Datasheet DP8432V-33 Datasheet (NSC)

Page 1

DP8430V/31V/32V-33 microCMOS Programmable 256k/1M/4M Dynamic RAM Controller/Drivers

General Description

The DP8430V/31V/32V dynamic RAM controllers provide a low cost, single chip interface between dynamic RAM and all 8-, 16- and 32-bit systems. The DP8430V/31V/32V generate all the required access control signal timing for DRAMs. An on-chip refresh request clock is used to automatically refresh the DRAM array. Refreshes and accesses are arbitrated on chip. If necessary, a WAIT put inserts wait states into system access cycles, including burst mode accesses. RAS RAS

precharge time after refreshes and back to back ac-

low time during refreshes and

cesses are guaranteed through the insertion of wait states. Separate on-chip precharge counters for each RAS can be used for memory interleaving to avoid delayed back to back accesses because of precharge. An additional feature of the DP8432V is two access ports to simplify dual accessing. Arbitration among these ports and refresh is done on chip.

Control

of Pins

(PLCC) Outputs

DP8430V 68 9 256 kbit 4 Mbytes Single Access Port

DP8431V 68 10 1 Mbit 16 Mbytes Single Access Port

DP8432V 84 11 4 Mbit 64 Mbytes Dual Access Ports (A and B)

of Address

or DTACK out-

output

Features

On chip high precision delay line to guarantee critical DRAM access timing parameters

microCMOS process for low power

High capacitance drivers for RAS, CAS,WEand DRAM address on chip

On chip support for nibble, page and static column DRAMs

Byte enable signals on chip allow byte writing in a word size up to 32 bits with no external logic

Can use a single clock source. Up to 33 MHz operating frequency

On board Port A/Port B (DP8432V only)/refresh arbitration logic

Direct interface to all major microprocessors

4 RAS and 4 CAS drivers (the RAS and CAS configuration is programmable)

Largest Direct Drive Access

DRAM Memory Ports

Possible Capacity Available

July 1993

DP8430V/31V/32V-33 microCMOS Programmable

256k/1M/4M Dynamic RAM Controller/Drivers

Block Diagram

DP8430V/31V/32V DRAM Controller

TRI-STATEÉis a registered trademark of National Semiconductor Corporation. Staggered Refresh

1995 National Semiconductor Corporation RRD-B30M75/Printed in U. S. A.

is a trademark of National Semiconductor Corporation.

TL/F/11118

FIGURE 1

TL/F/11118– 1

Page 2

Table of Contents

1.0 INTRODUCTION

2.0 SIGNAL DESCRIPTIONS

2.1 Address, R/W and Programming Signals

2.2 DRAM Control Signals

2.3 Refresh Signals

2.4 Port A Access Signals

2.5 Port B Access Signals (DP8432V)

2.6 Common Dual Port Signals (DP8432V)

2.7 Power Signals and Capacitor Input

2.8 Clock Inputs

3.0 PROGRAMMING AND RESETTING

3.1 Reset

3.2 Programming Methods

3.2.1 Mode Load Only Programming

3.2.2 Chip Selected Access Programming

3.3 Internal Programming Modes

4.0 PORT A ACCESS MODES

4.1 Access Mode 0

4.2 Access Mode 1

4.3 Extending CAS with Either Access Mode

4.4 Read-Modify-Write Cycles with Either Access Mode

4.5 Additional Access Support Features

4.5.1 Address Latches and Column Increment

4.5.2 Address Pipelining

4.5.3 Delay CAS

5.0 REFRESH OPTIONS

5.1 Refresh Control Modes

5.1.1 Automatic Internal Refresh

5.1.2 Externally Controlled Refresh

During Write Accesses

5.2 Refresh Cycle Types

5.2.1 Conventional Refresh

5.2.2 Staggered Refresh

5.2.3 Error Scrubbing Refresh

5.3 Extending Refresh

6.0 PORT A WAIT STATE SUPPORT

6.1 WAIT

6.2 DTACK

Type Output

6.3 Dynamically Increasing the Number of Wait States

6.4 Guaranteeing RAS

Low Time and RAS Precharge

Time

7.0 RAS

AND CAS CONFIGURATION MODES

7.1 Byte Writing

7.2 Memory Interleaving

7.3 Address Pipelining

7.4 Error Scrubbing

7.5 Page/Burst Mode

8.0 TEST MODE

9.0 DRAM CRITICAL TIMING PARAMETERS

9.1 Programmable Values of t

9.2 Calculation of t

RAH

and t

RAH

ASC

and t

ASC

10.0 DUAL ACCESSING (DP8432V)

10.1 Port B Access Mode

10.2 Port B Wait State Support

10.3 Common Port A and Port B Dual Port Functions

10.3.1 GRANTB Output

10.3.2 LOCK

Input

11.0 ABSOLUTE MAXIMUM RATINGS

12.0 DC ELECTRICAL CHARACTERISTICS

13.0 AC TIMING PARAMETERS

14.0 DP8430V/31V/32V USER HINTS

Page 3

1.0 Introduction

The DP8430V/31V/32V DRAM controllers are the latest devices based upon the DP8420A/21A/22A predecessors. The DP8430V/31V/32V implement changes which do not allow them to be pin compatible with any of the DP842XA or the DP842XV DRAM controllers. Two changes have been made: The limits for the input frequency to DELCLK have been increased making possible the use of a single clock source. A RESET input is now available making the reset procedure easier. These changes, although minimal, facilitate the use of the controllers and make them even more attractive for high performance applications. The controllers incorporate address latches, refresh counter, row/column/ refresh address multiplexer, delay line, refresh/access/precharge arbitration logic and high capacitive drivers. The DP8430V/31V/32V DRAM controllers allow any manufacturer’s CPU or bus to directly interface to DRAM arrays up to 64 Mbytes in size.

Reset:

The user must reset the controller before programming it. Reset is achieved by asserting the RESET 16 positive edges of clock.

Programming:

After reset, the user can program the controller by either one of two methods: Mode Load Only Programming or Chip Select Access Programming. The chip is programmed through the address bus.

Initialization Period:

Once the DP8430V/31V/32V has been programmed for the first time, a 60 ms initialization period is entered. During this time the DRC performs refreshes to the DRAM array so further warm up cycles are unnecessary. The initialization period is entered only after the first programming after a reset.

Accessing Modes:

After resetting and programming the chip, the DP8430V/ 31V/32V is ready to access the DRAM. There are two modes of accessing with these controllers. Mode 0, which indicates RAS RAS

asynchronously.

Refresh Modes:

Two refresh modes can be programmed. The user can choose Automatic Internal Refresh or Externally Controlled Refresh. With any refresh mode the user can perform burst refreshes.

Refresh Types:

There are three types of refreshing available: Conventional, Staggered and Error Scrubbing. Any refresh control mode can be used with any type of refresh.

Wait Support:

The DP8430V/31V/32V have wait support available as DTACK Transfer ACKnowledge, is useful for processors whose wait signal is active high. WAIT whose wait signal is active low. The user can choose either at programming. These signals are used by the on chip arbiter to insert wait states to guarantee the arbitration between accesses, refreshes and precharge. Both signals are independent of the access mode chosen and both signals can be dynamically delayed further through the WAITIN the DP8430V/31V/32V.

synchronously and Mode 1, which indicates

or WAIT. Both are programmable. DTACK, Data

is useful for those processors

input for at least

signal to

Sequential Accesses (Static Column/Page Mode):

The DP8430V/31V/32V have address latches, used to latch the bank, row and column address inputs. Once the address is latched, a COLumn INCrement (COLINC) feature can be used to increment the column address. The address latches can also be programmed to be fall through. COLINC can be used for Sequential Accesses of Static Column DRAMs. Also, COLINC in conjunction with ECAS be used for Sequential Accesses to Page Mode DRAMs.

RAS

and CAS Configuration (Byte Writing):

The RAS and CAS drivers can be configured to drive a one, two or four bank memory array up to 32 bits in width. The ECAS

signals can then be used to select one of four CAS

drivers for Byte Writing with no extra logic.

Memory Interleaving:

When configuring the DP8430V/31V/32V for more than one bank, Memory Interleaving can be used. By tying the low order address bits to the bank select lines B0 and B1, sequential back to back accesses will not be delayed since these controllers have separate precharge counters per bank.

Address Pipelining:

The DP8430V/31V/32V are capable of performing Address Pipelining. In address pipelining, the DRC will guarantee the column address hold time and switch the internal multiplexor to place the row address on the address bus. At this time, another memory access to another bank can be initiated.

Dual Accessing:

Finally, the DP8432V has all the features previously mentioned and unlike the DP8430V/31V, the DP8432V has a second port to allow a second CPU to access the same memory array. The DP8432V has four signals to support Dual Accessing, these signals are AREQB and GRANTB. All arbitration for the two ports and refresh is done on chip by the controller through the insertion of wait states. Since the DP8432V has only one input address bus, the address lines must be multiplexed externally. The signal GRANTB can be used for this purpose.

Terminology:

The following explains the terminology used in this data sheet. The terms negated and asserted are used. Asserted refers to a ‘‘true’’ signal. Thus, ‘‘ECAS0 the ECAS0 ed’’ means the COLINC input is at a logic 1. The term negated refers to a ‘‘false’’ signal. Thus, ‘‘ECAS0 means the ECAS0 negated’’ means the input COLINC is at a logic 0. The table shown below clarifies this terminology.

input is at a logic 0. The term ‘‘COLINC assert-

input is at a logic 1. The term ‘‘COLINC

Signal Action Logic Level

Active High Asserted High

Active High Negated Low

Active Low Asserted Low

Active Low Negated High

inputs can

, ATACKB, LOCK

asserted’’ means

negated’’

Page 4

Connection Diagrams

Top View

FIGURE 2

Order Number DP8430V-33

See NS Package Number V68A

TL/F/11118– 2

Top View

FIGURE 4

Order Number DP8432V-33

See NS Package Number V84A

Top View

TL/F/11118– 3

FIGURE 3

Order Number DP8431V-33

See NS Package Number V68A

TL/F/11118– 4

Page 5

2.0 Signal Descriptions

Pin Device (If not Input/

Name Applicable to All) Output

2.1 ADDRESS, R/W AND PROGRAMMING SIGNALS

R0–10 DP8432V I ROW ADDRESS: These inputs are used to specify the row address during an access R0–9 DP8430V/31V I

C0–10 DP8432V I COLUMN ADDRESS: These inputs are used to specify the column address during an C0–9 DP8430V/31V I

B0, B1 I BANK SELECT: Depending on programming, these inputs are used to select a group

ECAS0–3 I ENABLE CAS: These inputs are used to enable a single or group of CAS outputs

RESET I RESET: At power up, this input is used to reset the DRAM controller. The user must

WIN I WRITE ENABLE IN: This input is used to signify a write operation to the DRAM. If

COLINC I COLUMN INCREMENT: When the address latches are used, and RFIP is negated, (EXTNDRF) I

ML I MODE LOAD: This input signal, when low, enables the internal programming register

2.2 DRAM CONTROL SIGNALS

Q0–10 DP8432V O DRAM ADDRESS: These outputs are the multiplexed output of the R0 –9, 10 and Q0–9 DP8431V O Q0–8 DP8430V O

RAS0–3 O ROW ADDRESS STROBES: These outputs are asserted to latch the row address

CAS0–3 O COLUMN ADDRESS STROBES: These outputs are asserted to latch the column

WE O WRITE ENABLE or REFRESH REQUEST: This output asserted specifies a write (RFRQ

to the DRAM. They are also used to program the chip when ML R10).

access to the DRAM. They are also used to program the chip when ML (except C10).

of RAS and CAS outputs to assert during an access. They are also used to program the chip when ML

when asserted. In combination with the B0, B1 and the programming bits, these inputs select which CAS ECAS

signals can also be used to toggle a group of CAS outputs for page/nibble mode accesses. They also can be used for byte write operations. If ECAS negated during programming, continuing to assert the ECAS or AREQB during an access, will cause the CAS outputs to be extended while the RAS

outputs are negated (the ECASn inputs have no effect during scrubbing

refreshes).

keep RESET must remain negated (high) to avoid an unwanted reset.

ECAS0 is asserted during programming, the WE output will follow this input. This input asserted will also cause CAS bit C9 is asserted during programming.

this input functions as COLINC. Asserting this signal causes the column address to be incremented by one. When RFIP refresh cycle by any number of periods of CLK until it is negated.

that stores the programming information.

C0–9, 10 and form the DRAM address bus. These outputs contain the refresh address whenever RFIP series damping resistors.

contained on the outputs Q0–8, 9, 10 into the DRAM. When RFIP RAS

outputs are used to latch the refresh row address contained on the Q0–8, 9, 10 outputs in the DRAM. These outputs contain high capacitive drivers with 20X series damping resistors.

address contained on the outputs Q0–8, 9, 10 into the DRAM. These outputs have high capacitive drivers with 20X series damping resistors.

operation to the DRAM. When negated, this output specifies a read operation to the DRAM. When the DP8430V/31V/32V is programmed in address pipelining mode or when ECAS0 is negated during programming, this output will function as RFRQ. RFRQ

asserted, specifies that 13 msor15ms have passed. RFRQ can be used to externally request a refresh through the input RFSH capacitive driver and a 20X series damping resistor.

is asserted.

output or CAS outputs will assert during an access. The

low for at least 16 positive edges of clock. After programming this input

is asserted. They contain high capacitive drivers with 20X

Description

is asserted (except

is asserted

0is

0 while negating AREQ

to delay to the next positive clock edge if address

is asserted, this signal is used to extend the

is asserted, the

. This output has a high

Page 6

2.0 Signal Descriptions (Continued)

Pin Device (If not Input/

Name Applicable to All) Output

2.3 REFRESH SIGNALS

RFIP O REFRESH IN PROGRESS: This output is asserted prior to a refresh cycle and is

RFSH I REFRESH: This input asserted will request a refresh. If this input is continually

2.4 PORT A ACCESS SIGNALS

ADS I ADDRESS STROBE or ADDRESS LATCH ENABLE: Depending on programming, (ALE) I

CS I CHIP SELECT: This input signal must be asserted to enable a Port A access.

AREQ I ACCESS REQUEST: This input signal in Mode 0 must be asserted some time after

WAIT O WAIT or DTACK: This output can be programmed to insert wait states into a CPU (DTACK

WAITIN I WAIT INCREASE: This input can be used to dynamically increase the number of

2.5 PORT B ACCESS SIGNALS

AREQB DP8432V I PORT B ACCESS REQUEST: This input asserted will latch the row, column and bank

ATACKB DP8432V O ADVANCED TRANSFER ACKNOWLEDGE PORT B: This output is asserted when

only

negated when all the RAS

asserted, the DP8430V/31V/32V will perform refresh cycles in a burst refresh fashion until the input is negated.

this input can function as ADS when asserted along with CS an access will start from the positive clock edge of CLK as soon as possible. In Mode 1, the input functions as ADS RAS

to assert if no other event is taking place. If an event is taking place, RAS will be asserted from the positive edge of CLK as soon as possible. In both cases, the low going edge of this signal latches the bank, row and column address if programmed to do so.

the first positive clock edge after ALE has been asserted. When this signal is negated, RAS before ADS access.

access cycle. With R7 negated during programming, the output will function as a WAIT With R7 asserted during programming, the output will function as DTACK case, the output will be negated to signify a wait condition and will be asserted to signify the access has taken place. Each of these signals can be delayed by a number of positive clock edges or negative clock levels of CLK to increase the microprocessor’s access cycle through the insertion of wait states.

positive clock edges of CLK until DTACK during a DRAM access.

address if programmed, and requests an access to take place for Port B. If the access can take place, RAS RAS

the access RAS the appropriate DTACK

is negated for the access. In Mode 1, this signal must be asserted

can be negated. When this signal is negated, RAS is negated for the

type output. In this case, the output will be active low to signal a wait condition.

will assert as soon as possible from a positive edge of CLK.

is asserted for a Port B access. This signal can be used to generate

or WAIT type signal for Port B’s CPU or bus.

Description

outputs are negated for that refresh.

or ALE. In mode 0, the input functions as ALE and

causes an internal latch to be set. Once this latch is set

and when asserted along with CS, causes the access

. In this

will be asserted or WAIT will be negated

will assert immediately. If the access has to be delayed,

Page 7

2.0 Signal Descriptions (Continued)

Pin Device (If not Input/

Name Applicable to All) Output

2.6 COMMON DUAL PORT SIGNALS

GRANTB DP8432V O GRANT B: This output indicates which port is currently granted access to the DRAM

LOCK DP8432V I LOCK: This input can be used by the currently granted port to ‘‘lock out’’ the other

2.7 POWER SIGNALS AND CAPACITOR INPUT

GND I GROUND: Supply Voltage Reference.

CAP I CAPACITOR: This input is used by the internal PLL for stabilization. The value of the

2.8 CLOCK INPUTS

There are two clock inputs to the DP8430V/31V/32V, CLK and DELCLK. These two clocks may both be tied to the same clock input, or they may be two separate clocks, running at different frequencies, asynchronous to each other.

CLK I SYSTEM CLOCK: This input may be in the range of 0 Hz up to 25 MHz. This input is

DELCLK I DELAY LINE CLOCK: The input frequency to DELCLK should be in the range of

only

array. When GRANTB is asserted, Port B has access to the array. When GRANTB is negated, Port A has access to the DRAM array. This signal is used to multiplex the signals R0–8, 9, 10; C0 –8, 9, 10; B0 – 1; WIN when using dual accessing.

port from the DRAM array by inserting wait states into the locked out port’s access cycle until LOCK is negated.

I POWER: Supply Voltage.

ceramic capacitor should be 0.1 mF and should be connected between this input and ground.

generally a constant frequency but it may be controlled externally to change frequencies or perhaps be stopped for some arbitrary period of time.

This input provides the clock to the internal state machine that arbitrates between accesses and refreshes. This clock’s positive edges and negative levels are used to extend the WAIT RAS

precharge time and RAS low time during refresh.

All Port A and Port B accesses are assumed to be synchronous to the system clock CLK.

12 MHz to 40 MHz. This frequency will be internally divided by choosing a divisor when programming the part. The result of the division should be a frequency of 2 MHz. This is because the Phase Lock Loop that generates the delay line assumes an input clock frequency of 2 MHz. If after dividing DELCLK by one of the internal divisors (6, 8, 10, 12, 14, 16, 18 or 20) the resulting frequency is not 2 MHz, the delay line will suffer.

For example, if the DELCLK frequency is 18 MHz and a divide by 8 is chosen, programming bits C0–2, the resulting frequency will be 2.25 which is 12.5% off of 2 MHz. Therefore, the DP8430V/31V/32V will produce delays that are shorter (faster delays) than what is intended. On the other hand, if divide by 10 was chosen, the resulting frequency will be 1.8 MHz, this frequency will produce delays that are longer (slower delays) than intended.

This clock is also divided to create the internal refresh clock.

(DTACK) signals. Ths clock is also used as the reference for the

Description

; LOCK and ECAS0–3 to the DP8432V

Page 8

3.0 Programming and Resetting

The DP8430V/31V/32V must be reset before it can be programmed. After reset, the DRAM controller is programmed through the address bus by either one of two methods; Mode Load Only Programming or Chip Select Access Programming. After the first programming after a reset, the chip enters a 60 ms initialization period. During this period the controller performs refreshes every 13 msor15ms, this makes further DRAM warm up cycles unnecessary. After this stage the DRAM controller can be programmed as many times as the user wishes and the 60 ms initialization period will not be entered into unless the chip is reset and programmed again. During the 60 ms initialization period, RFIP

is asserted and RAS toggles every 13 msor15ms depending on the programming bit for refresh (C3). CAS be negated and the Q outputs will count from 0 to 2047 refreshing the entire DRAM array. The initialization time period is given by the following formula. T

4096 * (Clock

Divisor Select) * (Refresh Clock Fine Tune)/(DELCLK Frq.)

3.1 RESET

The DP8430V/31V/32V have a RESET

input pin which facilitates the reset procedure required for proper operation. Reset is accomplished by asserting the RESET least 16 positive edges of clock as shown in

Figure 5

The DRC may be programmed anytime on the fly, but the user must make sure that no access or refresh is in progress. RESET

is asynchronous.

will

input for at

3.2 PROGRAMMING METHODS

3.2.1 Mode Load Only Programming

To use this method the user asserts ML nal programming register. After ML

enabling the inter-

is asserted, a valid programming selection is placed on the address bus, B0, B1 and ECAS0

inputs, then ML is negated. When ML is negated the programming bits are latched into the internal programming register and the DP8430V/31V/32V is programmed, see controller must not be refreshing, RFIP

Figure 6

. When programming the chip, the

must be high (1) to

have a successful programming.

3.2.2 Chip Selected Access Programming

The chip can also be programmed by performing a chip selected access. To program the chip using this method, ML

is asserted, then CS is asserted and a valid programming selection is placed on the address bus. When AREQ asserted, the programming bits affecting the wait logic become effective immediately, then DTACK ing the access to terminate. After the access, ML

is asserted allow-

is negated

and the rest of the programming bits take effect.

FIGURE 5. Reset

FIGURE 6. ML Only Programming

FIGURE 7. CS Access Programming

TL/F/11118– 5

TL/F/11118– 6

TL/F/11118– 7

Page 9

3.0 Programming and Resetting (Continued)

3.3 PROGRAMMING BIT DEFINITIONS

Symbol Description

ECAS0 Extend CAS/Refresh Request Select

0 The CASn outputs will be negated with the RASn outputs when AREQ (or AREQB, DP8432V only) is negated.

1 The CASn outputs will be negated, during an acccess (Port A (or Port B, DP8432V only)) when their

B1 Access Mode Select

0 ACCESS MODE 0: ALE pulsing high sets an internal latch. On the next positive edge of CLK, the access (RAS

1 ACCESS MODE 1: ADS

B0 Address Latch Mode

0 ADS or ALE asserted for Port A or AREQB asserted for Port B with the appropriate GRANT latch the input row,

1 The row, column and bank latches are fall through.

C9 Delay CAS during WRITE Accesses

0 CAS is treated the same for both READ and WRITE accesses.

1 During WRITE accesses, CAS will be asserted by the event that occurs last: CAS asserted by the internal delay

C8 Row Address Hold Time

0 Row Address Hold Timee25 ns minimum

1 Row Address Hold Timee15 ns minimum

C7 Column Address Setup Time

0 Column Address Setup Timee10 ns miniumum

1 Column Address Setup Timee0 ns minimum

C6, C5, C4 RAS and CAS Configuration Modes/Error Scrubbing during Refresh

0, 0, 0 RAS0 – 3 and CAS0–3 are all selected during an access. ECASn must be asserted for CASn to be asserted.

0, 0, 1 RAS

0, 1, 0 RAS and CAS singles are selected during an access by B0 – 1. ECAS

0, 1, 1 RAS

1, 0, 0 RAS pairs are selected by B1. CAS0 – 3 are all selected. ECASn must be asserted for CASn to be asserted.

The WE

output pin will function as write enable. Automatic Internal Refresh selected.

corresponding ECAS outputs negating. Scrubbing refreshes are NOT affected. During scrubbing refreshes the CAS outputs will negate along with the RAS

Externally Controlled Refresh selected, WE will function as ReFresh ReQuest (RFRQ).

will start. AREQ

column and bank address.

line or CAS

B0 and B1 are not used during an access. Error scrubbing during refresh.

and CAS pairs are selected during an access by B1. ECASn must be asserted for CAS

0 during an access selects RAS0–1 and CAS0–1. B1e1 during an access selects RAS2–3 and CAS2–3. B0 is not used during an Access. Error scrubbing during refresh.

0, B0e0 during an access selects RAS0 and CAS0.

0, B0e1 during an access selects RAS1 and CAS1.

1, B0e0 during an access selects RAS2 and CAS2. B1e1, B0e1 during an access selects RAS3 and CAS3. Error scrubbing during refresh.

0–3 and CAS0 –3 are all selected during an access. ECASn must be asserted for CASn to be asserted. B1, B0 are not used during an access. No error scrubbing. (RAS

0 during an access selects RAS0–1 and CAS0–3. B1e1 during an access selects RAS2–3 and CAS0–3. B0 is not used during an access. No error scrubbing.

n inputs are negated. This feature allows the CAS outputs to be extended beyond the RAS

outputs regardless of the state of the ECAS inputs.

will terminate the access.

asserted starts the access (RAS) immediately. AREQ will terminate the access.

asserted on the positive edge of CLK after RAS is asserted.

n to be asserted.

n must be asserted for CASn to be asserted.

only refreshing)

)

Page 10

3.0 Programming and Resetting (Continued)

3.3 PROGRAMMING BIT DEFINITIONS (Continued)

Symbol Description

C6, C5, C4 RAS and CAS Configuration Modes (Continued)

1, 0, 1 RAS and CAS pairs are selected by B1. ECASn must be asserted for CASn to be asserted.

1, 1, 0 RAS singles are selected by B0 –1. CAS0 – 3 are all selected. ECASn must be asserted for CASntobe

1, 1, 1 RAS and CAS singles are selected by B0, 1. ECASn must be asserted for CASn to be asserted.

C3 Refresh Clock Fine Tune Divisor

0 Divide delay line/refresh clock further by 30 (If DELCLK/Refresh Clock Clock Divisore2 MHze15 ms

1 Divide delay line/refresh clock further by 26 (If DELCLK/Refresh Clock Clock Divisor

C2, C1, C0 Delay Line/Refresh Clock Divisor Select

0, 0, 0 Divide DELCLK by 20 to get as close to 2 MHz as possible. 0, 0, 1 Divide DELCLK by 18 to get as close to 2 MHz as possible. 0, 1, 0 Divide DELCLK by 16 to get as close to 2 MHz as possible. 0, 1, 1 Divide DELCLK by 14 to get as close to 2 MHz as possible. 1, 0, 0 Divide DELCLK by 12 to get as close to 2 MHz as possible. 1, 0, 1 Divide DELCLK by 10 to get as close to 2 MHz as possible. 1, 1, 0 Divide DELCLK by 8 to get as close to 2 MHz as possible. 1, 1, 1 Divide DELCLK by 6 to get as close to 2 MHz as possible.

R9 Refresh Mode Select

0 RAS0 – 3 will all assert and negate at the same time during a refresh. 1 Staggered Refresh. RAS

R8 Address Pipelining Select

0 Address pipelining is selected. The DRAM controller will switch the DRAM column address back to the row

1 Non-address pipelining is selected. The DRAM controller will hold the column address on the DRAM address

R7 WAIT or DTACK Select

0 WAIT type output is selected. 1 DTACK

R6 Add Wait States to the Current Access if WAITIN is Low

0 WAIT or DTACK will be delayed by one additional positive edge of CLK. 1 WAIT

0 during an access selects RAS0–1 and CAS0–1.

1 during an access selects RAS2–3 and CAS2–3. B0 is not used during an access. No error scrubbing.

asserted.

0, B0e0 during an access selects RAS0 and CAS0–3.

0, B0e1 during an access selects RAS1 and CAS0–3.

1, B0e0 during an access selects RAS2 and CAS0–3.

1, B0e1 during an access selects RAS3 and CAS0–3. No error scrubbing.

0, B0e0 during an access selects RAS0 and CAS0.

0, B0e1 during an access selects RAS1 and CAS1.

1, B0e0 during an access selects RAS2 and CAS2.

1, B0e1 during an access selects RAS3 and CAS3. No error scrubbing.

refresh period).

outputs during refresh are separated by one positive clock edge. Depending on the

configuration mode chosen, either one or two RAS

s will be asserted.

address after guaranteeing the column address hold time.

bus until the access RAS

s are negated.

(Data Transfer ACKnowledge) type output is selected.

or DTACK will be delayed by two additional positive edges of CLK.

2 MHze13 ms

Page 11

3.0 Programming and Resetting (Continued)

3.3 PROGRAMMING BIT DEFINITIONS (Continued)

Symbol Description

R5, R4 WAIT/DTACK during Burst (See Section 5.1.2 or 5.2.2)

0, 0 NO WAIT STATES; If R7e0 during programming, WAIT will remain negated during burst portion of access.

0, 1 1T; If R7e0 during programming, WAIT will assert when the ECAS inputs are negated with AREQ asserted.

1, 0 (/2T; If R7e0 during programming, WAIT will assert when the ECAS inputs are negated with AREQ asserted.

1, 1 0T; If R7e0 during programming, WAIT will assert when the ECAS inputs are negated. WAIT will negate when

R3, R2 WAIT/DTACK Delay Times (See Section 5.1.1 or 5.2.1)

0, 0 NO WAIT STATES; If R7e0 during programming, WAIT will remain high during non-delayed accesses. WAIT

0, 1 (/2T; If R7e0 during programming, WAIT will negate on the negative level of CLK, after the access RAS.

1, 0 NO WAIT STATES, (/2T; If R7e0 during programming, WAIT will remain high during non-delayed accesses.

1, 1 1T; If R7

R1, R0 RAS Low and RAS Precharge Time

0, 0 RAS asserted during refreshe2 positive edges of CLK.

0, 1 RAS asserted during refreshe3 positive edges of CLK.

1, 0 RAS asserted during refreshe2 positive edges of CLK.

1, 1 RAS

If R7

1 programming, DTACK will remain asserted during burst portion of access.

WAIT

will negate from the positive edge of CLK after the ECASs have been asserted.

If R7e1 during programming, DTACK will negate when the ECAS inputs are negated with AREQ asserted. DTACK

will assert from the positive edge of CLK after the ECASs have been asserted.

WAIT

will negate on the negative level of CLK after the ECASs have been asserted.

If R7

1 during programming, DTACK will negate when the ECAS inputs are negated with AREQ asserted.

DTACK

will assert from the negative level of CLK after the ECASs have been asserted.

the ECAS If R7e1 during programming, DTACK will negate when the ECAS inputs are negated. DTACK will assert when

the ECAS

will negate when RAS is negated during delayed accesses. NO WAIT STATES; If R7

1T; If R7e1 during programming, DTACK will be asserted on the positive edge of CLK after the access RAS.

WAIT (/2T; If R7

1(/2T; If R7 of CLK after the access RAS

RAS RAS

inputs are asserted.

1 during programming, DTACK will be asserted when RAS is asserted.

will negate on the negative level of CLK, after the access RAS, during delayed accesses.

1 during programming, DTACK will be asserted on the negative level of CLK after the access RAS.

0 during programming, WAIT will negate on the positive edge of CLK after the access RAS.

1 during programming, DTACK will be asserted on the negative level of CLK after the positive edge

precharge timee1 positive edge of CLK. will start from the first positive edge of CLK after GRANTB transitions (DP8432V).

precharge timee2 positive edges of CLK. will start from the second positive edge of CLK after GRANTB transitions (DP8432V).

precharge timee2 positive edges of CLK. will start from the first positive edge of CLK after GRANTB transitions (DP8432V).

asserted during refreshe4 positive edges of CLK. precharge timee3 positive edges of CLK. will start from the second positive edge of CLK after GRANTB transitions (DP8432V).

Page 12

4.0 Port A Access Modes

The DP8430V/31V/32V have two general purpose access modes. Mode 0 RAS chronous. One of these modes is selected at programming through the B1 input. A Port A access to DRAM is initiated by two input signals: ADS ways terminated by one signal: AREQ should be synchronous to the input clock.

4.1 ACCESS MODE 0

Mode 0, synchronous access, is selected by negating the input B1 during programming (B1 access, ALE is pulse high and CS time was met, a refresh of DRAM or a Port B access was not in progress, the RAS

synchronous and Mode 1 RAS asyn-

(ALE) and CS. The access is al-

(RASs) would be asserted on the

. These input signals

0). To initiate a Mode 0

is asserted. If precharge

first rising edge of clock. If a refresh or a Port B access is in progress or precharge time is required, the controller will wait until these events have taken place and assert RAS (RASs) on the next positive edge of clock.

Sometime after the first positive edge of clock after ALE and CS

have been asserted, the input AREQ must be asserted. In single port applications, once AREQ be negated. On the other hand, ALE can stay asserted several periods of clock; however, ALE must be negated before or during the period of CLK in which AREQ

The controller samples AREQ on the every rising edge of clock after DTACK AREQ

is sampled negated.

is asserted. The access will end when

is asserted, CS can

is negated.

FIGURE 8a. Access Mode 0

TL/F/11118– 8

Page 13

4.0 Port A Access Modes (Continued)

4.2 ACCESS MODE 1

Mode 1, asynchronous access, is selected by asserting the input B1 during programming (B1 cesses to start immediately from the access request input, ADS

. To initiate a Mode 1 access, CS is asserted followed

by ADS

asserted. If precharge time was met, a refresh of the DRAM or a Port B access was not in progress, the RAS (RASs) would be asserted from ADS being asserted. If a refresh or Port B access is in progress or precharge time is required, the controller will wait until these events have tak-

1). This mode allows ac-

en place and assert RAS of clock.

When ADS asserted. At this time, ADS continue the access. Also, ADS after AREQ new access will not start until ADS again. When address pipelining is not implemented, ADS and AREQ can be tied together.

The access will end when AREQ is negated.

is asserted or sometime after, AREQ must be

has been asserted and negated; however, a

(RASs) from the next rising edge

can be negated and AREQ will

can continue to be asserted

is negated and asserted

FIGURE 8b. Access Mode 1

TL/F/11118– 9

Page 14

4.0 Port A Access Modes (Continued)

4.3 EXTENDING CAS WITH EITHER ACCESS MODE

In both access modes, once AREQ DTACK

if programmed will be negated. If ECAS0 was as-

serted (0) during programming, CAS

is negated, RAS and

(CASs) will be negated

with AREQ CAS been negated, given that the appropriate ECAS

. If ECAS0 was negated (1) during programming,

(CASs) will continue to be asserted after RAS has

inputs are asserted. This allows a DRAM to have data present on the data out bus while gaining RAS

precharge time.

FIGURE 9a. Access Mode 0 Extending CAS

FIGURE 9b. Access Mode 1 Extending CAS

TL/F/11118– 10

TL/F/11118– 11

Page 15

4.0 Port A Access Modes (Continued)

4.4 READ-MODIFY-WRITE CYCLES WITH EITHER ACCESS MODE

There are 2 methods by which this chip can be used to do read-modify-write access cycles. The first method involves doing a late write access where the WIN some delay after CAS volves doing a page mode read access followed by a page mode write access with RAS

CASn must be toggled using the ECASn inputs and WIN has to be changed from negated to asserted (read to write) while CAS

is negated. This method is better than changing

is asserted. The second method in-

held low (see

input is asserted

Figure 9c

WIN

from negated to asserted in a late write access because here a problem may arise with DATA IN and DATA OUT being valid at the same time. This may result in a data line trying to drive two different levels simultaneously. The page mode method of a read-modify-write access allows the user to have transceivers in the system because the data in (read data) is guaranteed to be high impedance during the time the data out (write data) is valid.

*There may be idle states inserted here by the CPU.

TL/F/11118– 12

FIGURE 9c. Read-Modify-Write Access Cycle

Page 16

4.0 Port A Access Modes (Continued)

4.5 ADDITIONAL ACCESS SUPPORT FEATURES

To support the different modes of accessing, the DP8430V/ 31V/32V offer other access features. These additional features include: Address Latches and Column Increment (for page/burst mode support), Address Pipelining, and Delay CAS

(to allow the user with a multiplexed bus to ensure

valid data is present before CAS

4.5.1 Address Latches and Column Increment

The Address Latches can be programmed, through programming bit B0. They can be programmed to either latch the address or remain in a fall-through mode. If the address latches are used to latch the address, the controller will function as follows:

In Mode 0, the rising edge of ALE places the latches in fallthrough, once ALE is negated, the address present in the row, column and bank input is latched.

is asserted).

In Mode 1, the address latches are in fall through mode until ADS

is asserted. ADS asserted latches the address.

Once the address is latched, the column address can be incremented with the input COLINC. COLINC can be used for sequential accesses of static column DRAMs. COLINC can also be used with the ECAS tial accesses to page mode DRAMs as shown in COLINC should only be asserted when the signal RFIP negated during an access since this input functions as extended refresh when RFIP negated (0) when the address is being latched (ADS edge in Mode 1). If COLINC is asserted with all of the bits of the column address asserted (ones), the column address will return to zero.

inputs to support sequen-

Figure 10

is asserted. COLINC must be

falling

FIGURE 10. Column Increment

The address latches function differently with the DP8432V. The DP8432V will latch the address of the currently granted port. If Port A is currently granted, the address will be latched as described in Section 4.5.1. If Port A is not granted, and requests an access, the address will be latched on the first or second positive edge of CLK after GRANTB has been negated depending on the programming bits R0, R1.

For Port B, if GRANTB is asserted, the address will be latched with AREQB address will latch on the first or second positive edge of CLK after GRANTB is asserted depending on the programming bits R0, R1.

asserted. If GRANTB is negated, the

TL/F/11118– 13

Page 17

4.0 Port A Access Modes (Continued)

4.5.2 Address Pipelining

Address pipelining is the overlapping of accesses to different banks of DRAM. If the majority of successive accesses are to a different bank, the accesses can be overlapped. Because of this overlapping, the cycle time of the DRAM accesses are greatly reduced. The DP8430V/31V/32V can be programmed to allow a new row address to be placed on the DRAM address bus after the column address hold time has been met. At this time, a new access can be initiated with ADS AREQ supports address pipelining for Port A only. This mode cannot be used with page, static column or nibble modes of operations because the DRAM column address is switched back to the row address after CAS programmed through address bit R8 (see

11b

or ALE, depending on the access mode, while

is used to sustain the current access. The DP8432V

is asserted. This mode is

Figures 11a

and

During address pipelining in Mode 0, shown in ALE cannot be pulsed high to start another access until AREQ

has been asserted for the previous access for at least one period of CLK. DTACK negated once AREQ insert wait states, will be asserted once ALE and CS asserted.

In Mode 1, shown in AREQ

is asserted. After meeting the minimum negated pulse width for ADS new access. DTACK AREQ

is negated. WAIT, if programmed, will be asserted once ADS

In either mode with either type of wait programmed, the DP8430V/31V/32V will still delay the access for precharge if sequential accesses are to the same bank or if a refresh takes place.

is asserted.

is negated. WAIT, if programmed to

Figure 11d

, ADS can again be asserted to start a

, if programmed, will be negated once

, if programmed, will be

, ADS can be negated once

Figure 11c

are

FIGURE 11a. Non-Address Pipelined Mode

FIGURE 11b. Address Pipelined Mode

TL/F/11118– 14

TL/F/11118– 15

Page 18

4.0 Port A Access Modes (Continued)

TL/F/11118– 16

TL/F/11118– 17

is Sampled at the ‘‘T3’’ Falling Clock Edge)

WAIT

FIGURE 11c. Mode 0 Address Pipelining (WAIT of 0, (/2T Has Been Programmed.

FIGURE 11d. Mode 1 Address Pipelining (DTACK 1(/2T Programmed, DTACK is Sampled at the ‘‘T3’’ Falling Clock Edge)

Page 19

4.0 Port A Access Modes (Continued)

4.5.3 Delay CAS

Address bit C9 asserted during programming will cause CAS to be delayed until the first positive edge of CLK after RAS is asserted when the input WIN is asserted. Delaying CAS during write accesses ensures that the data to be written to DRAM will be setup to CAS

during Write Accesses

asserting as shown in

Figures

and

12b.

12a

be present after the first positive edge of CLK, CAS delayed further with the ECAS negated during programming, read and write accesses will be treated the same (with regard to CAS

If the possibility exists that data still may not

can be

inputs. If address bit C9 is

FIGURE 12a. Mode 0 Delay CAS

FIGURE 12b. Mode 1 Delay CAS

TL/F/11118– 18

TL/F/11118– 19

Page 20

5.0 Refresh Options

The DP8430V/31V/32V support two refresh control mode options:

1. Automatic Internally Controlled Refresh.

2. Externally Controlled Refresh.

With each of the control modes above, three types of refresh can be performed.

1. All RAS

2. Staggered Refresh.

3. Error Scrubbing During All RAS Refresh.

There are two inputs, EXTNDRF and RFSH puts, RFIP also ten programming bits: R0 – 1, R9, C0 –6 and ECAS0 used to program the various types of refreshing.

Asserting the input EXTNDRF, extends the refresh cycle for a single or multiple integral periods of CLK.

The output RFIP first refresh RAS progress, RFIP before the first refresh RAS ted for the access (see

The DP8430V/31V/32V will increment the refresh address counter automatically, independent of the refresh mode used. The refresh address counter will be incremented once all the refresh RAS

Refresh.

and two out-

and RFRQ, associated with refresh. There are

is asserted one period of CLK before the

is asserted. If an access is currently in

will be asserted up to one period of CLK

, after AREQ or AREQB is nega-

Figure 13

s have been negated.

In every combination of refresh control mode and refresh type, the DP8430V/31V/32V is programmed to keep RAS asserted a number of CLK periods. The time values of RAS low during refresh are programmed through programming bits R0 and R1.

5.1 REFRESH CONTROL MODES

5.1.1. Automatic Internal Refresh

To select Automatic Internal Refresh, the user must choose

ECAS0

0 during programming. The DP8430V/31V/32V have an internal refresh clock. The period of the refresh clock is generated from the programming bits C0 –3. Every period of the refresh clock, an internal refresh request is generated. As long as a DRAM access is not currently in progress and precharge time has been met, the internal refresh request will generate an automatic internal refresh. If a DRAM access is in progress, the DP8430V/31V/32V onchip arbitration logic will wait until the access is finished before performing the refresh. The refresh/access arbitration logic can insert a refresh cycle between two address pipelined accesses. However, the refresh arbitration logic can not interrupt an access cycle to perform a refresh.

Explanation of Terms

RFRQeReFresh ReQuest internal to the DP8430V/31V/32V. RFRQ has the ability to hold off a pending access.

RFSH

Externally requested ReFreSH

RFIP

ReFresh in Progress

ACIP

Port A or Port B (DP8432V only) ACcess in Progress. This means that either RAS is low for an access or is in the process of transitioning low for an access.

TL/F/11118– 20

FIGURE 13. DP8430V/31V/32V Access/Refresh Arbitration State Program

Page 21

5.0 Refresh Options (Continued)

5.1.2 Externally Controlled Refresh Mode

To choose this refresh mode, the user must program

ECAS0 sponsible for generating refresh requests by asserting the input RFSH this refresh mode, the output WE

When Externally Controlled Refresh is selected, the user may choose to monitor the output ReFresh ReQuest

1. When this mode is selected, the user is re-

every time a refresh cycle is to be performed. In

functions a RFRQ.

(RFRQ

) for an indication from the DRAM controller that a refresh is needed. When this output asserts, it indicates that the internal refresh clock has expired and that another refresh is necessary. Then the user has two options. First, he can answer immediately asserting the input RFSH ing a refresh cycle. In this case a refresh will take place immediately if no access is in progress and precharge time for the previous access has been met. See

request-

Figure 14a

FIGURE 14a. Automatic Internal Refresh with Refresh Request (3T of RAS Low during Refresh Programmed)

Second, the user may choose not to assert the RFSH delaying the refresh until later. RFRQ assert (toggle) if additional periods of the internal refresh clock have expired and the user has not performed a refresh by asserting the input RFSH critical event, or a long access like page or static column

will go high and then

. See

Figure 14b

FIGURE 14b. Refresh Request Timing

input

. If a time

mode can not be interrupted, RFRQ used to increment an external counter. This counter can later be used to perform a burst refresh of the number of refreshes missed (through the RFSH can be thought of as Refresh Request/Acknowledge.

pulsing high can be

input). This scheme

TL/F/11118– 21

TL/F/11118– 22

Page 22

5.0 Refresh Options (Continued)

In Externally Controlled Refresh the user does not have to wait for RFRQ time assert the RFSH ternal latch that is used to produce the internal refresh request. The refresh cycle will take place on the next positive edge of clock, as shown in DRAM is in progress or precharge time for the last access has not been met, the refresh will be delayed. Since pulsing RFSH low until the refresh starts. When the last refresh RAS negates, the internal refresh request latch is cleared.

to perform a refresh. The user can at any

input. Pulsing RFSH low, sets an in-

Figure 15a

. If an access to the

low sets a latch, the user doesn’t have to keep RFSH

By keeping the input RFSH of CLK which ends the refresh cycle as shown in

15b

, the user will perform another refresh cycle. Each re-

asserted past the positive edge

Figure

fresh cycle during a burst refresh will meet the refresh RAS low time and the RAS precharge time (programming bits R0–1). This scheme can be thought of as Externally Controlled Burst Refresh. If the user desires to burst refresh the entire DRAM, he could generate an end of count signal (burst refresh finished), by looking at one of the DP8430V/ 31V/32V high address outputs (Q7, Q8, Q9 or Q10) and the RFIP

output. The Qn outputs function as a decode of how many row addresses have been refreshed. (Q7 freshes, Q8

256 refreshes, Q9e512 refreshes and Q10

1024 refreshes).

128 re-

FIGURE 15a. Single Externally Refreshes (2 Periods of RAS Low during Refresh Programmed)

FIGURE 15b. External Burst Refresh

(2 Periods of RAS

Precharge, 2 Periods of Refresh RAS Low during Refresh Programmed)

TL/F/11118– 23

TL/F/11118– 24

Page 23

5.0 Refresh Options (Continued)

5.2 REFRESH CYCLE TYPES

Three different types of refresh cycles are available for use. The three different types are mutually exclusive and can be used with any of the three modes of refresh control. The three different refresh cycle types are: all RAS gered RAS fresh. In all refresh cycle types, the RAS guaranteed: between the previous access RAS the refresh RAS and access RAS

refresh and error scrubbing during all RAS re-

0 starting; between refresh RAS3 ending

beginning; between burst refresh RASs.

refresh, stag-

precharge time is

ending and

5.2.1 Conventional RAS

A conventional refresh cycle causes RAS0– 3 to all assert from the first positive edge of CLK after RFIP shown in number of positive edges of CLK programmed have passed. On the last positive edge, RAS ed. This type of refresh cycle is programmed by negating address bit R9 during programming.

Figure 16

Refresh

is asserted as

. RAS0 –3 will stay asserted until the

0–3, and RFIP will be negat-

FIGURE 16. Conventional RAS Refresh

5.2.2 Staggered RAS

A staggered refresh staggers each RAS or group of RASs by a positive edge of CLK as shown in ber of RAS of CLK, is determined by the RAS programming bits C4 – C6. If single RAS ed during programming, then each RAS cessive positive edges of CLK. If two RAS lected during programming then RAS

s, which will be asserted on each positive edge

Refresh

Figure 17

, CAS configuration mode

0 and RAS1 will assert

. The num-

outputs are select-

will assert on suc-

outputs are se-

FIGURE 17. Staggered RAS Refresh

TL/F/11118– 25

on the first positive edge of CLK after RFIP RAS

2 and RAS3 will assert on the second positive edge of CLK after RFIP during programming, all RAS first positive edge of CLK after RFIP or group of RASs will meet the programmed RAS low time and then negate.

is asserted. If all RAS outputs were selected

outputs would assert on the

is asserted. Each RAS

is asserted.

TL/F/11118– 26

Page 24

5.0 Refresh Options (Continued)

5.2.3 Error Scrubbing during Refresh

The DP8430V/31V/32V support error scrubbing during all RAS

DRAM refreshes. Error scrubbing during refresh is selected through bits C4 – C6 with bit R9 negated during programming. Error scrubbing can not be used with staggered refresh (see Section 8.0). Error scrubbing during refresh allows a CAS refresh as shown in from the DRAM array and passed through an Error Detection And Correction Chip, EDAC. If the EDAC determines that the data contains a single bit error and corrects that error, the refresh cycle can be extended with the input ex-

or group of CASs to assert during the all RAS

Figure 18

. This allows data to be read

tend refresh, EXTNDRF, and a read-modify-write operation can be performed by asserting WE the designer to ensure that WE has a 24-bit internal refresh address counter that contains the 11 row, 11 column and 2 bank addresses. The DP8430V/31V have a 22-bit internal refresh address counter that contains the 10 row, 10 column and 2 bank addresses. These counters are configured as bank, column, row with the row address as the least significant bits. The bank counter bits are then used with the programming selection to determine which CAS a refresh.

or group of CASs will assert during

. It is the responsibility of

is negated. The DP8432V

FIGURE 18. Error Scrubbing during Refresh (Two Refresh Cycles Shown)

TL/F/11118– 27

Page 25

5.0 Refresh Options (Continued)

5.3 EXTENDING REFRESH

The programmed number of periods of CLK that refresh RAS

s are asserted can be extended by one or multiple periods of CLK. Only the all RAS bing) type of refresh can be extended. To extend a refresh cycle, the input extend refresh, EXTNDRF, must be asserted before the positive edge of CLK that would have negated

(with or without error scrub-

all the RAS positive edge of CLK which starts all RAS refresh as shown in the next positive edge of CLK and EXTNDRF will be sampled again. The refresh cycle will continue until EXTNDRF is sampled low on a positive edge of CLK.

outputs during the refresh cycle and after the

Figure 19

. This will extend the refresh to

outputs during the

FIGURE 19. Extending Refresh with the Extend Refresh (EXTNDRF) Input

6.0 Port A Wait State Support

Wait states allow a CPU’s access cycle to be increased by one or multiple CPU clock periods. The wait or ready input is named differently by CPU manufacturers. However, any CPU’s wait or ready input is compatible with either the WAIT or DTACK output of the DP8430V/31V/32V. The user determines whether to program WAIT which value to select for WAIT ing upon the CPU used and where the CPU samples its wait input during an access cycle.

The decision to terminate the CPU access cycle is directly affected by the speed of the DRAMs used. The system designer must ensure that the data from the DRAMs will be present for the CPU to sample or that the data has been written to the DRAM before allowing the CPU access cycle to terminate.

The insertion of wait states also allows a CPU’s access cycle to be extended until the DRAM access has taken place. The DP8430V/31V/32V insert wait states into CPU access cycles due to; guaranteeing precharge time, refresh currently in progress, user programmed wait states, the WAITIN signal being asserted and GRANTB not being valid (DP8432V only). If one of these events is taking place and the CPU starts an access, the DP8430V/31V/32V will insert wait states into the access cycle, thereby increasing the

or DTACK (R7) and

or DTACK (R2, R3) depend-

TL/F/11118– 28

length of the CPU’s access. Once the event has been completed, the DP8430V/31V/32V will allow the access to take place and stop inserting wait states.

There are six programming bits, R2 –R7; an input, WAITIN and an output that functions as WAIT

6.1 WAIT

With the R7 address bit negated during programming, the user selects the WAIT asserted by the CPU, wait states (extra clock periods) are inserted into the current access cycle as shown in

completed by the CPU. WAIT a chip selected access and is programmed to negate a number of positive edges and/or negative levels of CLK from the event that starts the access. WAIT programmed to function in page/burst mode applications. Once WAIT inputs are negated with AREQ asserted, WAIT can be programmed to toggle, following the ECAS is negated, ending the access, WAIT will stay negated until the next chip selected access. For more details about WAIT Type Output, see Application Note AN-773.

TYPE OUTPUT

output. As long as WAIT is sampled

. Once WAIT is sampled negated, the access cycle is

is negated during an access, and the ECAS

or DTACK.

Figure

is asserted at the beginning of

can also be

inputs. Once AREQ

;

FIGURE 20. WAIT Type Output

TL/F/11118– 29

Page 26

6.0 Port A Wait State Support (Continued)

6.2 DTACK

With the R7 address bit asserted during programming, the user selects the DTACK sampled negated by the CPU, wait states are inserted into the current access cycle as shown in DTACK by the CPU. DTACK grammed to assert a number of positive edges and/or negative levels from the event that starts RAS DTACK burst mode accesses. Once DTACK ECAS be programmed to negate and assert from the ECAS toggling to perform a page/burst mode operation. Once AREQ ed and stays negated until the next chip selected access. For more details about DTACK Note AN-773.

TYPE OUTPUT

type output. As long as DTACK is

is sampled asserted, the access cycle is completed

Figure 21.

, which is normally negated, is pro-

for the access.

can also be programmed to function during page/

is asserted and the

inputs are negated with AREQ asserted, DTACK can

is negated, ending the access, DTACK will be negat-

type output see Application

Once

inputs

6.3 DYNAMICALLY INCREASING THE NUMBER OF WAIT STATES

The user can increase the number of positive edges of CLK before DTACK input WAITIN or WAIT CLK. The number of edges is programmed through address bit R6. If the user is increasing the number of positive edges in a delay that contains a negative level, the positive edges will be met before the negative level. For example if the user programmed DTACK grammed as 2T, would increase the number of positive edges resulting in DTACK larly, WAITIN a page/burst access. WAITIN in systems requiring an increased number of wait states. WAITIN the type of access. As an example, a user could invert the WRITE WAITIN 1 wait state and read accesses with 2 wait states as shown in

Figure 22b

is asserted or WAIT is negated. With the

asserted, the user can delay DTACK asserting

negating either one or two more positive edges of

of (/2T, asserting WAITIN, pro-

of 2(/2T as shown in

can increase the number of positive edges in

can be permanently asserted

can also be asserted and negated, depending on

line from the CPU and connect the output to

. This could be used to perform write accesses with

Figure 22a

. Simi-

FIGURE 21. DTACK Type Output

FIGURE 22a. WAITIN Example (DTACK is Sampled at the ‘‘T3’’ Falling Clock Edge)

TL/F/11118– 30

TL/F/11118– 31

Page 27

6.0 Port A Wait State Support (Continued)

FIGURE 22b. WAITIN Example (WAIT is Sampled at the End of ‘‘T2’’).

6.4 GUARANTEEING RAS AND RAS

The DP8430V/31V/32V will guarantee RAS precharge time between accesses; between refreshes; and between access and refreshes. The programming bits R0 and R1 are used to program combinations of RAS RAS low time is programmed for refreshes only. During an access, the system designer guarantees the time RAS serted through the DP8430V/31V/32V wait logic. Since inserting wait states into an access increases the length of the CPU signals which are used to create ADS AREQ

The precharge time is also guaranteed by the DP8430V/ 31V/32V. Each RAS

PRECHARGE TIME

low time referenced by positive edges of CLK. RAS

, the time that RAS is asserted can be guaranteed.

LOW TIME

precharge time and

is as-

or ALE and

output has a separate positive edge

TL/F/11118– 32

of CLK counter. AREQ of CLK to terminate the access. That positive edge is 1T. The next positive edge is 2T. RAS the programmed number of positive edges of CLK have passed as shown in charge time has been met, RAS positive edge of CLK. However, since there is a precharge counter per RAS delayed. Precharge time before a refresh is always referenced from the access RAS refresh asserting. After a refresh, precharge time is referenced from RAS RAS

asserting.

is negated setup to a positive edge

will not be asserted until

Figure 23

, an access using another RAS will not be

3 negating, for the refresh, to the access

. Once the programmed pre-

will be asserted from the

negating before RAS0 for the

FIGURE 23. Guaranteeing RAS Precharge (DTACK is Sampled at the ‘‘T2’’ Falling Clock Edge)

TL/F/11118– 33

Page 28

7.0 RAS and CAS Configuration Modes

The DP8430V/31V/32V allow the user to configure the DRAM array to contain one, two or four banks of DRAM. Depending on the functions used, certain considerations must be used when determining how to set up the DRAM array. Programming address bits C4, C5 and C6 along with bank selects, B0 –1, and CAS mine which RAS of CAS

s will be asserted during an access. Different memory schemes are described. The DP8430V/31V/32V is specified driving a heavy load of 72 DRAMs, representing four banks of DRAM with 16-bit words and 2 parity bits. The DP8430V/31V/32V can drive more than 72 DRAMs, but the AC timing must be increased. Since the RAS puts are configurable, all RAS used for the maximum amount of drive.

or group of RASs and which CAS or group

enables, ECAS0–3, deter-

and CAS out-

and CAS outputs should be

7.1 BYTE WRITING

By selecting a configuration in which all CAS selected during an access, the ECAS or group of CAS size of up to 32 bits. In this case, the RAS to select which of up to 4 banks is to be used as shown in

Figures 24a

the byte enables can be gated with a high order address bit to produce four byte enables which gives an equivalent to 8 banks of 16-bit words as shown in ry is required, each CAS in the 16-bit word as shown in

outputs to select a byte (or bytes) in a word

and

24b

. In systems with a word size of 16 bits,

should be used to drive each nibble

inputs enable a single

Figure 24d

Figure 24c

outputs are

outputs are used

. If less memo-

FIGURE 24a. DRAM Array Setup for 32-Bit System (C6, C5, C4e1, 1, 0 during Programming)

FIGURE 24b. DRAM Array Setup for 32-Bit, 1 Bank System (C6, C5, C4e0, 0, 0 Allowing Error Scrubbing

or C6, C5, C4

0, 1, 1 No Error Scrubbing during Programming)

TL/F/11118– 34

TL/F/11118– 35

Page 29

7.0 RAS and CAS Configuration Modes (Continued)

FIGURE 24c. DRAM Array Setup for 16-Bit System (C6, C5, C4e1, 1, 0 during Programming)

TL/F/11118– 36

FIGURE 24d. 8 Bank DRAM Array for 16-Bit System (C6, C5, C4e1, 1, 0 during Programming)

TL/F/11118– 37

Page 30

7.0 RAS and CAS Configuration Modes (Continued)

7.2 MEMORY INTERLEAVING

Memory interleaving allows the cycle time of DRAMs to be reduced by having sequential accesses to different memory banks. Since the DP8430V/31V/32V have separate precharge counters per bank, sequential accesses will not be delayed if the accessed banks use different RAS To ensure different RAS selected where either one or two RAS serted during an access. The bank select or selects, B0 and B1, are then tied to the least significant address bits, causing a different group of RAS tial access as shown in be at least one clock period of all RAS different RAS CAS

’s being asserted to avoid the condition of a

before RAS refresh cycle.

outputs will be used, a mode is

s to assert during each sequen-

Figure 25

outputs will be as-

. In this figure there should

’s negated between

outputs.

7.3 ADDRESS PIPELINING

Address pipelining allows several access RAS serted at once. Because RAS quires either a mode where one RAS per bank as shown in two CAS writing can be accomplished in a 16-bit word system if two RAS

s and two CASs are used per bank. In other systems,

s (or external gating on the CAS outputs) must be used to perform byte writing. If WE and data out buffers must be used. If the array is not layed out this way, a CAS will cause a refresh of the DRAM, not an access. To take full advantage of address pipelining, memory interleaving is used. To memory interleave, the least significant address bits should be tied to the bank select inputs to ensure that all ‘‘back to back’’ sequential accesses are not delayed, since different memory banks are accessed.

s can overlap, each bank re-

s are used per bank as shown in

Figure 26a

to a bank can be low before RAS, which

and one CAS are used

or where two RASs and

s are used separate data in

stobeas-

Figure 26b

. Byte

FIGURE 25. Memory Interleaving (C6, C5, C4e1, 1, 0 during Programming)

TL/F/11118– 38

Page 31

7.0 RAS and CAS Configuration Modes (Continued)

FIGURE 26a. DRAM Array Setup for 4 Banks Using Address Pipelining (C6, C5, C4e1, 1, 1

or C6, C5, C4

FIGURE 26b. DRAM Array Setup for Address Pipelining with 2 Banks (C6, C5, C4e1, 0, 1

or C6, C5, C4

0, 1, 0 (Also Allowing Error Scrubbing) during Programming)

0, 0, 1 (Also Allowing Error Scrubbing) during Programming)

7.4 ERROR SCRUBBING

In error scrubbing during refresh, the user selects one, two or four RAS

and CAS outputs per bank. When performing

error detection and correction, memory is always accessed

as words. Since the CAS individual bytes, the ECAS in

Figures 27a

and

27b

TL/F/11118– 39

TL/F/11118– 40

signals are not used to select

inputs can be tied low as shown

FIGURE 27a. DRAM Array Setup for 4 Banks Using Error Scrubbing (C6, C5, C4e0, 1, 0 during Programming)

TL/F/11118– 41

FIGURE 27b. DRAM Array Setup for Error Scrubbing with 2 Banks (C6, C5, C4e0, 0, 1 during Programming)

TL/F/11118– 42

Page 32

7.0 RAS and CAS Configuration Modes (Continued)

7.5 PAGE/BURST MODE

In a static column, page or burst mode system, the least significant bits must be tied to the column address in order to ensure that the page/burst accesses are to sequential memory addresses, as shown in

Figure 28.

In a nibble mode system, the least significant bits must be tied to the highest column and row address bits in order to ensure that sequential address bits are the ‘‘nibble’’ bits for nibble mode accesses

*See table below for row, column & bank address bit map. A0, A1 are used for byte addressing in this example.

(Figure 28)

. The ECAS inputs may then be tog-

Addresses Nibble Mode*

Column C9,R9eA2,A3 C0– 7eA2–9 C0–8eA2–10 C0–9eA2–11

Address C0 –8

Row

Address

B0 A4 A10 A11 A12 A13 B1 A5 A11 A12 A13 A14

Assume that the least significant address bits are used for byte addressing. Given a 32-bit system A0,A1 would be used for byte addressing.

DON’T CARE, the user can do as he pleases.

*Nibble mode values for R and C assume a system using 1 Mbit DRAMs.

XXX X X

256 Bits/Page 512 Bits/Page 1024 Bits/Page 2048 Bits/Page

X C8–10eX C9,10eX C10eX

FIGURE 28. Page, Static Column, Nibble Mode System

gled with the DP8430V/31V/32V’s address latches in fallthrough mode, while AREQ can also be used to select individual bytes. When using nibble mode DRAMS, the third and fourth address bits can be tied to the bank select inputs to perform memory interleaving. In page or static column modes, the two address bits after the page size can be tied to the bank select inputs to select a new bank if the page size is exceeded.

Page Mode/Static Column Mode Page Size

is asserted. The ECAS inputs

TL/F/11118– 43

C0–10

A2–12

Page 33

8.0 Test Mode

Staggered refresh in combination with the error scrubbing mode places the DP8430V/31V/32V in test mode. In this mode, the 24-bit refresh counter is divided into a 13-bit and 11-bit counter. During refreshes both counters are incremented to reduce test time.

9.0 DRAM Critical Timing Parameters

The two critical timing parameters, shown in must be met when controlling the access timing to a DRAM are the row address hold time, t dress setup time, t tain a precise internal delay line, the values of these param-

. Since the DP8430V/31V/32V con-

ASC

RAH

eters can be selected at programming time. These values will also increase and decrease if DELCLK varies from 2 MHz.

9.1 PROGRAMMABLE VALUES OF t

The DP8430V/31V/32V allow the values of t to be selected at programming time. For each parameter, two choices can be selected. t time, is measured from RAS

RAH

asserted to the row address starting to change to the column address. The two choices for t

are 15 ns and 25 ns, programmable through ad-

RAH

dress bit C8.

, the column address setup time, is measured from the

ASC

column address valid to CAS t

are 0 ns and 10 ns, programmable through address bit

ASC

C7.

asserted. The two choices for

Figure 29

, that

, and the column ad-

AND t

RAH

ASC

and t

ASC

, the row address hold

9.2 CALCULATION OF t

RAH

AND t

ASC

There are two clock inputs to the DP8430V/31V/32V. These two clocks, DELCLK and CLK can either be tied together to the same clock or be tied to different clocks running asynchronously at different frequencies.

The clock input, DELCLK, controls the internal delay line and refresh request clock. DELCLK should be a multiple of 2 MHz. If DELCLK is not a multiple of 2 MHz, t will change. The new values of t lated by the following formulas:

If t

was programmed to equal 15 ns then t

RAH

15*(((DELCLK Divisor)* 2 MHz/(DELCLK Frequency))b1)

15 ns.

was programmed to equal 25 ns then t

If t

RAH

25*(((DELCLK Divisor)* 2 MHz/(DELCLK Frequency))b1)

25 ns.

If t

was programmed to equal 0 ns then t

ASC

((DELCLK Divisor)* 2 MHz/(DELCLK Frequency))

RAH

and t

RAH

can be calcu-

ASC

and t

RAH

ASC

12.5*

12.5 ns.

If t

was programmed to equal 10 ns then t

ASC

((DELCLK Divisor)* 2 MHz/(DELCLK Frequency))

ASC

22.5*

12.5 ns.

Since the values of t creased, the time to CAS

and t

RAH

asserted will also increase or de-

are increased or de-

ASC

crease. These parameters can be adjusted by the following formula:

Delay to CAS Programmed t

Actual Spec.aActual t

Actual t

RAH

ASC

RAH

Programmed t

ASC

FIGURE 29. t

RAH

and t

ASC

TL/F/11118– 44

Page 34

10.0 Dual Accessing (DP8432V)

The DP8432V has all the functions previously described. In addition to those features, the DP8432V also has the capabilities to arbitrate among refresh, Port A and a second port, Port B. This allows two CPUs to access a common DRAM array. DRAM refresh has the highest priority followed by the currently granted port. The ungranted port has the lowest priority. The last granted port will continue to stay granted even after the access has terminated, until an access request is received from the ungranted port (see The dual access configuration assumes that both Port A and Port B are synchronous to the system clock. If they are not synchronous to the system clock they should be externally synchronized (Ex. By running the access requests through several Flip-Flops, see

Figure 32a

10.1 PORT B ACCESS MODE

Port B accesses are initiated from a single input, AREQB When AREQB

is asserted, an access request is generated. If GRANTB is asserted and a refresh is not taking place or precharge time is not required, RAS AREQB

is asserted. Once AREQB is asserted, it must stay asserted until the access is over. AREQB RAS

as shown in programming the CAS yond RAS ate ECAS

Figure 30b

. Note that if ECAS0e1 during

outputs may be held asserted (be-

n negating) by continuing to assert the appropri-

n inputs (the same as Port A accesses). If Port B is not granted, the access will begin on the first or second positive edge of CLK after GRANTB is asserted (See R0, R1 programming bit definitions) as shown in suming that Port A is not accessing the DRAM (CS ALE and AREQ

) and RAS precharge for the particular bank

Figure 30a

will be asserted when

negated, negates

Figure 30c

, as-

, ADS/

has completed. It is important to note that for GRANTB to transition to Port B, Port A must not be requesting an access at a rising clock edge (or locked) and Port B must be requesting an access at that rising clock edge. Port A can request an access through CS AREQ

. Therefore during an interleaved access where CS

and ADS/ALE or CS and

and ADS/ALE become asserted before AREQ from the previous access is negated, Port A will retain GRANTB whether AREQB

is asserted or not.

Since there is no chip select for Port B, AREQB must incorporate this signal. This mode of accessing is similar to Mode 1 accessing for Port A.

Explanation of Terms

AREQA

AREQB

LOCK

Chip Selected access request from Port A

Chip Selected access request from Port B

Externally controlled LOCKing of the Port that is currently GRANTed.

FIGURE 30a. DP8432V Port A/Port B Arbitration

State Diagram. This arbitration may take place

during the ‘‘ACCESS’’ or ‘‘REFRESH’’

state (see

Figure 13

TL/F/11118– 45

FIGURE 30b. Access Request for Port B

FIGURE 30c. Delayed Port B Access

TL/F/11118– 46

TL/F/11118– 47

Page 35

10.0 Dual Accessing (DP8432V) (Continued)

10.2 PORT B WAIT STATE SUPPORT

Advanced transfer acknowledge for Port B, ATACKB used for wait state support for Port B. This output will be asserted when RAS shown in

Figures 31a

will stay asserted until AREQB logic, ATACKB input as shown in

for the Port B access is asserted, as

and

31b

. Once asserted, this output

is negated. With external

can be made to interface to any CPU’s wait

Figure 31c

10.3 COMMON PORT A AND PORT B DUAL PORT FUNCTIONS

An input, LOCK

, and an output, GRANTB, add additional

functionality to the dual port arbitration logic. LOCK

,is

allows

Port A or Port B to lock out the other port from the DRAM. When a Port is locked out of the DRAM, wait states will be inserted into its access cycle until it is allowed to access memory. GRANTB is used to multiplex the input control signals and addresses to the DP8432V.

10.3.1 GRANTB Output

The output GRANTB determines which port has current access to the DRAM array. GRANTB asserted signifies Port B has access. GRANTB negated signifies Port A has access to the DRAM array.

FIGURE 31a. Non-Delayed Port B Access

FIGURE 31b. Delayed Port B Access

B. Extend ATACK to 1T after RAS goes low.

TL/F/11118– 50

A. Extend ATACK to (/2T((/2 Clock) after RAS goes low.

C. Synchronize ATACKB to CPU B Clock. This is useful if CPU B runs asynchronous to the DP8432.

FIGURE 31c. Modifying Wait Logic for Port B

TL/F/11118– 48

TL/F/11118– 49

TL/F/11118– 51

TL/F/11118– 52

Page 36

10.0 Dual Accessing (DP8432V) (Continued)

Since the DP8432V has only one set of address inputs, the signal is used, with the addition of buffers, to allow the currently granted port’s addresses to reach the DP8432V. The signals which need to be bufferred are R0 –10, C0 – 10, B0–1, ECAS common and do not have to be buffered as shown in

32a.

0–3, WE, and LOCK. All other inputs are not

Figure

If a Port, which is not currently granted, tries to access

the DRAM array, the GRANTB output will transition from a rising clock edge from AREQ precede the RAS ods. GRANTB will then stay in this state until the other port requests an access and the currently granted port is not accessing the DRAM as shown in

for the access by one or two clock peri-

or AREQB negating and will

Figure 32b

*If Port B is synchronous the Request Synchronizing logic will not be required.

FIGURE 32a. Dual Accessing with the DP8432V (System Block Diagram)

TL/F/11118– 53

Page 37

10.0 Dual Accessing (DP8432V) (Continued)

FIGURE 32b. Wait States during a Port B Access

10.3.2 LOCK

When the LOCK can ‘‘lock out’’ the other port through the insertion of wait states to that port’s access cycle. LOCK

Input

input is asserted, the currently granted port

does not disable

FIGURE 33. LOCK Function

refreshes, it only keeps GRANTB in the same state even if the other port requests an access, as shown in LOCK

can be used by either port.

TL/F/11118– 54

Figure 33

TL/F/11118– 55

Page 38

11.0 Absolute Maximum Ratings (Note 1)

If Military/Aerospace specified devices are required, please contact the National Semiconductor Sales Office/Distributors for availability and specifications.

Temperature under Bias ААААААААААААААААААА0

Storage Temperature АААААААААААААААААb65§Ctoa150§C

Ctoa70§C

All Input or Output Voltage

with Respect to GNDААААААААААААААААААА

0.5V toa7V

Power Dissipation@20 MHzААААААААААААААААААААААА0.5W

ESD RatingААААААААААААААААААААААААААААААААААААА2000V

Temperature Cycle АААААААААААААААААААААА300 of 0/125§C

12.0 DC Electrical Characteristics T

0§Ctoa70§C, V

5Vg10%, GNDe0V

Symbol Parameter Conditions Min Typ Max Units

OH1

OL1

OH2

OL2

IL ML

CC1

CC2

Logical 1 Input Voltage Tested with a Limited

Functional Pattern

Logical 0 Input Voltage Tested with a Limited

Functional Pattern

Q and WE Outputs I

All Outputs except Qs, WE I

Input Leakage Current V

ML Input Current (Low) V

Standby Current CLK at 12 MHz (V

Standby Current CLK at 20 MHz (V

Standby Current CLK at 33 MHz (V

10 mA V

10 mA 0.5 V

3mA V

3 mA 0.5 V

VCCor GND

GND 200 mA

VCCor GND) 6 15 mA

VCCor GND) 8 17 mA

VCCor GND) 10 20 mA

Supply Current CLK at 12 MHz (Inputs Active)

LOAD

25 pF) (V

VCCor GND)

Supply Current CLK at 20 MHz (Inputs Active)

LOAD

25 pF) (V

VCCor GND)

Supply Current CLK at 33 MHz (Inputs Active)

LOAD

25 pF) (V

VCCor GND)

2.0 V

0.5 0.8 V

1.0 V

10 10 mA

0.5 V

30 60 mA

65 90 mA

115 150 mA

CIN* Input Capacitance fINat 1 MHz 10 pF

*CINis not 100% tested.

Note 1: ‘‘Absolute Maximum Ratings’’ are those values beyond which the safety of the device cannot be guaranteed. They are not meant to imply that the devices should be operated at these limits. The tables of ‘‘Electrical Characteristics’’ provide conditions for actual device operation.

Note 2: Input pulse 0V to 3V; t

Note 3: AC production testing is done at 50 pF.

2.5 ns. Input reference point on AC measurements is 1.5V. Output reference point is 1.5V.

Page 39

13.0 AC Timing Parameters

Two speed selections are given, the DP8430V/31V/32V-20 and the DP8430V/31V/32V-33. The differences between the two parts are the maximum operating frequencies of the input CLKs and the maximum delay specifications. Low frequency applications may use the ‘‘-25’’ part to gain improved timing.

The AC timing parameters are grouped into sectional numbers as shown below. These numbers also refer to the timing diagrams.

1–36 Common parameters to all modes of operation

50–56 Difference parameters used to calculate;

RAS

low time,

RAS

precharge time,

CAS

high time and

CAS

low time

100–121 Common dual access parameters used for Port

B accesses and inputs and outputs used only in dual accessing

200–212 Refresh parameters

Note 1: ‘‘Absolute Maximum Ratings’’ are the values beyond which the safety of the device cannot be guaranteed. They are not meant to imply that the device should be operated at these limits. The table of ‘‘Electrical Characteristics’’ provides conditions for actual device operation.

Note 2: Input pulse 0V to 3V; tR

Note 3: AC Production testing is done at 50 pF.

etFe

2.5 ns. Input reference point on AC measurements is 1.5V. Output reference points are 2.4V for High and 0.8V for Low.

300–315 Mode 0 access parameters used in both single

and dual access applications

400–416 Mode 1 access parameters used in both single

and dual access applications

450–455 Special Mode 1 access parameters which super-

sede the 400 –416 parameters when dual accessing

500–506 Programming parameters

Unless otherwise stated V 70§C, the output load capacitance is typical for 4 banks of

5.0Vg10%, 0kT

18 DRAMs per bank, including trace capacitance (see Note

2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

FIGURE 34. Clock, DELCLK Timing

TL/F/11118– 56

Page 40

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

5.0Vg10%, 0§CkT

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

DP8430V/31V/32V-33

Number Symbol

Common Parameter

Description

Min Max Min Max

CLK

CLK Frequency 0 33 0 33

2 tCLKP CLK Period 30 30

3, 4 tCLKPW CLK Pulse Width 12 12

5 fDCLK DELCLK Frequency 12 40 12 40

6 tDCLKP DELCLK Period 25 83 25 83

7, 8 tDCLKPW DELCLK Pulse Width 12 12

9a tPRASCAS0 RAS Asserted to CAS Asserted

(tRAH

15 ns, tASCe0 ns)

9b tPRASCAS1 RAS Asserted to CAS Asserted

(tRAHe15 ns, tASCe10 ns)

9c tPRASCAS2 (RAS Asserted to CAS Asserted

(tRAH

25 ns, tASCe0 ns)

9d tPRASCAS3 (RAS Asserted to CAS Asserted

(tRAH

25 ns, tASCe10 ns)

30 30

40 40

50 50

10a tRAH Row Address Hold Time (tRAHe15) 15 15

10b tRAH Row Address Hold Time (tRAHe25) 25 25

11a tASC Column Address Setup Time (tASCe0) 0 0

11b tASC Column Address Setup Time (tASCe10) 10 10

12 tPCKRAS CLK High to RAS

following Precharge

Asserted

18 22

13 tPARQRAS AREQ Negated to RAS Negated 25 29

14 tPENCL ECAS0–3 Asserted to CAS Asserted 15 22

15 tPENCH ECAS0– 3 Negated to CAS Negated 14 21

16 tPARQCAS AREQ Negated to CAS Negated 36 43

17 tPCLKWH CLK to WAIT Negated 25 25

18 tPCLKDL0 CLK to DTACK Asserted

(Programmed as DTACK of 1/2, 1, 1(/2 23 23 or if WAITIN is Asserted)

19 tPEWL ECAS Negated to WAIT Asserted

during a Burst Access

29 29

20 tSECK ECAS Asserted Setup to CLK High to

Recognize the Rising Edge of CLK 13 13 during a Burst Access

Page 41

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

5.0Vg10%, 0§CkT

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

DP8430V/31V/32V-33

Number Symbol

Common Parameter

Description

Min Max Min Max

21 tPEDL ECAS Asserted to DTACK

Asserted during a Burst Access 29 29 (Programmed as DTACK

22 tPEDH ECAS Negated to DTACK

Negated during a Burst Access

30 30

23 tSWCK WAITIN Asserted Setup to CLK 5 5

24 tPWINWEH WIN Asserted to WE Asserted 18 28

25 tPWINWEL WIN Negated to WE Negated 18 28

26 tPAQ Row, Column Address Valid to

Q0–8, 9, 10 Valid

27 tPCINCQ COLINC Asserted to Q0–8, 9, 10

Incremented

28 tSCINEN COLINC Asserted Setup to ECAS

Asserted to Ensure tASC

0ns

29a tSARQCK1 AREQ, AREQB Negated Setup to CLK

High with 1 Period of Precharge

29b tSARQCK2 AREQ, AREQB Negated Setup to CLK High

withl1 Period of Precharge Programmed

14 16

25 25

11 11

20 29

24 33

30 tPAREQDH AREQ Negated to DTACK Negated 20 20

31 tPCKCAS CLK High to CAS Asserted

when Delayed by WIN

32 tSCADEN Column Address Setup to ECAS

Asserted to Guarantee tASCe0

10 11

21 28

33 tWCINC COLINC Pulse Width 10 10

34a tPCKCL0 CLK High to CAS Asserted following

Precharge (tRAH

15 ns, tASCe0 ns)

34b tPCKCL1 CLK High to CAS Asserted following

Precharge (tRAHe15 ns, tASCe10 ns)

34c tPCKCL2 CLK High to CAS Asserted following

Precharge (tRAH

25 ns, tASCe0 ns)

34d tPCKCL3 CLK High to CAS Asserted following

Precharge (tRAH

25 ns, tASCe10 ns)

35 tCAH Column Address Hold Time

(Interleave Mode Only)

36 tPCQR CAS Asserted to Row Address

Valid (Interleave Mode Only)

25 25

65 73

75 83

85 93

70 70

Page 42

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

5.0Vg10%, 0§CkT

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

Number Symbol

Difference

Parameter Description

50 tD1 (AREQ or AREQB Negated to RAS

Negated) Minus (CLK High to RAS 99 Asserted)

51 tD2 (CLK High to Refresh RAS Negated)

Minus (CLK High to RAS Asserted)

52 tD3a (ADS Asserted to RAS Asserted

(Mode 1)) Minus (AREQ to RAS

Negated)

Negated 4 4

53 tD3b (CLK High to RAS Asserted (Mode 0))

Minus (AREQ Negated to RAS Negated)

54 tD4 (ECAS Asserted to CAS Asserted)

Minus (ECAS

Negated to CAS Negated)

55 tD5 (CLK to Refresh RAS Asserted) Minus

(CLK to Refresh RAS

Negated)

56 tD6 (AREQ Negated to RAS Negated)

Minus (ADS

Asserted to RAS 55

Asserted (Mode 1))

DP8430V/31V/32V-33

Min Max Min Max

Page 43

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

Number Symbol

5.0Vg10%, 0§CkT

Common Dual Access

Parameter Description

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

DP8430V/31V/32V-33

Min Max Min Max

100 tHCKARQB AREQB Negated Held from CLK High 2 2

101 tSARQBCK AREQB Asserted Setup to CLK High 5 5

102 tPAQBRASL AREQB Asserted to RAS Asserted 29 33

103 tPAQBRASH AREQB Negated to RAS Negated 24 28

105 tPCKRASG CLK High to RAS Asserted for

Pending Port B Access

37 41

106 tPAQBATKBL AREQB Asserted to ATACKB Asserted 37 37

107 tPCKATKB CLK High to ATACKB Asserted

for Pending Access

45 45

108 tPCKGH CLK High to GRANTB Asserted 28 28

109 tPCKGL CLK High to GRANTB Negated 26 26

110 tSADDCKG Row Address Setup to CLK High That

Asserts RAS Change to Ensure tASR

111 tSLOCKCK LOCK Asserted Setup to CLK Low

to Lock Current Port

following a GRANTB 7 11

0 ns for Port B

112 tPAQATKBH AREQ Negated to ATACKB Negated 16 16

113 tPAQBCASH AREQB Negated to CAS Negated 38 45

114 tSADAQB Address Valid Setup to

AREQB

Asserted

610

116 tHCKARQG AREQ Negated Held from CLK High 5 5

117 tWAQB AREQB High Pulse Width

to Guarantee tASR

0ns

118a tPAQBCAS0 AREQB Asserted to CAS Asserted

(tRAHe15 ns, tASCe0 ns)

118b tPAQBCAS1 AREQB Asserted to CAS Asserted

118c tPAQBCAS2 AREQB Asserted to CAS Asserted

118d tPAQBCAS3 AREQB Asserted to CAS Asserted

(tRAH

15 ns, tASCe10 ns)

25 ns, tASCe0 ns)

25 ns, tASCe10 ns)

17 19

79 86

89 96

99 106

120a tPCKCASG0 CLK High to CAS Asserted

for Pending Port B Access 84 91

(tRAH

15 ns, tASCe0 ns)

120b tPCKCASG1 CLK High to CAS Asserted

for Pending Port B Access 94 101

(tRAH

15 ns, tASCe10 ns)

120c tPCKCASG2 CLK High to CAS Asserted

for Pending Port B Access 94 101 (tRAHe25 ns, tASCe0 ns)

120d tPCKCASG3 CLK High to CAS Asserted

for Pending Port B Access 104 111

(tRAH

25 ns, tASCe10 ns)

121 tSBADDCKG Bank Address Valid Setup to CLK

High That Starts RAS 55 for Pending Port B Access

Page 44

13.0 AC Timing Parameters (Continued)

FIGURE 35. 100: Dual Access Port B

FIGURE 36. 100: Port A and Port B Dual Access

TL/F/11118– 57

TL/F/11118– 58

Page 45

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

5.0Vg10%, 0§CkT

Number Symbol

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

Refresh Parameter

Description

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

DP8430V/31V/32V-33

Min Max Min Max

200 tSRFCK RFSH Asserted Setup to CLK High 16 16

201 tSDRFCK DISRFSH Asserted Setup to CLK High 16 16

202 tSXRFCK EXTENDRF Setup to CLK High 8 8

204 tPCKRFL CLK High to RFIP Asserted 26 26

205 tPARQRF AREQ Negated to RFIP Asserted 38 38

206 tPCKRFH CLK High to RFIP Negated 41 41

207 tPCKRFRASH CLK High to Refresh RAS Negated 26 30

208 tPCKRFRASL CLK High to Refresh RAS Asserted 20 24

209a tPCKCL0 CLK High to CAS Asserted

during Error Scrubbing 64 73

15 ns, t

RAH

ASC

0 ns)

209b tPCKCL1 CLK High to CAS Asserted

during Error Scrubbing 74 83

15 ns, t

RAH

ASC

10 ns)

209c tPCKCL2 CLK High to CAS Asserted

during Error Scrubbing 74 83

25 ns, t

RAH

ASC

0 ns)

209d tPCKCL3 CLK High to CAS Asserted

during Error Scrubbing 85 94

25 ns, t

RAH

ASC

10 ns)

210 tWRFSH RFSH Pulse Width 9 9

211 tPCKRQL CLK High to RFRQ Asserted 22 22

212 tPCKRQH CLK High to RFRQ Negated 22 22

FIGURE 37. 200: Refresh Timing

TL/F/11118– 59

Page 46

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

Number Symbol

5.0Vg10%, 0§CkT

Parameter Description

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

Mode 0 Access

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

DP8430V/31V/32V-33

Min Max Min Max

300 tSCSCK CS Asserted to CLK High 8 8

301a tSALECKNL ALE Asserted Setup to CLK High

Not Using On-Chip Latches or if Using On-Chip Latches and

11 11

B0, B1, Are Constant, Only 1 Bank

301b tSALECKL ALE Asserted Setup to CLK High,

if Using On-Chip Latches if B0, B1 20 20 Can Change, More Than One Bank

302 tWALE ALE Pulse Width 10 10

303 tSBADDCK Bank Address Valid Setup to CLK High 10 10

304 tSADDCK Row, Column Valid Setup to

CLK High to Guarantee 6 10

tASR

0ns

305 tHASRCB Row, Column, Bank Address

Held from ALE Negated 6 6 (Using On-Chip Latches)

306 tSRCBAS Row, Column, Bank Address

Setup to ALE Negated 1 1 (Using On-Chip Latches)

307 tPCKRL CLK High to RAS Asserted 19 23

308a tPCKCL0 CLK High to CAS Asserted

308b tPCKCL1 CLK High to CAS Asserted

308c tPCKCL2 CLK High to CAS Asserted

308d tPCKCL3 CLK High to CAS Asserted

15 ns, t

RAH

15 ns, t

RAH

25 ns, t

RAH

25 ns, t

RAH

ASC

0 ns)

10 ns)

0 ns)

10 ns)

65 72

75 82

85 92

309 tHCKALE ALE Negated Hold from CLK High 0 0

310 tSWINCK WIN Asserted Setup to CLK High

to Guarantee CAS

is Delayed

311 tPCSWL CS Asserted to WAIT Asserted 20 20

312 tPCSWH CS Negated to WAIT Negated 20 20

313 tPCLKDL1 CLK High to DTACK Asserted

(Programmed as DTACK0

)

314 tPALEWL ALE Asserted to WAIT Asserted

(CS

is Already Asserted)

27 27

21 21

315 AREQ Negated to CLK High That Starts

Access RAS to Guarantee tASRe0ns 27 31 (Non-Interleaved Mode Only)

316 tPCKCV0 CLK High to Column

Address Valid 58 67

15 ns, t

RAH

ASC

0 ns)

317 tPCKCV1 CLK High to Column

Address Valid 68 75

25 ns, t

RAH

ASC

0 ns)

Page 47

13.0 AC Timing Parameters (Continued)

FIGURE 38. 300: Mode 0 Timing

TL/F/11118– 60

Page 48

13.0 AC Timing Parameters (Continued)

(Programmed as C4e1, C5e1, C6e1)

TL/F/11118– 61

FIGURE 39. 300: Mode 0 Interleaving

Page 49

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

Number Symbol

5.0Vg10%, 0§CkT

Parameter Description

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

Mode 1 Access

DP8430V/31V/32V-33

Min Max Min Max

400a tSADSCK1 ADS Asserted Setup to CLK High 9 9

400b tSADSCKW ADS Asserted Setup to CLK

(to Guarantee Correct WAIT

19 19

or DTACK Output; Doesn’t Apply for DTACK0)

401 tSCSADS CS Setup to ADS Asserted 4 4

402 tPADSRL ADS Asserted to RAS Asserted 20 24

403a tPADSCL0 ADS Asserted to CAS Asserted

(tRAH

15 ns, tASCe0 ns)

403b tPADSCL1 ADS Asserted to CAS Asserted

(tRAHe15 ns, tASCe10 ns)

403c tPADSCL2 ADS Asserted to CAS Asserted

(tRAH

25 ns, tASCe0 ns)

403d tPADSCL3 ADS Asserted to CAS Asserted

(tRAH

25 ns, tASCe10 ns)

404 tSADDADS Row Address Valid Setup to ADS

Asserted to Guarantee tASR

0ns

611

70 77

80 87

90 97

405 tHCKADS ADS Negated Held from CLK High 0 0

406 tSWADS WAITIN Asserted Setup to ADS

Asserted to Guarantee DTACK000 Is Delayed

407 tSBADAS Bank Address Setup to ADS Asserted 6 6

408 tHASRCB Row, Column, Bank Address Held from

ADS Asserted (Using On-Chip Latches)

409 tSRCBAS Row, Column, Bank Address Setup to

Asserted (Using On-Chip Latches)

ADS

410 tWADSH ADS Negated Pulse Width 12 17

411 tPADSD ADS Asserted to DTACK Asserted

(Programmed as DTACK0

)

29 29

412 tSWINADS WIN Asserted Setup to ADS Asserted

(to Guarantee CAS

Delayed during

Writes Accesses)

413 tPADSWL0 ADS Asserted to WAIT Asserted

(Programmed as WAIT

0, Delayed Access)

414 tPADSWL1 ADS Asserted to WAIT Asserted

(Programmed WAIT

1/2 or 1)

19 19

22 22

415 tPCLKDL1 CLK High to DTACK Asserted

(Programmed as DTACK0, 27 27 Delayed Access)

416 AREQ Negated to ADS Asserted

to Guarantee tASR

0ns 27 29

(Non Interleaved Mode Only)

417 tPADSCV0 ADS Asserted to Column

Address Valid 51 60

15 ns, t

RAH

ASC

0 ns)

Page 50

13.0 AC Timing Parameters (Continued)

FIGURE 40. 400: Mode 1 Timing

TL/F/11118– 62

Page 51

13.0 AC Timing Parameters (Continued)

FIGURE 41. 400: COLINC Page/Static Column Access Timing

TL/F/11118– 63

Page 52

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

Number Symbol

5.0Vg10%, 0§CkT

Mode 1 Dual Access

Parameter Description

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

DP8430V/31V/32V-33

Min Max Min Max

450 tSADDCKG Row Address Setup to CLK High That

Asserts RAS Port Change to Ensure tASR

451 tPCKRASG CLK High to RAS Asserted

for Pending Access

452 tPCLKDL2 CLK to DTACK Asserted for Delayed

Accesses (Programmed as DTACK

following a GRANTB 10 12

0ns

30 34

37 37

453a tPCKCASG0 CLK High to CAS Asserted

for Pending Access 81 88 (t

RAH

15 ns, t

ASC

0 ns)

453b tPCKCASG1 CLK High to CAS Asserted

for Pending Access 91 98 (t

RAH

15 ns, t

ASC

10 ns)

453c tPCKCASG2 CLK High to CAS Asserted

for Pending Access 91 98 (t

RAH

25 ns, t

ASC

0 ns)

453d tPCKCASG3 CLK High to CAS Asserted

for Pending Access 101 108

25 ns, t

RAH

454 tSBADDCKG Bank Address Valid Setup to CLK High

that Asserts RAS

10 ns)

ASC

for Pending Access

455 tSADSCK0 ADS Asserted Setup to CLK High 8 8

Page 53

13.0 AC Timing Parameters (Continued)

Unless otherwise stated V per bank, including trace capacitance (Note 2).

Two different loads are specified:

50 pF loads on all outputs except

150 pF loads on Q0 –8, 9, 10 and WE;or

5.0Vg10%, 0§CkT

70§C, the output load capacitance is typical for 4 banks of 18 DRAMs

50 pF loads on all outputs except

125 pF loads on RAS0 –3 and CAS0–3 and

380 pF loads on Q0 –8, 9, 10 and WE.

DP8430V/31V/32V-33

Number Symbol

Programming

Parameter Description

Min Max Min Max

500 tHMLADD Mode Address Held from ML Negated 6 6

501 tSADDML Mode Address Setup to ML Negated 6 6

502 tWML ML Pulse Width 12 12

503 tSADAQML Mode Address Setup to AREQ Asserted 0 0

504 tHADAQML Mode Address Held from AREQ Asserted 30 30

505 tSCSARQ CS Asserted Setup to

AREQ

Asserted

506 tSMLARQ ML Asserted Setup to AREQ Asserted 6 6

FIGURE 42. 500: Programming

TL/F/11118– 64

Page 54

14.0 DP8430V/31V/32V User Hints

1. All inputs to the DP8430V/31V/32V should be tied high,

low or the output of some other device.

Note: One signal is active high. COLINC (EXTNDRF) should be tied low

to disable.

2. Each ground on the DP8430V/31V/32V must be decou-

pled to the closest on-chip supply (V ramic capacitor. This is necessary because these grounds are kept separate inside the DP8430V/31V/32V. The decoupling capacitors should be placed as close as possible with short leads to the ground and supply pins of the DP8430V/31V/32V.

3. The output called ‘‘CAP’’ should have a 0.1 mF capacitor

to ground.

4. The DP8430V/31V/32V has 20X series damping resis-

tors built into the output drivers of RAS and WE

/RFRQ. Space should be provided for external damping resistors on the printed circuit board (or wirewrap board) because they may be needed. The value of these damping resistors (if needed) will vary depending upon the output, the capacitance of the load, and the characteristics of the trace as well as the routing of the trace. The value of the damping resistor also may vary between the wire-wrap board and the printed circuit board. To determine the value of the series damping resistor it is recommended to use an oscilloscope and look at the furthest DRAM from the DP8430V/31V/32V. The undershoot of RAS

, CAS,WEand the addresses should be kept to less than 0.5V below ground by varying the value of the damping resistor. The damping resistors should be placed as close as possible with short leads to the driver outputs of the DP8430V/31V/32V.

5. The circuit board must have a good V plane connection. If the board is wire-wrapped, the V and ground pins of the DP8430V/31V/32V, the DRAM associated logic and buffer circuitry must be soldered to the V

and ground planes.

) with 0.1 mF ce-

, CAS, address

and ground

6. The traces from the DP8430V/31V/32V to the DRAM should be as short as possible.

7. Parameter Changes due to Loading All A.C. parameters are specified with the equivalent load

capacitances, including traces, of 64 DRAMs organized as 4 banks of 18 DRAMs each. Maximums are based on worst-case conditions. If an output load changes then the A.C. timing parameters associated with that particular output must be changed. For example, if we changed our output load to

250 pF loads on RAS0 –3 and CAS0–3

760 pF loads on Q0 –9 and WE

we would have to modify some parameters (not all calculated here)

$308a clock to CAS

15 ns, t

RAH

A ratio can be used to figure out the timing change per

ASC

asserted

0 ns)

change in capacitance for a particular parameter by using the specifications and capacitances from heavy and light load timing.

$308a w/Heavy Loadb$308a w/Light Load

Ratio

79 nsb72 ns

125 pFb50 pF

$308a (actual)

CH(CAS)bCL(CAS)

7ns

(capacitance difference

75 pF

ratio)a$308a (specified)

(250 pFb125 pF)

11.7 nsa79 ns

90.7 ns@250 pF load

7ns

75 pF

79 ns

Page 55

Page 56

Physical Dimensions inches (millimeters)

Order Number DP8430V-33 or DP8431V-33

Plastic Chip Carrier (V)

NS Package Number V68A

256k/1M/4M Dynamic RAM Controller/Drivers

DP8430V/31V/32V-33 microCMOS Programmable

Plastic Chip Carrier (V)

Order Number DP8432V-33

NS Package Number V84ALIFE SUPPORT POLICY

NATIONAL’S PRODUCTS ARE NOT AUTHORIZED FOR USE AS CRITICAL COMPONENTS IN LIFE SUPPORT DEVICES OR SYSTEMS WITHOUT THE EXPRESS WRITTEN APPROVAL OF THE PRESIDENT OF NATIONAL SEMICONDUCTOR CORPORATION. As used herein:

1. Life support devices or systems are devices or 2. A critical component is any component of a life systems which, (a) are intended for surgical implant support device or system whose failure to perform can into the body, or (b) support or sustain life, and whose be reasonably expected to cause the failure of the life failure to perform, when properly used in accordance support device or system, or to affect its safety or with instructions for use provided in the labeling, can effectiveness. be reasonably expected to result in a significant injury to the user.

National Semiconductor National Semiconductor National Semiconductor National Semiconductor National Semiconductores National Semiconductor Corporation GmbH Japan Ltd. Hong Kong Ltd. Do Brazil Ltda. (Australia) Pty, Ltd.

2900 Semiconductor Drive Livry-Gargan-Str. 10 Sumitomo Chemical 13th Floor, Straight Block, Rue Deputado Lacorda Franco Building 16 P.O. Box 58090 D-82256 F4urstenfeldbruck Engineering Center Ocean Centre, 5 Canton Rd. 120-3A Business Park Drive Santa Clara, CA 95052-8090 Germany Bldg. 7F Tsimshatsui, Kowloon Sao Paulo-SP Monash Business Park Tel: 1(800) 272-9959 Tel: (81-41) 35-0 1-7-1, Nakase, Mihama-Ku Hong Kong Brazil 05418-000 Nottinghill, Melbourne TWX: (910) 339-9240 Telex: 527649 Chiba-City, Tel: (852) 2737-1600 Tel: (55-11) 212-5066 Victoria 3168 Australia

National does not assume any responsibility for use of any circuitry described, no circuit patent licenses are implied and National reserves the right at any time without notice to change said circuitry and specifications.

Fax: (81-41) 35-1 Ciba Prefecture 261 Fax: (852) 2736-9960 Telex: 391-1131931 NSBR BR Tel: (3) 558-9999

Tel: (043) 299-2300 Fax: (55-11) 212-1181 Fax: (3) 558-9998 Fax: (043) 299-2500