ROM and BJT memory cell storing 1 or 0

ROM (Read Only Memory) is a non-volatile semiconductor memory in which data is written once (during manufacture or by programming) and afterwards can only be read. Data stays even when power is removed. It is used to store fixed programs (BIOS, firmware), look-up tables and code converters. Types: mask ROM, PROM, EPROM, EEPROM.

Storing 1 or 0 in a BJT ROM cell

A bipolar ROM is an array of row (word) lines and column (bit) lines. At each crossing there is a cell position where a BJT may or may not be connected.

• The base of the transistor is connected to the row line.
• The collector is connected to $V_{CC}$.
• The emitter is connected to the column line only if a 1 is to be stored. Each column line has a resistor to ground.

 Vcc        Row line (from decoder)
  |          |
  C          |
   \|--------+       (stored '1')
   /| B
  E
  |
==+===== Column line --> sense --> Data = 1
         |
        [R]
         |
        GND

 Stored '0': emitter link absent (or fuse blown),
 so column stays at 0 V through R.

Reading:

1. The address decoder makes the selected row line HIGH.
2. In a cell storing 1, the transistor's base is high, so it conducts (emitter follower). Current flows through the column resistor and the column line goes HIGH, which is read as 1.
3. In a cell storing 0, there is no emitter connection, so no current flows into that column; the resistor keeps it at 0 V, read as 0.
4. Unselected rows are LOW, so their transistors are off and do not affect the columns.

In a mask ROM the connection is made or left out by the metal mask. In a bipolar PROM every cell is made with a fusible link in the emitter; the user stores a 0 by blowing the fuse with a high current pulse, and the intact fuse keeps a 1. Since the stored pattern is fixed by a physical connection, the data is non-volatile.