Blockbuster
 
Xav de-mystifies a block diagram for the Team Tap...
 

Before we delve deep into the guts of the Team Tap, I thought it might be useful to look at it in block form.

The most important block of the team tap, and the bit which does much of the hard work, is the chip we covered last time so dig out the last issue and read it again.

Fig 1: Block Diagram of a Team Tap

To summarise; one output is taken low for each combination of inputs.
This gives sixteen possible output states, which are arranged such that four EJPs can be independently driven.
If things were arranged in the most obvious way (Q0, Q1, Q2 ... Q15) then the Tap would have to be removed in order to play
older games.
By using an "odd" arrangement, however, it is possible to use the same mask for a single joypad plugged directly into the
EJP as for a joypad plugged into the first socket on the Tap. In other words, the Tap becomes transparent to the software.

However, none of this would work correctly without the "diode decoding network"
For our purposes, the diode can be considered to be a one-way valve for electricity.
Quite simply, by sticking a couple of pieces of chemically altered silicon together, the current can be made to flow one way but not the other.

It's not always clear to beginners in electronics why you would want to stick dozens of diodes into circuits, in very precise positions
and orientations, but a brief look at the way the current flows will soon make things clearer.

Fig 2: A Simple Switch Matrix

In figure 2 we have a simple switch matrix - not dissimilar to that used in the joypad.
The idea is that voltages are applied to the input lines, and when the switches are pressed, these voltages are transferred to the outputs.
Therefore, if 5V is applied to I0, and switch A is pressed, 5V will appear at Q0. Similarly, if switch B is pressed, 5V will appear at Q1.
By applying a voltage to only one input at a time, it is possible to work out which particular button in the whole matrix is being pressed.

Now imagine that our elementary switch matrix represents the joypad direction and fire buttons.
What happens when you want to perform some ultra-death move in a fighting game, which requires you to hold down A, B and E simultaneously?
The obvious answer would be that Q0 and Q1 are at 5V when I0 is at 5V, but when I1 is at 5V only Q1 will be taken to that level.

Unfortunately this ignores one vital point: what are the input voltages when they're not being held at 5V? In other words, when I0 is at 5V,
what is the voltage of I1? Depending on the drive electronics, it could be at any level, of course - but the most likely case is they will be at 0V.

Now let's trace the path of the current. It comes sauntering along I1, and finds a nice exit route through switch A, pulling Q0
towards 5V as well. But not all of it takes that route - some carries on along I0, where it finds an equally nice path through switch B. Before it reaches the end of Q1, however, it gets distracted by an even nicer route through switch E, which will allow it to get straight to ground without having to meander through the electronics attached to Q1. This is the infamous "short circuit" - basically a way for the current to reach ground more easily. So now, instead of being pulled up towards 5V, Q1 is actually pulled down towards 0V, and it becomes impossible to tell precisely which buttons are being pressed..

Fig 3: A Switch Matrix with Diode Decoding

Now let's add some diodes to the circuit (figure 3). This time our current trundles happily along I0, and some of it
travels down to Q0 through switch A as before. The diode lets it though, because it happens to be travelling in the right
direction. Had it been coming the other way, the diode would have blocked it.
This is exactly what happens when the current passes through switch B and then tries to double back via switch E.
This time the diode stops it, so the easier route is the one through the electronics on Q1.
Consequently the current continues on its merry way, pulling Q1 towards 5V as it does so.

Naturally the actual Team Tap slightly complicates this scenario, as all the voltages are reversed (remember that negative logic stuff),
but the general principle remains the same. Also, the number of diodes in a Team Tap (or joypad) puts our simple circuit to shame,
but essentially they perform the exact same task.

Next time we'll look at the full circuit diagram for the Team Tap. In the meantime if you have any questions or comments
about this series, get in touch and I'll do my best to help.
 

Email: xav@compsoc.man.ac.uk
 
 
 

[Home] [Info] [Back] [Next]