AHC/AHCT Designer's Guide: September 1998
AHC/AHCT Designer's Guide: September 1998
AHC/AHCT Designer's Guide: September 1998
SCLA013D
February 2000
Texas Instruments and its subsidiaries (TI) reserve the right to make changes to their products
or to discontinue any product or service without notice, and advise customers to obtain the latest
version of relevant information to verify, before placing orders, that information being relied on
is current and complete. All products are sold subject to the terms and conditions of sale supplied
at the time of order acknowledgement, including those pertaining to warranty, patent
infringement, and limitation of liability.
In order to minimize risks associated with the customer’s applications, adequate design and
operating safeguards must be provided by the customer to minimize inherent or procedural
hazards.
TI assumes no liability for applications assistance or customer product design. TI does not
warrant or represent that any license, either express or implied, is granted under any patent right,
copyright, mask work right, or other intellectual property right of TI covering or relating to any
combination, machine, or process in which such semiconductor products or services might be
or are used. TI’s publication of information regarding any third party’s products or services does
not constitute TI’s approval, warranty or endorsement thereof.
The Advanced High-Speed CMOS (AHC) logic family from Texas Instruments (TI) provides an
effortless migration path for HCMOS users who require higher speed and lower power without
paying a noise or price premium. The AHC logic family also offers the broadest selection of logic
choices, ranging from simple gates/MSI/octals (SN74AHCxxx) to single-gate (SN74AHC1Gxx)
and Widebus (SN74AHC16xxx) devices. Add to that the ability to operate at both 3.3 V and
5 V, and you have a reliable migration path from HCMOS.
Performance characteristics of the AHC family are:
• Low noise – The AHC family allows designers to maintain the same low noise
characteristics of HCMOS without the overshoot and undershoot typical of higher-drive
devices usually required to achieve AHC speeds.
• Low power – The AHC family, by using CMOS technology, has low power consumption
(40-µA maximum static current, one-half that of HCMOS).
• Speed – With typical propagation delays of 5.5 ns (’245), AHC offers three times the speed
of HCMOS.
• Drive – Output-drive current is ±8 mA at 5-V VCC and ±4-mA at 3.3-V VCC.
• 5-V input tolerance at 3.3 V – With the input diode to VCC removed, AHC is specified for
both 5-V and 3.3-V operation.
• Pin-for-pin compatibility – All AHC devices are pin-for-pin compatible with
industry-standard functional pinouts.
• Options – With CMOS- (AHC) and TTL- (AHCT) compatible devices available in
gates/MSI/octals, single gates, and Widebus, the AHC family offers the widest selection of
logic choices on the market.
• Packaging – AHC devices are available in D and DW (SOIC), N (PDIP), DB and DL
(SSOP), DGG and PW (TSSOP), DGV (TVSOP), and DBV (SOT) and DCK (SC–70)
packages. Selected AHC devices are available in military versions (SN54AHCxx).
For more information on these or other TI products, please contact your local TI representative,
authorized distributor, the TI technical support hotline at 972-644-5580, or visit the TI logic home
page at http://www.ti.com/sc/logic.
For a complete listing of all TI logic products, please order our logic CD-ROM (literature number
SCBC001) or Logic Selection Guide (literature number SDYU001) by calling our literature
response center at 1-800-477-8924.
iii
iv
Contents
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2. DC Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.1 Input Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2 Output Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Protection Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3.1 Electrostatic Discharge (ESD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
2.3.2 Latch-Up Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3. Dynamic Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1 Power Dissipation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.2 Quality of the Waveforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.1 Cross Talk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2.2 Ground Bounce . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3 Signal Transmission . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.1 Point-to-Point Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.3.2 Bus Lines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.4 Behavior With Slow Signal Edges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4. Special Application Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.1 Level Matching and Conversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.2 Partial Switching Off of Parts of a System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
5. Comparison of AHC and HC Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
6. Package Construction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
6.1 Single-gate Logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
7. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
8. References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
v
Contents (Continued)
vi
Your Next Choice: Advanced High-Speed CMOS Logic (AHC)
Do you use HCMOS logic in telecom, computer, industrial, automotive, or consumer applications?
For the same price as HCMOS, would you like to plug in devices with:
• 3-times the speed of HCMOS
• Half the static power consumption of HCMOS
• The same low noise as HCMOS
• A wide supply-voltage range
If so, this designer’s guide is for you . . . and so is advanced high-speed CMOS (AHC).
Convert HC to AHC
Family Positioning
5V
64 BCT 3.3 V
F
Drive – I OL – mA
24 LVC AC/ACT
12
AHC/ HC/HCT
8
AHCT LV
5 10 15 20
Performance – max tpd – ns
1
Power vs Frequency
70 ’AC11245
60
Power Per Bit – mW
50
’HC240 VCC = 5 V
40
30
’LV244
’AC11245 VCC = 3.3 V
20
’HC240 VCC = 3.3 V
10 ’AHC244 VCC = 5 V
2
Improve Switching Performance With AHC/AHCT
0 2 4 6 8 10 12 14 16 18 20
HC AHC AC Time – ns
Slew Rate (tPHL) 0.9 0.8 1.8 V/ns
5.5 V
5.0 V
4.5 V
AC244
4.0 V
3.5 V
3.0 V
VI HC244
2.5 V
AHC244
2.0 V
1.5 V
1.0 V
0.5 V
VCC = 5 V
0.0 V
Load = 50 Ω/50 pF
–0.5 V
0 2 4 6 8 10 12 14 16 18 20
HC AHC AC Time – ns
Slew Rate (tPHL) 0.8 0.7 1.5 V/ns
3
Ground-Bounce Comparison
AC244
HC244
Noise – 500 mV/Division
AHC244
Time – 1 ns/Division
Technical Comparison of AHC Versus Other CMOS Logic Families (’245 Function)
AHC/AHCT HC/HCT AC/ACT
VCC 5V 3.3 V 5V 2V 5V 3.3 V
Drive –8/8 mA –4/4 mA -8/8 mA 20 µA –24/24 mA –12/12 mA
Speed (typical) 5.5 ns 8.3 ns 18 ns 54 ns 3.5 ns 5 ns
Ground bounce 0.5 V (–0.2 V) N/A 0.6 V (–0.3 V) N/A 1.5 V (–1.8 V) N/A
8.6 pF
Power dissipation capacitance† N/A 40 pF N/A 45 pF N/A
(at 1 MHz)
Quiescent power dissipation 40 µA 80 µA 40 µA
Input 3.3 V 5V 3.3 V 3.3 V
Level conversion option
Output 5V 3.3 V 5V 5V
Widebus package available Yes Yes No No No Yes
† CL = 50 pF, f = 10 MHz unless otherwise specified
4
Widebus Minimizes Board Space
(SN74AHC16xxx/SN74AHCT16xxx)
The trend toward 16-bit and 32-bit Widebus systems to increase data throughput continues unabated, requiring bus drivers that
support these formats.
Many 16-bit bus systems can be supported easily by TI Widebus devices. These are designed to replace the commonly used
8-bit functions. A single 16-bit Widebus package replaces 2 × 8-bit packages. A typical Widebus example is the
SN74AHC16244, which incorporates twice the functionality of an SN74AHC244.
×1 16-Bit
+ =
5
Multiple Package Options
AHC Packages
PDIP (N) SOIC (D/DW) SOT-23 (DBV) SC-70 (DCK)
DUAL-IN-LINE SMALL OUTLINE SMALL OUTLINE SMALL OUTLINE
PIN COUNT 14 16 18 20 28 16 20 28 5 5
Width 6.60 6.60 6.60 6.60 14.22 7.59 7.59 7.59 1.80 1.35
Length 19.69 19.69 23.37 24.77 36.83 10.41 12.95 18.03 3.10 2.2
Pitch 2.54 2.54 2.54 2.54 2.54 1.27 1.27 1.27 0.95 0.65
Height 5.08 5.08 5.08 5.08 5.08 2.65 2.65 2.65 1.3 1.0
PACKAGES
TI NOMENCLATURE PHILIPS MOTOROLA FAIRCHILD TOSHIBA
SOT-23 (new) DBV (5 pin) N/A N/A M5X F
SC-70 (new) DCK (5 pin) N/A N/A P5X FU
PDIP N, NT (24 pin) N P N P
SOIC D, DW (20 pin–>) D M M FN
SSOP DB (24 pin–>), DL (48 pin–>) DB, DL SD MSA, MEA FS
TSSOP PW (–>24 pin), DGG (48 pin–>) PW, DGG DT MTC, MTD ST
TVSOP (new) DGV (≤ 56 pin) N/A N/A N/A N/A
Data sheets are available in the AHC/AHCT Logic Advanced High-Speed CMOS Data Book, literature number SCLD003A,
or at http://www.ti.com/sc/logic.
6
Single-Gate Logic
(SN74AHC1Gxxx/SN74AHCT1Gxxx)
TI’s Single-gate Logic helps designers of handheld systems, such as portable computers, remote control units, and cellular
telephones, to reduce the size and weight of their designs. The extremely small logic devices simplify the layout of printed
circuit boards, and can be used to make simple functional modifications of ASICs without the cost and delay of redesigning.
Nomenclature Package
SN 74 AHC 1G xx DBV R Area:
9 mm2
SOT-23
Tape and Reel 1.8 mm 3.0 mm
Height (max): (DBV)
SOT-23 Package 1.45 mm
Nomenclature Package
SN 74 AHC 1G xx DCK R Area:
5 mm2
SC-70
Tape and Reel 1.25 mm 2.1 mm
Height (max): (DCK)
SC-70 Package 1.10 mm
The Principle
Standard Gates Single-gate
Board Space Comparison
Quad Gate Single-gate Area Saving
TSSOP 14 Pin SOT-23 5 Pin 24.36 mm2 or
33.66 mm2 (max) 9.3 mm2 (max) 72.7%
Benefits
• Small Package (SOT-23)
• Optimized PCB Layout
• Reduced EMI Noise
• Enhanced ASIC Functionality
Performance
• 3.5-ns typical propagation delay
• ±8-mA output drive
• 20-µA static current
• CMOS (AHC)- and TTL (AHCT)-compatible versions
Cross-Reference Examples
TEXAS INSTRUMENTS TOSHIBA
SN74AHC1G00DBV TC7SH00F
SN74AHCT1G00DBV TC7ST00F
SN74AHCU1G04DBV TC7SHU04F
7
8
Abstract
With the advanced high-speed CMOS family of logic devices, TI has brought to market a series of components that fully meets
today’s requirements for increased speed, that is, reduced signal delay time, and for operation from supply voltages of 5 V and
3.3 V. This document first addresses the electrical characteristics of these new devices. A detailed investigation of
dc parameters, input/output characteristics, and dynamic behavior follows. Power consumption, cross talk between signal
lines, and electromagnetic compatibility also are discussed.
1. Introduction
The introduction of the high-speed CMOS family SN74HC device at the beginning of the 1980s provided the system designer
with a sensible and logical alternative to bipolar logic devices, which, until then had been so widely used. These CMOS devices
featured delay times approximately comparable to those of the low-power Schottky family. The output currents that these
components could deliver were also comparable to those of their bipolar predecessors. An advantage of the CMOS devices
is the wide range of supply voltages (2 V to 6 V) with which these components could be operated. It allowed their effective
application in battery-operated equipment. However, as with all other CMOS devices, lowering the supply voltage meant that
increased delay times had to be tolerated. A few years later, improvements in semiconductor technology made possible the
introduction of the advanced CMOS (AHC) devices. In addition to the advantage of a wide range of supply voltage, these
devices featured significant improvements in drive capability and delay time. As a result, CMOS devices were for the first time
able to penetrate a domain that had been the reserve of fast bipolar logic devices from the SN74F and SN74AS series. Also,
whereas CMOS devices were seldom used in applications with extreme requirements of drive capability and speed, such as
backplane wiring in large computer systems, advanced CMOS devices established a firm position in all applications. Examples
include applications in personal computers and workstations.
With the introduction of notebook computers at the beginning of the 1990s, new requirements were placed on logic devices
to perform well at the supply voltage of 3.3 V, which is usual in battery-operated equipment. The HC and AC devices performed
inadequately with regard to drive capability and delay time at this low supply voltage. As fast logic circuits, the LVC and ALVC
devices were acceptable successors to the AC devices. In drive capability and delay time, these new logic families operating
at a supply voltage of only 3.3 V provided the same results as the well-known AC devices using a supply voltage of 5 V. In
3.3-V applications, a useable successor to the HC devices was still missing. In many applications, the outstanding
characteristics of the LVC and ALVC devices are not required. In fact, the interference resulting from the steep-edges
characteristic of such devices is a disadvantage. They require additional circuit-design precautions, such as multilayer circuit
boards, which, in turn, increase equipment costs unnecessarily. With the introduction of the LV series, the attempt was made
to create quickly an appropriate logic family. Clever modifications of the process steps in semiconductor manufacturing
allowed better performance at a lower supply voltage, but the long-term result was inadequate. Consequently, a new logic
family, which, at a supply voltage of 3.3 V, would have the same, or better characteristics as its well-known predecessor was
needed. Also, several problems associated with interfaces of circuits operating at 3.3 V and 5 V needed to be addressed. For
various reasons, future systems are expected to use both supply voltages.
The logical answer to all these questions is the series of advanced high-speed CMOS devices manufactured in a process that
permits gate lengths of 1 µm. The result is typical delay times of 6 ns at 3-V VCC that, in the past, were measured on ALS
circuits. This document acquaints the system designer with the characteristics of these advanced components. In addition, a
large number of questions with which the designer is often confronted are discussed. Many application problems that can be
solved elegantly with this new logic family are shown.
9
2. DC Characteristics
Q1 Q5 Q3
Input
D1
Q2 Q4
10
6
VCC = 5 V
5
4
VCC = 3 V
VO – V
3
VCC = 2 V
2
0
0 1 2 3 4 5
VI – V
6
VCC = 5.5 V
5
VCC = 4.5 V
4
VO – V
0
0 1 2 3 4 5
VI – V
11
5
4
VCC = 5 V
3
VO – V 2
0
0 1 2 3 4 5
VI – V
4
VCC = 3.3 V
3
VO – V
0
0 1 2 3 4 5
VI – V
3
VO – V
VCC = 2 V
2
0
0 1 2 3 4 5
VI – V
12
5
4
VCC = 5 V
VO – V
3
0
0 1 2 3 4 5
VI – V
2.5
2.0
I CC – mA
1.5
1.0
0.5
0.0
0 1 2 3 4 5
VI – V
13
Table 1. Specification of ∆ICC
SN74AHCT245
PARAMETER TEST CONDITIONS Vcc UNIT
MIN MAX
∆ICC One input at 3.4 V, Other inputs at VCC or GND 5.5 V 1.5 mA
Input characteristics of an AHC device depend on the input of the inverter with a diode connected in parallel (see Figure 1)
that is part of the electrostatic discharge (ESD) protection circuit for the input. In addition, the diode limits negative-going
overshoots caused by line reflections and improves the quality of the signal. Over a voltage range of 0 ≤ VI ≤ 7 V, the circuit
has an extremely high resistance, as indicated by the value of II in Table 2. When a device’s output that is in an inactive
high-impedance state is connected internally in parallel with an input, as in bidirectional circuits, for example, SN74AHC245,
IOZ should be used as the effective input current (see Table 2). The value of IOZ is the sum of the leakage currents of the input
and output circuits.
Table 2. Specification of the Input Current
SN74AHC245
PARAMETER TEST CONDITIONS UNIT
MIN MAX
II VI = VCC or GND ±1 µA
IOZ† VO = VCC or GND, VI(OE) = VIL or VIH ±2.5 µA
† The parameter IOZ includes the input leakage current.
Input voltages greater than 7 V must be avoided to preclude damage to the gate oxide of the input stage. This damage is not
necessarily permanent, but will adversely affect the expected lifetime of the circuit. The gate oxide of AHC devices is only
200 Å thick. An input voltage of 7 V corresponds to a field strength over the gate oxide of 350 kV/cm. Although breakdown
of the oxide is expected only at input voltages above 10 V, electrons tunnel increasingly into the gate oxide at field strengths
greater than 350 kV/cm, influencing characteristics of the transistors and causing failure.
In practice, negative input voltages are of greater interest. These voltages result from negative-going overshoots generated by
line reflections. To limit these negative overshoots and improve the quality of the signal, an effective clamping diode (D1 in
Figure 1) is used. Figure 7 shows a typical input characteristic of an AHC device. The input is at a high resistance with positive
input voltages (0 V ≤ VI ≤ 7 V). With negative input voltages, the clamping diode conducts. It also limits negative-going
overshoots at higher currents to voltages of about –1 V (see Figure 7).
7
6
5
4
VI – V
3
2
1
0
–1
–2
–100 –80 –60 –40 –20 0 20
II – mA
14
2.2 Output Circuit
The simplified output circuit of an AHC device is shown in Figure 8. Only those components necessary to understand the
behavior of the circuit are shown.
VCC
Q1
D1
Internal
Logic Output
Q2
D2
Figure 9 shows the high- and low-logic output characteristics of AHC devices for various supply voltages. Figure 10 shows
the capacitive loading effect on AHC devices.
Output characteristics of AHC devices with 3-state outputs in the inactive high-impedance state are shown in Figure 11. The
output data are based on the simplified circuit of the output stage in Figure 8. In the operating state discussed here, output
transistors Q1 and Q2 are nonconducting. Over a range of output voltage from 0 V ≤ VO ≤ VCC, the circuit is, accordingly,
at a high resistance. If the output voltage is raised to a value above VCC + 0.7 V, or reduced to below –0.7 V, diode D1 or D2,
respectively, conducts and will limit the output voltage. The output characteristic with VCC = 0 V is shown in Figure 11. At
this supply voltage, the output transistors do not conduct. The circuit then behaves like two diodes connected in parallel but
with opposite polarities. These curves apply to circuits with 3-state outputs and to those with the push-pull output stage, which
is usual with all CMOS devices.
15
0 30
VCC = 4.5 V TA = 25°C
VCC = 4.5 V
(MAX)
I OH – Output High Current – mA
–20 10
TA = 25°C TA = 25°C
(MIN) (TYP)
–30 0
2.5 2.9 3.3 3.7 4.1 4.5 0 0.4 0.8 1.2 1.6 2.0
VOH – Output High Voltage – V VOL – Output Low Voltage – V
0 30
TA = 85°C
VCC = 3.0 V (MIN) VCC = 3.0 V
I OH – Output High Current – mA
TA = 25°C
(TYP) TA = 25°C
–20 10 (MAX)
TA = 85°C
(MAX)
–30 0
1.0 1.4 1.8 2.2 2.6 3.0 0 0.4 0.8 1.2 1.6 2.0
VOH – Output High Voltage – V VOL – Output Low Voltage – V
16
16
TA = 85°C (MAX)
0.13 ns/pF
TA = 25°C (TYP)
4 0.07 ns/pF
0
50 100 150 200
CL – Load Capacitance – pF
7
VCC = 5 V
6
4
VCC = 3.3 V
VO – V
2
VCC = 0 V
1
–1
–2
–100 –50 0 50 100
IO – mA
Figure 11. Output Characteristics in High-Impedance State With Supply Voltage Switched Off
17
2.3 Protection Circuits
Because of their small internal structures, all integrated circuits are susceptible to ESD. An additional problem arises with
complementary MOS circuits whose internal structures form parasitic thyristors, which under certain conditions, can be fired
and cause a short circuit. Destruction of the device usually is the unavoidable consequence. Therefore, when developing and
manufacturing integrated circuits, semiconductor manufacturers must take precautions to protect them from ESD.
Human-Body Model
This model simulates the situation in which the energy stored in the human body is discharged into the device under test. In
this case, a 100-pF capacitor is charged to ±2000 V, then discharged through a resistor of 1.5 kΩ into the device under test.
The rise time of the discharge current must be less than a nanosecond.
Machine Model
In this model, immunity to disturbances that contain considerably more energy but have a significantly longer current rise time
is tested. For this purpose, a 200-pF capacitor is charged to ±200 V, then discharged without any series resistor into the device
under test. The inductances of the lines in the measurement setup (L > 500 nH) reduce the rate of rise of the discharge current
sufficiently.
Charged-Device Model
This test simulates the situation in which an integrated circuit is charged, for example, by sliding along a plastic transport rail
before insertion by an automatic insertion machine, then is discharged when it touches the printed circuit board. The
capacitance of the integrated circuit including package, in which the energy is stored, is then only a few picofarads, but at the
instant of the discharge extremely short rise times can be expected. With integrated circuits as currently used, withstanding
±1000 V can be regarded as sufficient in this test.
The engineer who designs integrated circuits must provide protection circuits that will withstand the stresses of the tests
described above. A distinction must be made between two destructive processes. High energy levels with relatively long rise
times (machine model) in which the protection circuit must be designed with sufficient ability to conduct current away. With
the two other test methods, the danger is that, because of the extremely short rise times, the protection circuit will only partially
conduct and is overloaded in this region.
Conventional protection circuits consist of diodes or zener diodes that conduct away the currents and limit the voltages.
Resistors in series with the circuit to be protected limit the current. Besides reliably diverting the current, whereby the circuit
must be protected against thermal overload, the device also must be protected against excessive voltages, for example, to avoid
a breakdown of the gate oxide of an MOS transistor. In general, a combination of various methods is used to obtain
optimum results.
Figure 12 shows protective circuits used for advanced high-speed CMOS devices. To meet the requirements outlined
previously, the protective circuit is constructed in two stages. The input is first protected by a thyristor consisting of transistors
Q2 and Q3. This provides coarse protection. If the input voltage rises above about 15 V, transistor Q1 breaks down and fires
the thyristor. The latter then short circuits the high currents. Resistors R1 and R2 have values of only a few ohms. Therefore,
the holding current of the thyristor is several tens of mA. When the current is reduced again at the end of the discharge, the
thyristor is extinguished. Transistors Q4, Q5, and Q6 operate as fine protection and are intended principally to protect the input
from excessive voltages. When there are overvoltages at the input, these transistors are driven into breakdown and limit the
voltage, while resistor R3 limits the current.
18
VCC
Q7 Q9
Q4
Q5 D1
Input Output
R3
R1
Q3 D2
Q8 Q10
Q2
Q1 Q6
R2
N-Channel P-Channel
Transistor Transistor
VCC
S G D D G S
P+ N+ N+ N+ P+ P+
RS RW
N-Well
P-Substrate
19
In the early days of CMOS technology, the latch-up effect was a major problem for system designers. Often, many additional
precautions had to be taken in a system to avoid excessive currents in the connections to integrated-circuit devices. This
inevitably increased the cost of the complete equipment. To counteract the disadvantage of CMOS devices at that time,
precautions were taken later when designing the device to prevent latch-up from occurring. This began with the choice of a
high-resistance substrate to prevent the spreading of undesired currents. In addition, n- or p-doped guard rings (see Figure 14)
were placed around critical parts of the circuit that were connected to the corresponding supply-voltage rails. These guard rings
function as additional collectors of the parasitic transistors. Since these collectors are considerably closer to the corresponding
base-emitter areas than the bases of the complementary transistors, they take the major part of the current that wanders about
in the substrate. In this way, the thyristor is not completely eliminated. However, its sensitivity is reduced to such an extent
that, under normal operating conditions, triggering the thyristor is not expected. During the characterization of a new
component (type testing), its resistance to latch-up also is checked. With AHC circuits, a current of ±300 mA is applied to
all relevant pins of the device under test. At an ambient temperature of 125°C and VCC = 7 V, latch-up must not occur. At room
temperature, currents of more than 1 A typically are necessary to cause latch-up.
Four Guard Rings
P-Channel
N-Channel
Transistor
Transistor VCC
S G D D G S
P+ N+ N+ P+ N+ P+ N+ P+ P+
RS RW
N-Well
P-Substrate
20
3. Dynamic Behavior
An important parameter when choosing a device is the delay time. Table 4 gives a comparison between HC/HCT and AC/ACT
devices. Advanced high-speed CMOS devices are about three times faster than comparable HC devices; AHC and the
TTL-compatible AHCT devices have only minor differences with regard to their dynamic characteristics.
Table 4. Comparison of the Delay Times of HC and AHC Devices
DEVICE SN74HC SN74HCT SN74AHC SN74AHCT
’244 buffer 13 ns 15 ns 5.8 ns 5.4 ns
’245 transceiver 15 ns 14 ns 5.8 ns 4.5 ns
’373 latch 15 ns 20 ns 5 ns 5 ns
’374 flip-flop 17 ns 25 ns 5.4 ns 5 ns
The quiescent power dissipation, Pr, is calculated as the product of the supply voltage, VCC, and the quiescent current, ICC,
as given in the data sheet. This quiescent current results primarily from the leakage currents of the reverse-biased p-n junctions
in the integrated circuit. At room temperature, it is only a few nanoamperes. This current usually can be neglected, but leakage
currents in depletion layers typically double with a temperature increase of 10°C. In equipment that is operated at high
temperatures, this leakage current can be significant.
The switching loss, Ps, results from charging, discharging, and switching processes inside the device. The charge and discharge
of the internal capacitances of the circuit make up a minor part of the total. The major part comes from the current spikes that
occur when switching every CMOS stage and which, in this case, primarily affect the output stage. If a CMOS output stage
(shown in Figure 8) is switched from a high to low logic level or vice versa, the control voltage on the gate of the transistor
within the device only rises (or falls) in a finite time from low to high. The complementary transistor also is being driven in
this finite time. Thus, at the moment of switchover, both transistors conduct simultaneously for several nanoseconds.
Therefore, a considerable current flows for a short time (see Figure 15) in the circuit. When measuring this current, care must
be taken not to capacitively load the output being measured.
ICC
VI
2
Ch1: 4mA Ch2: 5V 12.5ns
21
The charge, Q, and the energy consumed can be calculated at every switching cycle from the amplitude of the current and its
waveform over a period of time. In this way, switching loss, Ps, can be calculated. In practice, a simpler process is used. The
supply current, ICCS, of the circuit being considered is measured at a specific input frequency, fI, but the output must not be
loaded. This current consumption can be thought of as generated by an equivalent power-dissipation capacitance, Cpd, at the
output of the circuit. The following expression then applies:
or
C pd + VI @ fCCS
CC I
(3)
This power-dissipation capacitance, Cpd, is given in the data sheet. In a particular application, the following formula can be
used to calculate the switching loss Ps:
Ps +C @V @f
pd CC
2
I (4)
Where:
PI +C @V @f
L CC
2
O (5)
Neglecting the quiescent power dissipation, Pr, the following expression gives total power dissipation:
P ges + ǒC @ f ) C @ f ǓV
pd I L O CC
2
(6)
For a SN74AHC244, the data sheet gives a power-dissipation capacitance, Cpd, of 8.6 pF. With a capacitive load, CL, of 50 pF
and VCC = 5 V, the following then applies:
P ges + (8.6 pF @ f ) 50 pF @ f )5
I O
2
(7)
With a buffer such as the SN74AHC244, the input and output frequencies are the same (fI = fO). In this case, the resulting power
dissipation per output becomes:
Figure 16 provides a comparison between the theoretical power dissipation calculated from the formula above and the
dissipation actually measured. There is good correlation between the theoretical result and the measurements made. Figure 17
shows the measurement results at a supply voltage VCC = 3.3 V and load of 50 pF or with no load at the output.
22
700
SN74AHC244
VCC = 5 V
600 CL = 47 pF
Eight Outputs Switching Measured
Power Dissipation – mW
500
Calculated
400
300
200
100
0
0 10 20 30 40 50 60
Frequency – MHz
500
SN74AHC244
VCC = 3.3 V
400 Eight Outputs Switching
Power Dissipation – mW
CL = 47 pF
300
200
CL = 0 pF
100
0
0 10 20 30 40 50 60
Frequency – MHz
23
3.2.1 Cross Talk
The cross talk between adjacent signal lines results from the undesirable inductive and capacitive coupling between them. A
precise mathematical treatment of this phenomenon is very complicated, particularly because the precise electrical
characteristics of the lines, such as the line inductance and capacitance and the line mutual inductance and mutual capacitance,
must be known. For the system designer, it usually is sufficient to know the behavior of typical configurations to draw
conclusions about similar situations in other applications. Typical line configurations are illustrated in Figure 18.
VCC
G4 G3
G1 G2
VCC
G3 G4
Disturbing Signal
2
Ch1: 2V Ch2: 2V 12.5ns
24
The behavior is different with transmission in opposite directions (see Figure 18b). The interfering signal, which is coupled
into the disturbed line when gate G1 is switched, encounters the high-resistance input of G3 and has significant effects. The
disturbance then runs to the end of line G3–G4 (output at G4). Because the output impedance of this gate typically is
significantly lower than the line impedance, the interfering waveform is reflected with reversed polarity. After its return, the
disturbance will arrive at the input of gate G3. At this point, an interfering pulse can be expected, the length of which is
determined as a result of its doubled signal propagation time on this line. Line length directly influences the magnitude of the
interference. This considerably more critical manifestation of cross talk is known as near-end cross talk. In the example shown,
AHC devices should not have been disturbed. Their switching threshold is typically about 2.5 V, providing an adequate noise
margin. In contrast, the situation is different when G3 (see Figure 18b) has a TTL-compatible input stage with a threshold
voltage of 1.5 V. In this case, the switching threshold of the circuit that is disturbed is clearly exceeded and experience has
shown that this can lead to false triggering. As mentioned previously, the length of the interference pulse, tw, is in accordance
with the doubled signal propagation time on the line in question. With a line having a length of 25 cm and a typical signal
propagation time of 6 ns/m (see Figure 20), the width of the resulting interference pulse is:
tw + 2 @ t @ l + 2 @ 6 nsm @ 25 cm + 3 ns
p (9)
Disturbing Signal
2
Ch1: 2V Ch2: 2V 12.5ns
25
Disturbing Signal
Figure 21. Near-End Cross Talk With Screening Between the Lines and Line Length of 25 cm
m + L @ didt (10)
It is this change of voltage that appears with undiminished amplitude at the output of inverter Q3/Q4, and the output potential
should remain constant. Circuits connected to its output may be influenced by this disturbance. The same effect, but with
opposite polarity, occurs when the output in question is at a high logic level and the other outputs of the circuit are switched
from a low to a high logic level.
26
VCC
Lg
Q1 Q3
Lg C1
VCC
Lg
C2
Q2 Q4
Lg
VCC To
VCC DIR VCC VCC Oscilloscope
A1 OE Ri = 50 Ω
A2 B1 R2
A3 B2
A4 B3
A5 B4 C C C R1 R2
A6 B5 C R1
A7 B6 C R1
A8 B7 C R1
GND B8 C R1
C R1
C = 47 pF
R1 = 500 Ω
R2 = 450 Ω
Figure 23. Circuit for Evaluating Simultaneous-Switching Noise (8-Bit SN74AHC245, VCC = 5 V)
Figure 24 shows the interference voltage that arises when simultaneously switching several outputs as measured on an
SN74AHC245 bus-interface device in a dual-in-line (N) package. The measured output B2 (see Figure 23) is at a low logic
level, while seven other outputs are switched simultaneously from high to low.
27
VCC = 5.5 V
Seven Outputs Switching
Simultaneously
1
Output B2
2
VCC = 3.3 V
Seven Outputs Switching
Output B3 Simultaneously
1
Output B2
2
28
Besides the effect of the voltage drop across the inductances of the supply lines, the amplitude of the noise voltage also is
determined by the cross talk between the pins of the package. One of the consequences is that the measured interference voltage
is at a maximum at those pins that have simultaneously switching outputs on both sides. Conversely, interference voltages at
the ends of the package are significantly lower. When low distortion of the signal is particularly important in specific
applications, the latter situation can be attained by appropriate routing of the signals. As a result of the many supply voltage
connections distributed around the perimeter of Widebus packages (see Figure 26), harmful inductance is reduced in
accordance with the number of parallel electrical connections. Also, supply lines between the signal lines reduce the coupling
between signal lines, further contributing to the low level of interference voltage.
1OE 1 48 2OE
1Y1 2 47 1A1
1Y2 3 46 1A2
GND 4 45 GND
1Y3 5 44 1A3
1Y4 6 43 1A4
VCC 7 42 VCC
2Y1 8 41 2A1
2Y2 9 40 2A2
GND 10 39 GND
2Y3 11 38 2A3
2Y4 12 37 2A4
3Y1 13 36 3A1
3Y2 14 35 3A2
GND 15 34 GND
3Y3 16 33 3A3
3Y4 17 32 3A4
VCC 18 31 VCC
4Y1 19 30 4A1
4Y2 20 29 4A2
GND 21 28 GND
4Y3 22 27 4A3
4Y4 23 26 4A4
4OE 24 25 3OE
29
G1 G2
D1
Beginning of Line
End of Line
2
Ch1: 1V Ch2: 1V 5ns
a) G1 G2
D1
b) G1 G2
Rt
c) G1 G2
Rt
Ct
d) G1 G2
Rs
30
The use of clamping diodes (D1 in Figure 29a) is the most effective method of termination, especially when the protection
diodes already incorporated in the input circuits of AHC devices can take on this job. In some cases, limiting positive
overshoots may also be advisable. In this case, additional diodes should be connected between the input and the positive
supply-voltage connection terminal.
The use of termination resistors of the proper value at the end of the line (Rt in Figure 29b) produce ideal waveforms. However,
the higher power dissipation in the termination resistors, which results from this arrangement, usually outweighs the advantage
of the low distortion of the signal.
If a termination at the end of the line cannot be avoided, connecting a termination resistor and a capacitor in series is
recommended. This blocks dc from the terminating network and reduces power consumption of the circuit. Capacitor Ct in
Figure 29c is chosen so that the time constant Rt × Ct is approximately four times the signal propagation time along the line.
A more elegant method of preventing undershoots and overshoots at the end of the line consists of matching the output
impedance of the line driver with a series resistor (Rs in Figure 29d) to the line impedance. This makes optimum matching
possible, without adversely affecting the balancing of the line.
Bus
31
Zo = 30 Ω
tp = 10 ns
Beginning of Line
End of Line
2
Ch1: 1V Ch2: 1V 25ns
VCC = 5 V
RL = 500 Ω
Input Output
1
Ch1: 1V Ch2: 1V 12.5ns
32
4. Special Application Problems
For several years, systems have been designed and manufactured to use two or more supply voltages, 3.3 V and 5 V. The reason
is that, with the introduction of the so-called low-voltage logic circuits, all components needed were not available and, in some
cases, still are not. Therefore, often there was no alternative but to use integrated circuits requiring a supply voltage of 5 V in
systems conceived for a supply voltage of 3.3 V. Special circuit techniques are then required at the interfaces. The problems
involving the use of several supply voltages can be expected to increase in the future. With components having structures of
<0.5 µm being manufactured, still lower operating voltages will be needed, and the problem mentioned above will appear again
in another form. Level conversion and matching will remain an applications problem.
AHCT
Any 3-V ABT, BCT
Logic Circuit ALS, AS, F
etc.
AHC
Any 5-V
LVC
Logic Circuit
LVT
FROM TO VCC = 3 V
VCC = 5 V LV LVC ALVC HC AC AHC LVT ALVT
Bipolar TTL No Yes No No No Yes Yes Yes
BiCMOS (ABT, BCT) No Yes No No No Yes Yes Yes
CMOS No Yes No No No Yes Yes Yes
33
4.2 Partial Switching Off of Parts of a System
Partial switching off of parts of a system occurs when part of an equipment or installation is switched off (without supply
voltage) while other parts of the equipment remain in normal operation. This operating situation occurs regularly at the
interfaces with other equipment. The same state can be observed frequently within a module that operates with several supply
voltages, for example, 3.3 V and 5 V. Since the individual power supplies are not switched on and off simultaneously and in
coordination, the case in which one or other power supply does not deliver the required voltage must be considered. The
simplified output circuit of an AHC/AHCT device is shown in Figure 8. Diode D1 short circuits the output to ground when
the supply voltage is switched off (VCC = 0 V). Since this diode has a very low resistance (see Figure 11), this operating state
is a defined low-logic level. To this extent, such a circuit provides a defined level. This behavior has disadvantages in bus
systems. If the supply voltage of one of the subscribers connected to the bus is switched off, its output short circuits the complete
bus line. A solution in such a case can be provided only by using bipolar and BiCMOS circuits, which do not have the diodes
shown in Figure 8 in their output stages. In this connection, special mention should be made of the circuits from the SN74ABT
and SN74LVT series.
Many of the interface problems discussed here can be solved very easily using integrated circuits specially developed for this
purpose, such as the bidirectional 8-bit Widebus transceiver SN74LVC4245 (see Figure 34), or its 16-bit Widebus version
SN74ALVC164245. These components have two separate supply-voltage connections [VCCA (5 V) and VCCB (3.3 V)]. In this
way it is possible to solve the problems previously discussed by means of appropriate circuitry within the component. The
engineer developing a system will no longer be concerned with these problems.
34
5. Comparison of AHC and HC Circuits
High-speed CMOS and advanced CMOS circuits have been used for more than a decade in many diverse applications. The
HC circuits feature comparatively simple application rules, and this has encouraged their widespread use. AC circuits are found
in applications in which high speed (i.e., short delay times) and high drive capability are required. The latter advantages must
be weighed against the considerable internal noise (ground bounce, cross talk, etc.) that these circuits generate.
The ideal situation was a combination of the advantages of both logic families. Maintaining the moderate drive capability of
HC circuits, which ensures a low internal-noise level, and incorporating the technical advantages offered by a modern
manufacturing process with structures of 1 µm, the creation of the advanced high-speed CMOS family became a reality. In
addition, particular attention was paid to the increasing trend toward applications operating with supply voltages of only 3.3 V.
A number of improvements were also incorporated that facilitate applications with these components: changes to the input
circuits, and improved ESD protection. The most important parameters are summarized in Table 8.
Table 8. Comparison of the Logic Families
PRODUCT FAMILY
AHC HC LVC AC
Technology CMOS CMOS CMOS CMOS
Structure (gate length) 1 µm 2–3 µm 0.8 µm 1 µm
5-V tolerant? Yes No Yes No
Gate and bus-interface circuits available? Yes Yes Yes Yes
Widebus circuits (16 bit) available? Yes No Yes Yes
Bus-hold circuit? No No Yes No
VCC = 5 V VCC = 3.3 V VCC = 5 V VCC = 3.3 V VCC = 5 V
Supply current, ICC (’245) 40 µA 40 µA 80 µA 10 µA 40 µA
Output current –8/8 mA –4/4 mA –6/6 mA –24/24 mA –24/24 mA
Delay time, tpd(max) (’245) 6.5 ns 10 ns 26 ns 7.5 ns 9 ns
Input capacitance, Ci (’245) 2.5 pF 4.6 pF 3.3 pF 4.5 pF
Input/output capacitance, Cio (’245) 8 pF 16 pF 5.4 pF 15 pF
6. Package Construction
The trend toward further miniaturization of equipment and appliances is continuing, as indicated by the huge range of portable
battery-operated equipment now available. Manufacturers of semiconductors are making a major contribution to this trend,
because miniaturization can be realized only with smaller packages and corresponding progress in manufacturing technology.
System designers always should remain aware of the problems involved in the use of modern packages.
Special manufacturing techniques when encapsulating the integrated circuits (chips) in their packages are employed to
overcome problems that can occur. Everything possible must be done to eliminate humidity inside the package. This humidity
has less to do with possible corrosion of the integrated circuits because, for the last 20 years, surfaces of all chips have been
passivated with a glass layer (nitride), and possible corrosion has lost its significance. Any humidity trapped in the package
shows up as a problem with the soldering techniques, for example, flow-soldering baths, now used for surface-mounted
components. During the soldering process, humidity can vaporize and cause the package to burst (the “popcorn” effect).
Immediately after manufacture, the devices must be stored in a special packing (Dry Pack) and, in some cases, in
air-conditioned rooms.
The handling of ever-smaller packages presents a problem for the manufacturing engineer. With a pin spacing of only 0.4 mm,
such as that attained with the thin shrink small-outline packages (TSSOP), exceptional demands are placed on soldering
techniques, such as the accuracy with which the components are placed in assembly and the precise control of the soldering
process. In the past, difficulty in controlling the soldering process often has been responsible for delaying the introduction of
smaller packages.
35
Although maximum permissible power dissipation of the small packages is of secondary significance only for the AHC
circuits, miniaturization of components obviously has reduced their ability dissipate heat. The relationships are explained in
Table 9. With AHC circuits in the middle-speed class, thermal impedance usually is of little importance; these devices have
an extremely low quiescent current drain. Also, in the frequency range up to about 10 MHz, in which these components should
be used, the dynamic power dissipation is kept within reasonable limits. In individual cases, for example, at high clock
frequencies or with the use of Widebus circuits, the system designer should calculate the power dissipation that can be expected
to prevent overloading of these components.
Table 9. Thermal Impedance of 20-Pin Packages
PACKAGE
PARAMETER UNIT
DIL SOP SSOP TSSOP TVSOP
Thermal impedance, θJA 67 96.6 104.2 148.9 179.5 °C/W
Figure 35 provides mechanical dimensions of the various packages in which AHC/AHCT families are available. This table
is not all inclusive because the many variants of different numbers of package pins from 14 to 56 cannot be shown in the space
available. The spectrum of available packages extends from the very well-known and much-used dual-in-line package (DIL),
through the well-established small outline package (SO) and up to the thin very small-outline package (TVSOP). With a pin
spacing of only 0.4 mm (16 mil) and a height of 1.2 mm, this package is ideal for use in chip cards.
24-Pin SSOP
24-Pin SOIC
Height = 2 mm
Height = 2.65mm 24-Pin SSOP Volume = 140 mm3
24-Pin SOIC Volume = 437 mm3 Area = 70 mm2 Lead pitch = 0.65 mm
Area = 165 mm2 Lead pitch = 1.27 mm
24-Pin TSSOP
48-Pin SSOP
Height = 1.1 mm
24-Pin TSSOP Volume = 59 mm3
Area = 54 mm2 Lead pitch = 0.65 mm
Height = 2.74 mm
48-Pin SSOP Volume = 469 mm3
24-Pin TVSOP
Area = 171 mm2 Lead pitch = 0.635 mm
Height = 1.2 mm
48-Pin TSSOP 24-Pin TVSOP Volume = 38 mm3
Area = 32 mm2 Lead pitch = 0.4 mm
36
6.1 Single-gate Logic
System designers often are confronted with the need for another gate or inverter to complete the design. The reason for this
additional component may be, for example, that a signal from one circuit can supply the logic level needed by the subsequent
circuit only after inversion. Or, at the last moment, it may be realized that the logical combination of two signals (AND, OR)
is needed to implement the required function. Finally, it can be determined that the input signal needs to be amplified or that
a Schmitt trigger is required to make an edge steeper so that a following circuit will operate properly.
In the past when this situation arose, it was necessary to incorporate an additional 14- or 16-pin package, which might have
been only 25% utilized. Besides the cost of this additional component, the space required becomes of great importance when
equipment and systems need to be miniaturized. To meet this need, the Microgate Logic and Picogate Logic packages have
been developed. The Microgate Logic circuits are supplied in a 5-pin SOT-23 package, and the Picogate Logic circuits in the
still smaller SC-70 package (Figure 35). The dimensions of Microgate Logic conform to those of the SOT-23, which has long
been used for small-signal transistors and has been extended with two additional pins. It should be emphasized that the 5-pin
SOT-23 package originally was introduced for use with analog circuits. In analog circuit practice there are far fewer
opportunities to construct circuits (such as amplifiers) with standardized components than is the case in digital circuitry where
all circuits are basically derived from gates or inverters. Because amplifiers or comparators are chosen for specific functions
in the application, the SOT-23 package containing the required function is the logical choice.
Because the SOT-23 package has only five pins, of which two need to be reserved for the supply voltage, the functions that
can be integrated into them are limited: AND, NAND, OR, NOR, EXOR gates, and inverters. Other functions, such as the
Schmitt trigger, are available that are particularly needed in interfaces. An available often-used function is the unbuffered
inverter, designated as ’04U (U = unbuffered). This device has applications in oscillators, and can be used as an analog
wideband amplifier.
7. Summary
With their advanced high-speed CMOS logic family, TI has created a series of components that combines the advantages of
many integrated circuits that are already well known, without having to accept many of their disadvantages:
• All CMOS circuits have low power requirements in common.
• Delay times have been much improved in comparison with HC devices.
• Values have been reached that were previously possible only with AC devices.
The high driveability of the latter family has not been incorporated – this is reserved for the AC, LVC, and ALVC families –
but instead they have been limited in this respect to values that are usual for high-speed CMOS. From the point of view of
interference that integrated circuits themselves generate, these components are easy to use. This ease of use extends from their
dynamic-power dissipation and low cross talk between signal lines to the precautions necessary to ensure the electromagnetic
compatibility of a circuit or system.
8. References
1. Texas Instruments, AHC/AHCT, HC/HCT, and LV CMOS Logic Data Book, literature number SCLD004.
2. Texas Instruments, Semiconductor Group Package Outlines Reference Guide, literature number SSYU001.
3. Fachverband Bauelemente der Elektronik: Messung der EME von integrierten Schaltungen (Professional
Association for Electronic Components: Measuring the EME of Integrated Circuits).
4. Texas Instruments, Digital Design Seminar, literature number SDYDE01A.
37
38
Appendix A
Product Portfolio
1
2
ADDITIONAL LITERATURE
For more information on the AHC product line, please visit:
http://www.ti.com/sc/ahc
http://www.ti.com/sc/littlelogic
If you would like additional AHC literature, please call 1-800-477-8924 and ask for the following items:
Design Considerations for Logic Products Application Book Volume 2 (1999) . . . . . . . . SDYA018
PRODUCT AVAILABILITY
Refer to the following codes for column entries on the following pages.