  The Computox has an additional uniselector and nine more relays than Haufe's coin-op version.  This could be explained by the fact that Haufe's machine only supports single player games, but it's more probable that the Computox has a different design.

Nonetheless, let's start with the hypothesis that the Computox is a derivative or extension of Haufe's work and see if we can refute the proposition using the small amount of reverse engineering we have done thus far.  If we can't, maybe we can use the Haufe patent to guide our reverse engineering efforts.

So far we have focused on understanding the connections of the switch inputs and lamp outputs.  These aspects of the Haufe machine are shown in figure 9b of the patent.

Haufe machine switch input and lamp output

The lamps that illuminate on the game display are found in a column at the left of the diagram, with the machine or O lamps shown above the player or X lamps.  In both cases two lamps, labelled 62 and 27 or 63 and 28, are shown for each position.  Haufe suggests that in addition to the game display being illuminated by lamps 62 or 63, the controller buttons could also be lit by lamps 27 or 28.

To the right of the lamps, the switches the player uses to make a move, equivalent to the Computox's X controller are shown in the second column of the diagram.  These are 24A through 24J.  So far, so good.

Let's consider what will happen if we close one of the player switches, say 24E.  The left side of 24E is connected to switch 99 at the top of the diagram.  It's not clear what this does, but for the moment, let's assume that this is at logic 1.  This would mean that when we close 24E the logic 1 is transmitted to the break switch ME-1, connected to the right hand contact.  Interesting.  ME-1 is controlled by the ME or Machine E relay.  From there the signal goes to the PE-1 (Player E relay switch 1) break switch, hmm this could be a problem, to relay 120 and to ground.

With relay 120 active, switch 122 opens, switch 123 moves from position 124 to 125 and switch 126 closes.  Switches 122 and 123 create a chain bringing a logic 1 down through the positions to the bottom of circuit 'a'.  This logic 1 is carried through conductor 135 and 136 back to switch 126 which as we said, is now closed.  From there it is carried to relay PE, which activates.  This breaks PE-1 disconnecting the input switch.  However, it also closes PE-2 which latches both relay 120 and PE in their active state, despite the fact that the input from switch E has been disconnected.

The other point to note is that when relay PE activates, switch PE-3 will close.  This illuminates both the 63 and 28 lamps.

So, the key question.  Does this input logic map onto the Computox?

Well, the X switch inputs on the Computox initially do connect to a break switch on a machine relay, for example EOA.  However, it then goes to a make switch on EXA, rather than a break switch.  There are only two break switches on the *XA relays (top and bottom bank switch 4).  The switches on the top bank are connected to the O controller inputs.  So that just leaves the bottom bank switch 4, and these don't appear to be connected to the break switch on EOA.  Additionally, relay EXA top bank switch 2 pin 1, which is connected to the EOA input switch, also appears to be directly connected to the AXA relay coil, meaning this relay is directly connected to the switch input without the complexity of a second relay seen in the Haufe design.  Therefore, the Computox input is more like the diagram below (yes I am trying to find something that can draw these electrical diagrams better than Eagle, it seems Dia isn't that tool).

Computox X input logic

It's not clear from the work done so far how the EXA relay is latched once it is activated.  However, I think it is clear that given the radically simpler input handling, it's unlikely that Computox is a close relative of the Haufe machine.