Ch3
- 1. Copyright © 2008 by Dr.K.K.Thyagharajan
1
3 . I n t e r f a c i ng M e mo r y w i t h 8 0 8 5
A microprocessor will have its own address space for accessing physical memory.
The memory locations that are directly addressed by the microprocessor is called physical
memory space. For example a microprocessor like 8085 has 16 address lines, and it can
access a physical memory space of 64K starting from 0000H to FFFFH as shown in
Table 3.1. The process of interfacing memories to microprocessor and allocating address
to each memory location is called memory mapping. The complete address space may be
considered as a single memory block. But practically, when ICs are used as memory
devices, instead of a single IC (Integrated Chip), few devices that fit into the address
space will be used. This is due to the fact a microprocessor based system requires at least
one ROM/EPROM and a RAM. For example instead of using a memory device of size
64KB (Kilo Bytes), we can use 8 memory devices with a capacity of 8KB each. This will
reduce the chip replacement cost while servicing the microprocessor-based system. When
memory blocks are used, to access all the locations in a block n address lines are needed
so that
2n = 8 K
= 23 K
= 23 x 210
= 213
Therefore n = 13
3.1
Address Decoding and Memory Mapping
Memory address decoding is nothing but to assign an address for each location in
the memory chip. The data stored in the memory is accessed by specifying its address.
Memory address can be decoded in two ways
i)
Absolute or Fully decoding
and
ii)
Linear Select or Partial decoding
There are many advantages in absolute address decoding. They are
i)
Each memory location has only one address, there is no duplication in the
address
ii)
Memory can be placed contiguously in the address space of the
microprocessor
iii)
Future expansion can be made easily without disturbing the existing
circuitry
There are few disadvantages in this method
i)
Extra decoders are necessary
ii)
Some delay will be produced by these extra decoders.
- 2. Copyright © 2008 by Dr.K.K.Thyagharajan
2
The main advantage of linear select decoding is its simplified decoding circuit.
This reduces the hardware design cost. But there are many disadvantages in this
decoding.
i)
Multiple addresses are provided for the same location
ii)
Complete memory space of the microprocessor is not efficiently used
iii)
Adding or interfacing ICs with already existing circuitry is difficult.
3.1.1
Absolute Address Decoding
The 8085 microprocessor has 16 address lines. Therefore it can access 216
locations in the physical memory. If all these lines are connected to a single memory
device, it will decode these 16 address lines internally and produces 216 different
addresses from 0000H to FFFFH so that each location in the memory will have a unique
address as shown in table 3.1 and figure 3.1. This is called absolute address decoding.
Figure 3.1 is called the memory map of 8085 address space and it shows how each
location in the physical memory gets an address from the address space of the
microprocessor.
A15 A14 A13 A12
0 0 0
0
A11 A10 A9 A8
0 0 0 0
A7 A6 A5 A4
0 0 0 0
A3 A2 A1 A0
0 0 0 0
Hex Address
0000H
0
0
0
0
0
0
0
0
0 0
0
0
0
0
0
1
0001H
0
0
0
0
0
0
0
0
0 0
0
0
0
0
1
0
0002H
-
-
-
-
-
-
-
-
-
-
-
-
- - - - -
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
FFFEH
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
FFFFH
Table 3.1: Memory Address
- 3. Copyright © 2008 by Dr.K.K.Thyagharajan
3
FFFF H
FFFE H
FFFD H
8001 H
8000 H
7FFF H
0002 H
0001 H
0000 H
Figure 3.1: 64KB Memory Map
(Single Block)
Instead of connecting all the 16 address lines to a single 64 KB memory device,
we can also use memory devices of lower capacity. In such cases the physical memory
space of the microprocessor is divided into smaller memory spaces or memory blocks.
Each block is selected by a block select address signal and the memory locations with in
a block are accessed by the processor’s address lines. This is called address portioning.
For example if the memory device (memory block) has only 13 address lines, we can use
the lower 13 address lines (A0 – A12) to access the locations with in a memory block and
the remaining 3 lines (A13, A14 and A15) to access 8 such blocks. These three lines are
called the block select address signals and this is called address partitioning.
Depending on the number of memory blocks, the number of block select address lines
will change. The starting address and ending address of each block can be found as
shown in Table 3.2. Since each block is an Integrated Chip (IC), and each chip has a
Chip Select signal, the block select address lines must be used to select the ICs. We have
to produce eight chip select signals from the three address lines. Therefore we have to
decode these lines using a 3 to 8 decoder or 1 of 8 decoder. In this case the lower thirteen
address lines of the processor are connected to the 13 address lines of the memory chip
and hence they are internally decoded in the memory. The higher 3 address lines (A13,
A14 and A15) are externally decoded by a 3 to 8 decoder. Since three lines can provide a
maximum of 8 addresses, in this case they are said to be fully decoded and hence each
location in each block has a specific unique address as shown in table 3.2. This is called
as absolute address decoding. The diagrammatic representation of each block with its
address range shown in figure 3.2 is called the memory map of the microprocessor
system.
- 4. Copyright © 2008 by Dr.K.K.Thyagharajan
4
Hex
Address
range
0000H to
Block
No
A15A14A13 A12
A11 A10 A9A8
A7 A6 A5 A4
A3 A2 A1 A0
1
0 0
0
0
0
0
0 0
0
0
0
0
0
2
0 0
0 0
0
1
1
0
1
0
1
0
1 1
0 0
1
0
1
0
1
0
1
0
1 1
0 0
1 1 1FFFH
0 0 2000H to
3
0
0
0
1
1
0
1
0
1
0
1
0
1 1
0 0
1
0
1
0
1
0
1
0
1 1
0 0
1 1 3FFFH
0 0 4000H to
4
0
0
1
1
0
1
1
0
1
0
1
0
1 1
0 0
1
0
1
0
1
0
1
0
1 1
0 0
1 1 5FFFH
0 0 6000H to
5
0
1
1
0
1
0
1
0
1
0
1
0
1 1
0 0
1
0
1
0
1
0
1
0
1
0
1
0
1 1
0 0
7FFFH
8000H to
6
1
1
0
0
0
1
1
0
1
0
1
0
1 1
0 0
1
0
1
0
1 1
0 0
1
0
1
0
1 1
0 0
9FFFH
A000H to
7
1
1
0
1
1
0
1
0
1
0
1
0
1 1
0 0
1
0
1
0
1 1
0 0
1
0
1
0
1 1
0 0
BFFFH
C000H to
8
1
1
1
1
0
1
1
0
1
0
1
0
1 1
0 0
1
0
1
0
1 1
0 0
1
0
1
0
1 1
0 0
DFFFH
E000H to
1
1
1
1
1
1
1 1
1
1
1 1
1
1
1 1
FFFFH
0
Table 3.2: Address Space for Memory Blocks
0 0
- 5. Copyright © 2008 by Dr.K.K.Thyagharajan
5
FFFF Memory
RAM 7
Block8(8KB)
E000
DFFF
Memory
RAM 6
Block7(8KB)
C000
BFFF Memory
RAM 5
A000 Block6(8KB)
9FFF Memory
RAM 4
8000 Block5(8KB)
7FFF Memory
RAM 3
6000 Block4(8KB)
5FFF Memory
RAM 2
4000 Block3(8KB)
3FFF Memory
RAM 1
Block2(8KB)
2000
1FFF Memory
EPROM
Block1(8KB)
0000
Figure 3.2: 64KB Memory Map
(Eight Blocks)
We use 74LS138 address decoder to generate the chip select signals for each
memory block. In this decoder when the address lines A13, A14 and A15 are 000, the
output line Y0 will be activated as shown in figure 3.3. This in turn selects the first
memory block. Similarly when these lines are 001 (C=0, B=0 and A=1) Y1 will be
activated and the second memory block will be selected. When the 8085 microprocessor
is reset the contents of the program counter will be 0000H. Since this content is placed on
the address bus while accessing the memory, the microprocessor will start executing the
program from the address 0000H. A monitor program must be placed starting from this
address, so that it reads the inputs and takes action based on that program. The input may
even be a function key that requires a specific action to be carried over. Since this
program is fixed and developed by the manufactures of the product it is called as
firmware and it is stored in a ROM or in an EPROM. Therefore, in any microprocessor
based system an EPROM should be placed at the reset address (0000H). In addition to
EPROM, any microprocessor based system would require at least few RAM locations to
store temporary data. Specifically there should be a RAM for stack operations. So,
though we are using eight memory chips of 8KB size each, at least one of the eight ICs
must be an EPROM and it should be placed at the starting address of 0000H. The
memory map for this case can be given as shown in figure 3.2. The EPROM chip must be
selected using the chip select signal CS1 and the RAM1 chip must be selected using CS2.
The memory interface diagram may be given as shown in Figure 3.4.
- 6. Copyright © 2008 by Dr.K.K.Thyagharajan
6
A13
A
Y0
CS1
A14
B
Y1
CS2
A15
Block
Select
Address
Signals
C
Y2
CS3
74LS138 Y3
CS4
Y4
CS5
G1
Y5
CS6
G2
Y6
CS7
G3
Y7
CS8
+5V
IO/M
}
Block
Select
Signals
Figure 3.3: Memory Block Decoder
In this type of memory interfacing, all the address lines (A0 to A15) have been
used. Each location in the memory will have a single address. This type of address
decoding is called as absolute or fully decoded addressing.
Most of the microprocessor based systems do not use the complete 64 KB
memory space. Even one EPROM and a RAM will be sufficient. For example in the
memory map shown in figure 3.2, if only the EPROM and RAM 2 are used in the
practical system, the memory map of such a system can be given as shown in figure 3.5.
Still each location has single address. Therefore it is also called absolute address
decoding. The main advantage of this type of decoding is, you can add memory devices
for this system without disturbing the already connected devices. The memory interface
diagram for this case is given in figure 3.6.
- 8. Copyright © 2008 by Dr.K.K.Thyagharajan
8
FFFF
E000
DFFF
Unused memory space
C000
BFFF
A000
9FFF
8000
7FFF
6000
5FFF Memory
RAM 2
4000 Block3(8KB)
3FFF
2000
1FFF Memory
EPROM
Block1(8KB)
0000
Figure 3.5: Memory Map with EPROM and RAM 2
A13
CS1
Y0
A
A14
B
Y1
A15
C
CS2
Y2
+5V
74LS138
Y3
CS3
CS4
Y4
G1
Y5
CS6
G2
Y6
CS7
G3
IO/M
CS5
Y7
CS8
CS
RAM 2
CS
EPROM
A0 - A12 Address Bus
Figure 3.6: Interfacing EPROM and RAM
Instead of using 3 to 8 decoder, NAND/AND/OR gates can also be used to
generate chip select signals as shown in figure 3.7. In this case, CS1 will be activated to
select the EPROM, when all the three inputs to the OR gate (OR1) are zero. Similarly the
chip select signal CS3 will be activated to select RAM 2 only when A13 = 0, A14 = 1 and
A15 = 0. This is also an absolute address decoding method. But, here we need more
- 9. Copyright © 2008 by Dr.K.K.Thyagharajan
9
decoders (gates) when we want to add more memory chips. The main advantage of the
absolute address decoding is the contiguous placement of the memory locations. So no
memory space of the microprocessor is wasted without being used.
A13
OR 2
OR 1
A13
A14
CS1
A14
CS3
A15
A15
CS
CS
RAM 2
8K
EPROM
8K
A0 - A12 Address Bus
Figure 3.7: Address Decoding using OR Gates
3.1.2
Linear Select Address Decoding
In the circuit given in figure 3.8, the address line A15 of the microprocessor has
not been used for decoding the chip select signals. The outputs Y0 and Y2 of the 74LS138
decoder will be activated as given in table 3.3. Since the address line A15 is not connected
to the decoder and the pin C of the decoder has been grounded, irrespective of the signal
A15 (i.e.A15 is 0 or 1), the pin C is 0. So, you will get Y0 selected twice i.e. when A15 = 0
and when A15 = 1. Similarly Y2 will also be selected twice. Therefore the address map for
the EPROM and RAM is found as shown in Table 3.4.
A15 C
0
0
1
1
0
0
0
0
A14
B
0
1
0
1
A13
A
0
0
0
0
Y0
Y2
0
1
0
1
1
0
1
0
Table 3.3: Linear Select Address Decoding – Chip Select Signals
- 10. Copyright © 2008 by Dr.K.K.Thyagharajan
A13
A
B
Y1
C
CS1
Y0
A14
Y2
+5V
74LS138
Y3
CS2
CS3
CS4
Y4
CS5
G1
Y5
CS6
G2
Y6
CS7
G3
IO/M
10
Y7
CS8
CS
CS
EPROM
RAM 2
A0 - A12 Address Bus
Figure 3.8: Linear Select Address Decoding - Circuit Diagram
Block
No
A15 A14 A13 A12 A11 A10 A9 A8 A7 A6 A5 A4 A3 A2 A1 A0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0 0
0
0
0
1
0
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1 1
0 0
1 1
0 0
0
1
1
0
0
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1 1
0 0
1 1
0 0
1
1
0
1
0
0
1
0
1
0
1
0
1
0
1
0
1
0
1
0
1 1
0 0
1 1
0 0
1 1
0 0
1
1
0
1
1
1
1
1
1
1
1 1
1 1
1 1
1
3
5
7
Hex
Address
range
0000H
to
1FFFH
4000H
to
5FFFH
8000H
to
9FFFH
C000H
to
DFFFH
Table 3.4: Linear Select Decoding – Memory Address Ranges
The same EPROM is selected both for the address range 0000H to 1FFFH and
8000H to 9FFFH. This is due to the signals A13 and A14 which have same values in both
the cases and hence activates the same chip select signal CS1. Similarly the RAM is
selected both for the address range 4000H to 5FFFH and C000H to DFFFH. The memory
map for this case is as shown in figure 3.9.
- 11. Copyright © 2008 by Dr.K.K.Thyagharajan
11
FFFF
E000
DFFF
Memory
C000 Block7(8KB)
BFFF
A000
9FFF Memory
8000 Block5(8KB)
7FFF
6000
5FFF Memory
4000 Block3(8KB)
3FFF
RAM 2
Unused memory
space
EPROM 1
RAM 2
2000
1FFF Memory
0000
Block1(8KB) EPROM 1
Figure 3.9: Linear Select Address Decoding – Memory Map
A13
OR 2
OR 1
A13
CS1
CS3
A14
A14
CS
EPROM
8K
CS
RAM 2
8K
A0 - A12 Address Bus
Figure 3.10: Linear Select Address Decoding using OR gates
In this type of address decoding, if we want to add a RAM chip at the address
space 8000H to 9FFFH, which was unnecessarily occupied by the first EPROM, we have
to change the decoding circuit. We may not simply add additional circuitry without
disturbing the already existing circuit. In this method each location in the physical
memory is accessed by more than one address (here two addresses) i.e. address space is
wasted. This type of address decoding is called as Linear Select or partially decoded
addressing. This happens when we have few address lines without being used for
decoding. The same circuit can also be implemented using OR gates as shown in figure
3.10. In this circuit the OR gates require only two inputs. This is possible, since we have
left A15 address line unconnected. Therefore, linear select addressing reduces the
hardware requirements.
- 12. Copyright © 2008 by Dr.K.K.Thyagharajan
3.2
12
Bus Contention
Consider the timing diagram shown in figure 3.11 that reads a byte (opcode) from
the memory. At the end of the ALE signal, all the 16 bits of the address will be available
on the bus. The decoder (74LS138) produces the chip select signals from these address
lines after a small decoding delay. The chip select signal will be generated with in 30 ns
after the trailing edge of ALE signal. When the chip select signal enables the memory
chip, since the address has been already placed on the address pins of the memory, the
memory will place the data on the data lines after a small access delay called the memory
access time. The output buffer of the memory will be enabled within 20 ns after getting
CS signal. The memory access time is the time delay between the address placed on the
address bus by the address decoder and the data placed by the memory on the data bus. If
the data are placed before the end of the T1 state, both the lower order address and the
data from the memory will be placed on the AD0 – AD7 lines i.e. both the microprocessor
and the memory try to access the same lines as shown in figure 3.11 by hashed lines. If
the microprocessor tries to place 1 (+5V) and the memory tries to place 0 (0 V or
Ground), then the microprocessor pin will be shorted through the memory. This may
even damage the processor. This is called bus contention. So, the bus contention will
occur when two ICs try to place data on the same bus at the same instant of time. This
effect will be very adverse in the case of high speed memories interfaced with slow
processors. Since the RD or WR signal is issued by the processor only after the first Tstate, if these signals are used along with the output of the address decoders to select the
chip, the content of the memory will not be placed before the start of the T2 state.
- 13. Copyright © 2008 by Dr.K.K.Thyagharajan
T2
T1
13
T3
T4
CLK
A15
A8
AD7
41H
Unspecified
High-Order Memorry Address
Low-Order address
Data from memory
00H
F
4FH Opcode
AD0
C
Delay in the decoder
ALE
CS
Access Time
Bus Contention
G
Data from
the Memory
4FH Opcode
H
Bus Contention
CS
E
B
RD
D
A
Avoiding Bus
Contention
Figure 3.11: Bus Contention – Reading Opcode from the Memory
- 14. Copyright © 2008 by Dr.K.K.Thyagharajan
A13
A
B
Y1
A15
C
CS1
Y0
A14
Y2
+5V
CS2
CS3
CS4
CS5
G1
RD
Y3
Y4
IO/M
14
Y5
CS6
G2
Y6
CS7
Y7
CS8
74LS138
G3
WR
CS
RAM 2
CS
EPROM
A0 - A12 Address Bus
Figure 3.12: Avoiding Bus Contention
By using a standard 3 to 8 decoder as shown in figure 3.12, the CS signal can be
delayed until the end of T1 state. The address decoder is enabled by the G3 signal and
hence the CS signal is generated only when RD or WR is low. In figure 3.11, point A
activates point B and point B in turn activates point C to avoid bus contention.
If you are using logic gates to decode the addresses, bus contention may be
avoided by combining the output of the final stage of the decoder with RD or WR signal
to produce an active low chip select signal.
NB: Instead of connecting the RD or WR signals to the input or output of the
address decoder, the best practice to avoid bus contention is to connect the RD signal to
the OE pin of EPROM/RAM and WR signal to R/W pin of RAM.
Example 3.1:
Interface a 4K EPROM, one 4K RAM and one 8K RAM to a microprocessor with
the following Memory Map.
- 15. Copyright © 2008 by Dr.K.K.Thyagharajan
15
Figure 3.13: Memory Map
A memory chip select decoder is used to provide chip select signal for each
memory device (IC). This will decide the address range that is allotted for each memory
IC. 74LS138 is a 3 to 8 decoder and it can be used for this purpose. In this example the
minimum memory block size is 4K. To access 1K locations 10 address lines must be used
(210 =1K = 1024 locations). So to access 4K locations (4 X 1K = 22 X 210 = 212) 12
address lines (A0 – A11) must be used. Since 8085 has 16 address lines the decoding can
be indicated as shown below.
Figure 3.14: Variable Address Lines
While accessing 4 K locations the lower 12 bits (A0 – A11) can have either 0 or 1.
If all the 12 bits are 0 then that will be the starting address of that memory block if all 12
bits are 1 that will be the end address of the block. The remaining 4 address lines (A12 A15) are the block select address signals, which decide the memory block number. For
example, if A15 – A12 is 0000 then memory block that can be accessed by the remaining
address lines is block 0. If it is 0001 the memory block that can be accessed is 1. Note
that the memory blocks 5 and 6 are combined because a 8-K RAM is to be placed there.
The following table shows the starting and ending address of each memory block.
- 16. Copyright © 2008 by Dr.K.K.Thyagharajan
Block
No
A15 A14 A13 A12 A11 A10 A9 A8
16
A7 A6 A5 A4
0
0
0
0 0
0
0 0 0
0 0
1
0
0
0
0
0 0
0 1
1
0
1 1 1
0 0 0
1 1 1 1
0 0 0 0
1 1 1 1
0 0 0 0
2
0
0
0
0
0
1
1
0
1
0
1 1 1
0 0 0
1 1 1 1
0 0 0 0
1 1 1 1
0 0 0 0
3
0
0
0
0
1
1
0
1
1
0
1 1 1
0 0 0
1 1 1 1
0 0 0 0
1 1
0 0
1
0
1
0
4
0
0
0
1
1
0
1
0
1
0
1 1 1
0 0 0
1 1 1 1
0 0 0 0
1 1
0 0
1
0
1
0
5
0
0
1
1
0
0
0
1
1
0
1 1 1
0 0 0
1 1 1 1
0 0 0 0
1 1
0 0
1
0
1
0
0
1
0
1
1
1
1 1
1
1
1
1
1
1
0
0 0 0
1 1 1 1
:
:
:
0 0 0 0
0 0 0
0
1
1
1
1
1
1
1 1 1
1
15
1 1
1 1
0 0
A3 A2 A1 A0
1 1 1 1
0
0 0
0
Hex
Address
Range
0000H
to
0FFFH
1000H
to
1FFFH
2000H
to
2FFFH
3000H
to
3FFFH
4000H
to
4FFFH
5000H
to
5FFFH
F000H
to
FFFFH
Table 3.5: Memory Map
The 74LS138 decoder has three input pins marked as CBA when all these three
pins are 0 then the output pin O0 will be activated i.e. O0 becomes zero. When CBA is
001 the O1 will be activated and all other output pins will be at high state. i.e. any one of
the output pins O0 to O7 is selected based on the input A, B and C and this will happen
only if the control pins G1, G2 and G3 are properly activated. Table 3.5 shows that
address line A15 is always zero for the address range 0000H to 6FFFH. So this can be
connected to G3 pin of the 74LS138. The output pins of 74LS138 must be selected only
for memory operations. During memory operations IO/M = 0. So IO/ M is connected to
G2 pin of the 74LS138 and hence for any IO operation the chip 74LS138 will not
activate any of the output pins. The pin G1 is connected to +5V through a resistor to
enable 74LS138.
The O0 pin of 74LS138 is connected to the chip select (CS) signal of the 4K
EPROM. This allows selecting this EPROM chip when the microprocessor sends
address in the range 0000H to 0FFFH. Similarly O1 pin of 74LS138 is connected to the
CS of 4K RAM. Since no memory chip is connected in the address range 2000H to
3FFFH, the output pin Q2 is left free. The 8K RAM requires address from 4000H to
- 17. Copyright © 2008 by Dr.K.K.Thyagharajan
17
5FFFH and these addresses are covered by Q4 and Q5 pins. So these two pins are
connected to a two input AND gate. The output of the AND gate is connected to the chip
select pin of the RAM and this CS will be activated (active low) if either Q4 or Q5 is low.
i.e. the output of the AND gate will be 0 for the addresses 4000H to 5FFFH and hence the
8K RAM chip is selected for this address range.
C B
0 0
0 0
0 1
0 1
1 0
1 0
1 1
1 1
A
0
1
0
1
0
1
0
1
Output pin Activated
O0
O1
O2
O3
O4
O5
O6
O7
Table 3.1: Function Table of 74LS138
The address lines A0 to A11 of 8085 MPU must be connected to the corresponding
address pins of the memory chip and the data lines D0 to D7 of MPU must be connected
to the corresponding data pins of all the memory chips. Since 8K RAM has 13 address
lines in addition to the twelve address lines (A0 to A11) A12 of MPU must also be
connected to the A12 pin of the 8K RAM. You note that bus contention is not addressed in
this circuit.
Y0
A
A12
0000 - 0FFF
A13
B
Y1
1000 - 1FFF
A14
C
Y2
2000 -2FFF
Y3
3000 - 3FFF
+5V
74LS138
Y4
G1
A15
5000 - 5FFF
G2
IO/M
4000 - 4FFF
Y5
Y6
6000 - 6FFF
Y7
7000 - 7FFF
G3
From MPU
Address Bus A11 . . . A0
CS
4K
EPROM
From MPU
CS
4K
RAM
Data Bus D7 . . . D0
Figure3.15: Memory Interfacing
A12
CS
8K
RAM
- 18. Copyright © 2008 by Dr.K.K.Thyagharajan
3.3
18
Interfacing low speed memory
To avoid bus contention, the RD or WR signal of the microprocessor is used
along with the address decoding to produce the CS signal for the memory. The RD or
WR signal starts approximately in the middle of the T2 state and available up to the end
T3 state. If the microprocessor operates at 3 MHz, these signals are available for
approximately 1.5 T state or 500 ns (1.5 / 3 MHz = 0.5 micro seconds) and hence the
memory access time should be less than 500 ns. If the memory access time is more than
this value the data cannot be read. If a low speed memory has access time greater than
this value, Wait cycles are introduced between T2 and T3. In such a case, the CS select
signal need not be delayed until RD or WR signals are issued by the microprocessor. The
memory can decode the address as soon as it gets the address from the microprocessor. In
this case the total delay introduced will be equal to the sum of the delays introduced by
the decoder, wait cycles and the delay introduced due to memory access time. For
example 2716 EPROM has access time of 450 ns and therefore it can be directly
interfaced with 3 MHz 8085, but one wait cycle is necessary if it is interfaced with a
microprocessor operating at 5 MHz. A low speed memory can also be interfaced with a
high speed processor by using low frequency crystal but this will slow down not only the
memory read/write operation but all activities of the processor.
The READY pin of 8085 processor is used to introduce wait states. For normal
operations of the processor this pin should be connected to logic high. When this pin is
made low, the processor will enter into wait state until it is made high again. A special
circuitry called Wait State Generator as shown in figure 3.16 is required for this purpose.
This circuit will introduce a wait state equal to one T-state of the processor cycle as
shown in figure 3.17. The ALE signal fed to the clock input of the first D-type flip-flop
will transfer the logic high input (+5V) placed on its D0 pin to Q0 during its positive edge.
The Q0 output is connected to D1 input of the second flip-flop. So, the logic high output
on Q0 will be transferred to the output Q1 of the second flip-flop during the positive edge
of T2 state. The complement value of Q1 is available on Q1 pin of the second flip-flop.
This will reset the first flip-flop i.e. Q0 will become logic 0 as shown in figure 3.17. At
the same time since Q1 has been connected to the READY pin of the processor, the
processor enters into wait state. It will remain in that state until Q1 becomes high. This
extends the length of the RD signal as shown in figure 3.17 and allows memory to take
one more clock period time to place the contents on the bus. The logic 0 placed on D1 pin
during this time will be transferred to Q1 in the positive edge of the next clock cycle (T3
state). This will make Q1 logic high and hence the READY signal becomes high and the
processor continues its work. It should be noted that wait state can be introduced in any
machine cycle of 8085 except in bus idle cycle. The 8085 checks the READY signal at
the second T-state of every machine cycle.
- 19. Copyright © 2008 by Dr.K.K.Thyagharajan
D Flip-Flop
+5V
ALE
D Flip-Flop
Q0
D0
19
Q1
D1
CLK
CLK
Q0
Q1
R
R
To READY
Pin of 8085
CLKOUT from 8085
Figure 3.16: Wait State Generator
Memory Read
T1
T2
Memory Read
T3
TWAIT
POSITIVE OR LEADING EDGE OF THE CLOCK CYCLE
CLK
A15
A8
AD7
AD0
41H
High-Order Memorry Address
Low-Order address
00H
Data
Delayed Data
4FH
+VE EDGE
ALE
Q0
READY/ Q1
RD
EXTENDED READ SIGNAL
Figure 3.17: Timing Diagram with Wait State
T1
- 20. Copyright © 2008 by Dr.K.K.Thyagharajan
3.4
20
Interfacing More than 64K Memory
The 8085 microprocessor can address only 64K memory at a time, if you want to
interface more than 64K memory; you can use a jumper or an IO port bit as shown in
figure 3.18. the chip select signal CS3 will select RAM 2 if the jumper connects B and C
and it will select RAM 3 if the jumper connects A and B. Similarly CS4 will select RAM
4 if the port bit is zero and RAM 5 if the port bit is one. The main difference between
jumper selection and port bit selection is the manual selection in the first case and
automatic or program activated selection in the second case.
CS1
A14
A
Y0
A15
B
Y1
CS2
C
Y2
CS3
+5V
74LS138 Y3
CS4
Y4
G1
G2
C
B
Y6
G3
IO/M
A
Y5
Y7
CS
PORT BIT
16K
RAM 5
CS
16K
RAM 4
CS
CS
16K
RAM 3
16K
RAM 2
CS
16K
RAM1
CS
16K
EPROM
A0 - A13 Address Bus
Figure 3.4: Memory Interface Diagram
Figure 3.18: Interfacing more than 64 K Memory
Summary
A microprocessor has address space for memory. Any memory chip that is being
interfaced with processor must fit into this memory space. A memory map diagram
shows the address boundary for each memory chip interfaced with the processor and an
address decoder decides the address range for each chip. There are two types of address
decoding viz absolute address decoding and linear select address decoding. When high
speed memories are interfaced with slow processors bus contention may occur and when
a low speed memory is to be accessed by a fast processors wait state must be introduced
using external circuitry.