Pass Transistor

A pass transistor is an enhancement mode transistor that is used in a certain way. Rather than being used as a pull-up or pull-down, it is used simply as a switch to control or steer a logic signal. It does not require Vdd or Vss as the pull up or pull down transistor would, but instead needs only a control connection, an input connection and an output connection. The control signal is connected to the gate and this signal controls whether the logic signal propagates from the input to the output. Note that a MOSFET used in this way can carry a current in either direction, so its not immediately obvious in what direction the signal would be propagating until you work out what each side is connected to.

The 6561 die shot shows a great many of this type of transistor. They are used a lot to control what happens within certain clock phases.

The following image shows a number of pass transistors:

The four contacts that are labelled are connected to short poly lines that cross a diffusion path. Two of these are labelled f1 and two are labelled f2. This indicates that one of those metal lines carries the phase 1 clock signal and the other metal line carries the phase 2 clock signal. These two signals are non-overlapping phases of the same two-phase clock. When phase 1 is high, the pass transistors connected to the phase 1 metal line will allow the logic signal on the diffusion path to pass through from one side to the other. When phase 2 is high, the pass transistors connected to the phase 2 metal line will allow the logic signal on the diffusion path to pass through from one side to the other.

The image below shows more examples of pass transistors:

In the lower middle of the image, we see four metal lines coming up from the bottom that each have a metal to polysilicon contact at the end of them. A polysilicon line extends from each of those four contacts and in each case they pass over four different diffusion areas. So we have a total of 16 pass transistors.

This second image shows an interesting example as it appears to show how the different colours are selected. The four metal lines going horizontally across the middle of the image and then turning downward are where the colour for the pixel is going to end up going. The four metal lines in the lower middle of the image that are connected to the pass transistors by way of the polysilicon lines are what controls which colour is selected, acting very much like a multiplexor. Presumably then only one of those four metal lines will be high at any one time. The purpose of the pass transistors in this example are to let a signal through to the metal to diffusion contacts that are labelled. In the case of the CA0-CA3 contacts, these actually receive both the auxiliary colour and the current character's foreground colour. The two polysilicon paths either side of the CA0-CA3 contacts control which of those two colours pass through to the contacts. The situation for the CE and CB contacts are a bit simpler as they only have one possible input to them.

Note that in the case of two of the colours (border and character foreground colour), the high bit is hard wired for logic 0. So this still needs to be controlled by a pass transistor though as it only applies to two of the colours.