JP2008305429A - Non-volatile storage device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To easily connect chip selection signal output terminals of first semiconductor integrated circuit chips respectively to chip selection signal input terminals of corresponding non-volatile memory chips by means of bonding wires without being disturbed by other bonding wire. <P>SOLUTION: Terminals (chip selection signal input terminals CE) for inputting chip selection signals from the first semiconductor integrated circuit chips 33 are positioned at the end of an external terminal array, and a plurality of non-volatile memory chips 34a and 34b are laminated by being shifted alternately in the crossing directions. In this way, external terminals of the laminated non-volatile memory chips are individually exposed for wire bonding. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a technique for suppressing electrostatic breakdown (also referred to as ESD (Electrostatic Discharge) breakdown) of a semiconductor integrated circuit chip mounted on an IC card, and is effective when applied to a memory card such as a multimedia card. Regarding technology.

  Various memory cards that have been reduced in size and weight for multimedia data storage have been provided. For example, there is provided a multimedia card characterized in that a memory and a memory controller are mounted on a card substrate and can be interfaced with a host device with a small number of signals.

  Since this type of memory card prioritizes reduction in size and weight, a connection terminal connected to the host device is exposed from the card substrate, and a special terminal protective cover or other mechanism is not provided. Therefore, when the memory card is removed from the host device, if the exposed terminal is touched, the semiconductor integrated circuit chip connected to the exposed terminal may be destroyed. Normally, an input protection circuit for preventing electrostatic breakdown of the input circuit is integrated together in the semiconductor integrated circuit chip. The input protection circuit is configured, for example, by disposing an element such as a diode that is reversely connected to the input signal amplitude voltage of the input terminal between the power supply terminal. However, such a memory card is expected to be carried alone or frequently detached from the host device, and the present inventors have found the usefulness of enhancing prevention of electrostatic breakdown.

  Here, although the technical field is different from that of the memory card, there is a technique described in Japanese Patent Laid-Open No. 10-209379 as a technique for enhancing input protection against electrostatic breakdown. This is because when a metal wiring layer is formed with an interval (discharge gap) at which static electricity can be discharged from the electrode layer on the semiconductor substrate, when static electricity enters the electrode layer, the static electricity is transferred to the metal wiring layer. It is intended to prevent static electricity that has been discharged toward the electrode layer from entering the inside of the semiconductor element. Japanese Patent Application Laid-Open No. 7-271937 discloses a circuit that employs a gate-source electrode protection diode of an external MOSFET for preventing electrostatic breakdown in addition to a semiconductor integrated circuit chip.

  Also, varistors using semiconductor ceramics are provided from the viewpoint of overvoltage protection of various circuits.

JP-A-10-209379 JP 7-271937 A

    The present inventor conducted the following investigation from the viewpoint of enhancing prevention of electrostatic breakdown related to an IC card such as a memory card with exposed connection terminals.

  First, when a Zener diode or the like having a large element size is integrated on a semiconductor integrated circuit chip to obtain an energy resistance that is useful for strengthening prevention of electrostatic breakdown, the area efficiency deteriorates as the circuit element is miniaturized. It became clear that it would increase the cost.

  Secondly, when taking measures to prevent electrostatic breakdown by externally attaching an overvoltage protection element to a semiconductor integrated circuit chip, consider the relationship with the characteristics and capabilities of the overvoltage protection circuit built into the semiconductor integrated circuit chip. Otherwise, it has been clarified that prevention of electrostatic breakdown is not effective, and there is a possibility that the size and thickness of the IC card may be increased due to excessive or excessive external circuit elements. Such a viewpoint is not shown in the above prior art. In the present specification, the overvoltage means a surge voltage or a transient voltage generated electrostatically.

  Third, even if countermeasures to prevent electrostatic breakdown with external circuit elements are taken, there is no guarantee that they will be absolutely immune from destruction even if they are treated unexpectedly by the ignorance of the operator. It is necessary to.

  Fourth, even if the input circuit of the semiconductor integrated circuit chip is electrostatically destroyed, it can be assumed that only the data in the memory is safe. In such a case, data recovery of the memory card is possible. This is excellent in terms of relief, and the safety of the memory card as a storage medium can be increased.

  Fifth, if measures for strengthening the prevention of electrostatic breakdown with external circuit elements are taken, at least the amount of free space on the card substrate is reduced, and even in such a case, malfunction due to undesired leakage of signal lines It is also necessary to devise a technique that can avoid dense wiring patterns and dense bonding wires. This is also a necessary consideration when increasing the storage capacity of the memory card.

  An object of the present invention is to provide an IC card capable of enhancing prevention of electrostatic breakdown without increasing the cost of a semiconductor integrated circuit chip.

  Another object of the present invention is to provide an IC card capable of strengthening prevention of electrostatic breakdown by externally attaching an overvoltage protection element to a semiconductor integrated circuit chip without greatly changing the size and thickness of the card. It is in.

  Another object of the present invention is to provide an IC card that can be expected to prevent electrostatic breakdown due to unexpected handling due to ignorance of the operator.

  Another object of the present invention is to provide an IC card capable of easily recovering data in a memory card when the data in the memory is safe even if the input circuit of the semiconductor integrated circuit chip is electrostatically damaged. There is to do.

  Another object of the present invention is that even if the empty area on the card substrate is reduced by the countermeasure for preventing electrostatic breakdown by an external circuit element, the wiring pattern is densely bonded or bonded, which may cause malfunction due to undesired leakage of the signal line. An object of the present invention is to provide an IC card capable of avoiding crowding of wires.

  Still another object of the present invention is to provide an IC card having a relatively large storage capacity in a relatively small size.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  The following is a brief description of an outline of typical inventions disclosed in the present application.

  [1] Considering the relationship with the first overvoltage protection element integrated in the semiconductor integrated circuit chip, a second overvoltage protection element capable of enhancing prevention of electrostatic breakdown is externally attached to the semiconductor integrated circuit chip. That is, an IC card having a semiconductor integrated circuit chip on a card substrate and exposing a plurality of connection terminals, wherein the connection terminals are connected to predetermined external terminals of the semiconductor integrated circuit chip. A first overvoltage protection element connected to the external terminal is integrated on the chip, and a second overvoltage protection element connected to the connection terminal is mounted on the card substrate.

  At this time, according to the first aspect, the second overvoltage protection element is a variable resistance element having a current permissible capacity exceeding the first overvoltage protection element.

  According to the second aspect, an applied voltage required to flow a prescribed pulse current by applying a voltage larger than a rated voltage to the second overvoltage protection element is the regulation voltage for the first overvoltage protection element. It is a voltage that allows only a smaller amount of current to flow through.

  According to a third aspect, the second overvoltage protection element is a variable resistance element having a breakdown voltage larger than that of the first overvoltage protection element.

  According to a fourth aspect, the second overvoltage protection element is an element having a larger capacity than the first overvoltage protection element.

  As a result, the high-speed surge pulse can be bypassed with a low resistance.

  According to a fifth aspect, the breakdown voltage of the second overvoltage protection element is smaller than the breakdown voltage of the first overvoltage protection element.

  According to a sixth aspect, the breakdown voltage of the second overvoltage protection element is smaller than the breakdown voltage of the circuit protected by the first overvoltage protection element.

  In any of the above viewpoints, the second overvoltage protection element takes into consideration the relationship with the characteristics and capabilities of the first overvoltage protection element incorporated in the semiconductor integrated circuit chip. The effect can be effective.

  The second overvoltage protection element may have one end connected to the power supply connection terminal of the card substrate and the other end connected to the signal connection terminal. This signal connection terminal is connected to a corresponding external terminal of the semiconductor integrated circuit chip. At this time, the signal propagation distance from the signal connection terminal to the corresponding second overvoltage protection element is shorter than the signal propagation distance from the signal connection terminal to the corresponding external terminal of the semiconductor integrated circuit chip. . Thereby, it is possible to prevent the semiconductor integrated circuit chip from being directly affected by the destructive effect of the overvoltage before the second overvoltage protection element functions due to the overvoltage.

  The second overvoltage protection element may be a surface mount type varistor mainly composed of semiconductor ceramics, a chip diode array, a chip capacitor, or a chip transistor. This makes it possible to reduce the mounting area or occupied area of the second overvoltage protection element. Manufacturing costs can be reduced by surface mounting.

  When a memory card such as a multimedia card is assumed as an IC card, the semiconductor chip is a controller chip, and one or more memory chips (for example, a nonvolatile memory chip) connected to the controller chip are further included in the card. Mounted on the board. The controller chip has a memory control function for controlling a read / write operation with respect to the memory chip in accordance with an instruction from the outside.

  When considering data security or copyright protection, the controller chip encrypts the data to be written to the memory chip and decrypts the data read from the memory chip. A function may be adopted.

  In consideration of prevention of electrostatic breakdown even in the manufacturing process of an IC card, a second overvoltage protection element connected to the connection terminal is first mounted on the card substrate, and then the semiconductor integrated circuit is connected to the connection terminal. A predetermined external terminal of the chip may be connected. Thus, the second overvoltage protection element can be protected in the step of connecting the semiconductor integrated circuit chips.

  [2] A semiconductor integrated circuit chip is provided on the card substrate, a plurality of connection terminals are exposed, predetermined external terminals of the semiconductor integrated circuit chip are connected to the connection terminals, and the semiconductor integrated circuit chip is connected to the semiconductor integrated circuit chip. A first overvoltage protection element connected to an external terminal is integrated, and the card board is mounted with a second overvoltage protection element connected to the connection terminal. The second overvoltage protection element is a card board. The conductive pattern may be connected to the conductive pattern by surface mounting. The mounting cost of the second overvoltage protection element can be reduced.

  When a memory card such as a multimedia card is assumed as an IC card, the semiconductor chip is a controller chip, and one or more memory chips connected to the controller chip are further mounted on the card substrate. Become. At this time, a bonding wire may be used for connection between the connection terminal and the external terminal of the controller chip, and a bonding wire may be used for connection between the controller chip and the memory chip. Thereby, a large number of wiring patterns having the same function as the connection by the bonding wires need not be formed densely on the card substrate. The space above the controller chip and memory chip can be used for wiring. Therefore, it can contribute to the cost reduction of the card substrate.

  When a plurality of memory chips are connected in parallel to the controller chip with bonding wires, the memory chip is shifted in position so that each external terminal is exposed from the viewpoint of shortening the length of the bonding wire. It is good to mount in the said card | curd board | substrate in the state piled up. Thereby, compared with the case where each memory chip is not stacked, the distance from the controller chip is shortened, and the length of the bonding wire is shortened. Therefore, the possibility of undesired contact and disconnection of the bonding wire can be reduced. In particular, at this time, it is preferable to maintain a condition that the surface area of the one surface of the card substrate is larger than the total area of the memory chip and the controller chip. This is a consideration for ensuring that the card board has a sufficient space enough to cope with the constraint that the wiring layer is formed only on one surface of the card board. The idea is different from simply stacking memory chips in order to reduce the area of the card substrate.

  [3] With respect to an IC card in which a plurality of memory chips and a controller chip for controlling the memory chips are mounted on one surface of the card substrate, the memory chips are shifted in position so as to expose respective external terminals. When mounted on the card substrate in a stacked state, the external terminals of the memory chips that receive the same signal from the controller chip are sequentially connected in series with bonding wires. A bonding technique such as so-called stitch sewing is employed. The bonding wire can be shortened as a whole as compared with the case where the controller chip is connected to each external terminal separately by bonding wires, and also in this respect, the possibility of undesired contact and disconnection due to the dense bonding wires can be reduced. .

  The present invention also relates to an IC card in which a plurality of memory chips and a controller chip for controlling the memory chip are mounted on one surface of a card substrate, and the memory chips are shifted in position so as to expose respective external terminals. When mounted on the card substrate in a stacked state, the external chip for inputting the chip selection signal of the memory chip is positioned at the end of the external terminal array of the nonvolatile memory chip, and the controller chip is separately bonded with a bonding wire. Connect to. In a configuration in which a plurality of memory chips are to be selected separately, the external terminals for chip selection signal input must be connected separately to the external terminals for chip selection signal output of the controller chip. Although such a technique cannot be adopted, the external terminal for chip selection is arranged at the end of the memory chip, so that it becomes easy to make a necessary connection without being obstructed by other bonding wires.

  [4] Regarding the arrangement of memory chips, controller chips and the like mounted on the card substrate, a row arrangement may be adopted. That is, the memory chip is connected to the controller chip, the connection terminal formed on the card substrate is connected to a predetermined external terminal of the controller chip, and the controller chip has a first overvoltage protection element connected to the external terminal. A second overvoltage protection element that is integrated and connected to the connection terminal is mounted on the card substrate. Then, the distance away from the connection terminal is increased in the order of the second overvoltage protection element, the controller chip, and the plurality of memory chips, and these are arranged in a row from one side of the card substrate to the opposite side. . With this arrangement, the second overvoltage protection element for finally releasing the overvoltage is closest to the connection terminal which is the overvoltage application terminal, the memory chip storing the data is the farthest, and the electrostatic breakdown of the semiconductor chip is prevented. And high reliability in terms of data protection.

  Also in this case, as described above, the memory chip may be mounted on the card substrate in a state where the memory chips are stacked while being shifted in position so that the external terminals are exposed.

  The arrangement of memory chips, controller chips and the like mounted on the card substrate is not limited to a row arrangement. When the plurality of connection terminals are arranged along one of the two adjacent sides of the card substrate, a memory controller is arranged along the longitudinal direction of the other side of the two adjacent sides, The plurality of memory chips are arranged in parallel in a direction substantially perpendicular to the arrangement direction of the connection terminals. A connection terminal exposed from the card substrate is connected to a predetermined external terminal of the controller chip, a first overvoltage protection element connected to the external terminal is integrated in the controller chip, and the memory chip is integrated in the controller chip. Connected. According to the layout configuration in which the connection terminals and the controller chip are arranged close to the two adjacent sides of the card substrate, it is easy to increase the mounting density or the mounting number of the memory chips. The memory chips may be arranged in parallel in a state where the memory chips are divided into a first group in which a plurality of sheets are stacked in a state shifted to expose each external terminal, and a second group in which a plurality of sheets are similarly stacked. It becomes easy to suppress the height of the IC card. A second overvoltage protection element connected to the connection terminal may be mounted on the card board along the arrangement direction of the connection terminals.

  [5] When an IC card is formed by forming a conductive pattern on both sides of a card substrate, generally a through hole penetrating the card substrate can be used for connection of the conductive pattern. At this time, the through hole may be disposed outside a mold region that covers the other surface of the card substrate together with the semiconductor integrated circuit chip. When molding is performed by applying pressure, it is possible to eliminate the possibility that the mold resin leaks to the back side of the card substrate through the through hole.

  When forming a through hole in the connection terminal exposed from the IC card, the through hole may be formed at a position biased with respect to the sliding surface of the connection terminal. As a result, even if the IC card is removed from the mounting slot, the slot terminals do not slide into the through holes, and no mechanical force is applied, so the connection terminal pattern may crack from the through holes. This can prevent the possibility of damage.

  A plurality of connection terminals are exposed on one surface of the card substrate, a semiconductor integrated circuit chip is mounted on the other surface of the card substrate, a predetermined external terminal of the semiconductor integrated circuit chip is connected to the connection terminal, and the semiconductor integrated circuit The circuit chip is integrated with a first overvoltage protection element connected to the external terminal, and the second overvoltage protection element connected to the connection terminal is mounted on the other surface of the card substrate. The other surface of the card substrate together with the integrated circuit chip and the second overvoltage protection element may be covered with a metal cap. The metal cap can be formed by sheet metal drawing, forging, or die casting. Thereby, compared with a resin cap, it becomes a measure against EMI (Electro Magnetic Interference), and sealing by mechanical tightening and high temperature cap sealing are also possible. Also in the resin cap, an electromagnetic wave absorbing material such as ferrite can be mixed. As ESD countermeasures, conductive particles such as carbon can be mixed.

  In order to mitigate the influence of electrostatic discharge generated in the vicinity of the card substrate, a conductive shield pattern may be adopted for the card substrate. That is, an IC card in which a plurality of connection terminals are exposed on one surface of a card substrate and a semiconductor integrated circuit chip is mounted on the other surface of the card substrate, the connection terminals being a predetermined external portion of the semiconductor integrated circuit chip. A first overvoltage protection element connected to the external terminal is integrated on the semiconductor integrated circuit chip, and a second overvoltage protection element connected to the connection terminal is mounted on the other surface of the card substrate. A conductive shield pattern is formed on one surface of the card substrate in a region excluding the connection terminal, and the conductive shield pattern is connected to the connection terminal for supplying ground power, or is not in contact with any connection terminal. To do. The conductive shield pattern disperses static electricity.

  [6] From the viewpoint of prevention of electrostatic breakdown when the IC card handler receives unexpected handling due to ignorance, etc., an IC card having a semiconductor integrated circuit chip mounted with a plurality of connection terminals exposed. A display for clearly indicating the position of the IC card to be held with a finger (for example, a display of a finger shape printed at a position to be held by the finger at the time of attachment / detachment) is provided on the surface. In addition, a cautionary note is provided on the surface of the IC card to prevent touching the connection terminal. Further, a cautionary note is provided on the packaging material that wraps the IC card so as not to touch the connection terminal of the IC card.

  [7] An IC card focused on the recovery of stored data has a plurality of connection terminals exposed, a plurality of memory chips and a controller chip for controlling the memory chips mounted on a card substrate, and the connection terminals Is connected to an external terminal of the first group of the controller chip, the memory chip is connected to an external terminal of the second group of the controller chip, and a data evaluation terminal connected to the external terminal of the second group is connected to the card It is formed on the substrate.

  According to the above, when the memory control operation is disabled due to electrostatic breakdown or the like, the memory chip can be directly controlled from the outside via the data evaluation terminal. Thus, even if the controller chip is destroyed, if data remains in the memory chip, it can be easily recovered.

  A control terminal for controlling output terminals included in the second group of external terminals of the controller chip to a high output impedance state may be further provided on the card substrate. When the destroyed controller chip is brought into an undesired signal output state, this can be easily solved.

  The controller chip may have a security function that encrypts data written to the memory chip and decrypts data read from the memory chip. In this case, the data is recovered by decrypting the data read from the memory chip by the IC card maker or a person who has obtained the permission.

  [8] The simplest method of data recovery for the IC card having the data evaluation terminal includes a first process for making the memory chip uncontrollable by the controller chip, and a memory from the data evaluation terminal to the memory. And a second process of reading data by controlling the chip. A data recovery method when assuming that the controller chip has the security function includes a first process for controlling an output terminal included in the second group of external terminals of the controller chip to a high output impedance state, A second process for reading data by controlling the memory from the data evaluation terminal; a third process for decoding the data read in the second process; and a second process for writing the data decoded in the third process to another IC card. 4 processes.

  As a result, even if the input circuit of the semiconductor integrated circuit chip is electrostatically destroyed, if the data in the memory is safe, the data in the memory card can be easily recovered.

  The effects obtained by the representative ones of the inventions disclosed in the present application will be briefly described as follows.

  That is, it is possible to provide an IC card that can enhance the prevention of electrostatic breakdown without increasing the cost of the semiconductor integrated circuit chip.

  Without greatly changing the card size and thickness of the IC card, an overvoltage protection element can be externally attached to the semiconductor integrated circuit chip to enhance prevention of electrostatic breakdown.

  It can also be expected to prevent electrostatic breakdown of the IC card due to unexpected handling due to the ignorance of the handler.

  Even if the input circuit of the semiconductor integrated circuit chip is electrostatically destroyed, if the data in the memory is safe, an IC card that can easily recover the data in the memory card can be provided.

  Even if the vacant area on the IC card card board is reduced by measures to prevent electrostatic breakdown with external circuit elements such as varistors, dense wiring patterns and bonding wires that cause malfunction due to undesired leakage of signal lines Avoid crowding.

  An IC card having a relatively small size and a relatively large storage capacity can be realized.

<Strengthening the function of preventing electrostatic breakdown with varistors>
First, a principle configuration for enhancing the electrostatic breakdown preventing function for a semiconductor integrated circuit by an external circuit element such as a varistor will be described.

  FIG. 1 shows an example of an IC card according to the present invention with respect to one connection terminal. The IC card shown in the figure has a semiconductor integrated circuit chip 2 on a card substrate 1 and exposes connection terminals 3 shown representatively. The connection terminal 3 is an interface terminal for electrically connecting the IC card to a host device to / from which the IC card is attached / detached.

  The connection terminal 3 is connected to a predetermined external terminal 4 of the semiconductor integrated circuit chip 2. The external terminal 4 is, for example, an input terminal, and is connected to, for example, a CMOS inverter at the first stage of the input circuit via the signal line 5. The CMOS inverter includes a p-channel field effect transistor (also referred to as a unit MOS transistor) Q1 and an n-channel MOS transistor Q2 arranged in series between the circuit ground terminal Vss and the power supply terminal Vcc. The semiconductor integrated circuit chip 2 is integrated with diodes 7 and 8, a thyristor 9 and a clamp MOS transistor Q 5 as a first overvoltage protection element connected to the external terminal 4, and is connected to the connection terminal 3 on the card substrate 1. A varistor 11 is mounted as a second overvoltage protection element. The diodes 7 and 8, the thyristor 9, and the clamp MOS transistor Q5 constitute an input protection circuit 6.

  Note that the sword and drain of the MOS transistor are relatively determined according to the direction of the operating voltage, but in this specification, for the sake of convenience, the name determined in the normal operating state by the operating power supplies Vss and Vcc is used as the terminal name.

  The anode of the diode 7 is connected to the input signal line 4, the cathode is connected to the power supply terminal Vcc, the cathode of the other diode 8 is connected to the input signal line 4, and the anode is connected to the ground terminal Vss. The thyristor 9 is equivalently composed of a pnp transistor Q3 and an npn transistor Q4, and has an anode connected to the input signal line 4 and a cathode connected to the ground terminal Vss. The MOS transistor Q5 is a so-called diode-connected clamp MOS transistor having a gate and a source connected to the ground terminal Vss and a drain connected to the input signal line 4.

  Reference numerals 12 and 13 denote input protection resistors. Q6 and Q7 are so-called diode-connected p-channel clamp MOS transistors and n-channel clamp MOS transistors having gates and sources connected to each other. The clamp MOS transistors Q6 and Q7 are circuit elements having an auxiliary function to cope with an overvoltage leaking from the input protection circuit 6, and cannot be a first overvoltage protection element by itself. It can be an overvoltage protection element in cooperation with other circuit elements.

  A signal having a voltage amplitude between the ground voltage Vss and the power supply voltage Vcc is input to the connection terminal 3 in a normal state. At this time, the diodes 7 and 8, the thyristor 9, and the clamp MOS transistors Q5, Q6, and Q7 are all reversely connected.

  When a positive overvoltage is applied to the connection terminal 3 due to electrostatic discharge or the like, the diode 7 is in the forward connection state, and the anode of the thyristor 9 is turned on exceeding the forward element voltage. The overvoltage flows into the power supply voltage Vcc and the ground voltage Vss, and transmission to the subsequent stage is prevented or alleviated. Even if the positive overvoltage leaks slightly, the clamp MOS transistor Q6 is turned on and tries to escape to the power supply voltage Vcc.

  On the other hand, when a negative overvoltage is applied to the connection terminal 3 by electrostatic discharge or the like, this time, the diode 8 is in the forward connection state, and the clamp MOS transistor Q5 is turned on. By flowing into the ground voltage Vss, transmission to the subsequent stage is prevented or alleviated. Even if the negative overvoltage slightly leaks, the clamp MOS transistor Q7 is turned on and tries to escape to the ground voltage Vss.

  The varistor 11 is a circuit element set so that the overvoltage element operation can be started before the overvoltage element operation of the input protection circuit 6 reaches the limit, and an attempt is made to strengthen the electrostatic breakdown prevention function or the overvoltage protection function. To do. The varistor 11 can be equivalent to or replaced with a circuit in which a Zener diode or the like is connected back-to-back.

  Here, a multilayer chip varistor using a semiconductor ceramic is adopted as the varistor 11. The varistor 11 has a small chip shape that can be surface-mounted as illustrated in the axial sectional view of FIG. 2, and has conductive side electrodes 20 and 21 at both ends. A pair of interlayer electrodes 22, 23 are provided toward the side electrode 21, and another interlayer electrode 24 positioned between the pair of interlayer electrodes 22, 23 is provided on the other side electrode 21. The space between the side electrodes 20 and 21 and the interlayer electrodes 22, 23 and 24 is filled with a semiconductor ceramic 25.

  FIG. 3 shows the characteristics of the varistor 11. The varistor 11 is a variable resistance element, and has the current-voltage (IV) characteristics shown in FIG. 3. In a normal normal use state, the varistor 11 operates at a leakage current of 50 μA or less and is actually used, that is, connected terminal The signal input from 3 is not affected. This state is obtained by using the device at a rated voltage (also referred to as a working voltage) Vwm or less described in data sheets of varistors provided as various devices. When an abnormally high voltage starts to enter the semiconductor integrated circuit chip, the input protection circuit inside the semiconductor integrated circuit chip starts to work at a relatively low voltage, but the current of an overvoltage protection element such as a diode of the input protection circuit against the overvoltage. Insufficient capacity and current saturates. As a result, the operating current of the varistor 11 begins to flow. Then, when the overvoltage reaches near the breakdown voltage (Vb), the resistance becomes low so that the voltage becomes almost constant regardless of the current, and the clamp voltage (Vc) is theoretically increased for a transient voltage higher than that. The electrostatic breakdown of the semiconductor integrated circuit chip is prevented with a high energy allowable value.

  For example, an IC card to which an varistor 11 having an energy resistance level of 0.2 J (joule) is externally attached to 10 kV (10 kV) far exceeding several hundred volts to 2 kV, which is the electrostatic breakdown resistance of the built-in semiconductor integrated circuit chip. Is assumed to be applied at 1000 A for 10 nanoseconds (10 nS). The amount of energy at this time is 10 kV × 1000 A × 10 nS = 0.1 J, and this amount of energy is below the energy resistance level of the varistor 11, so that electrostatic breakdown is prevented.

  In FIG. 3, the clamp voltage Vc is defined as a terminal voltage (voltage between side electrodes) when a specified pulse current, for example, 1A is allowed to flow for 8.20 seconds, and the breakdown voltage Vb is specified as a terminal voltage when a current of 1 mA is applied, for example. be able to. To explain qualitatively, the breakdown voltage Vb can be defined as a voltage that can easily maintain the reversibility of the IV characteristics even when a direct current is applied within that range. The clamp voltage Vc can be defined as a voltage that has a very high possibility of being destroyed or exceeding destruction several times.

  The characteristics of the varistor 11 can be defined as follows in consideration of the characteristics of the overvoltage protection element of the input protection circuit 6.

  First, the varistor 11 may be defined as a variable resistance element having a current permissible capacity that exceeds the diodes 7 and 8, the thyristor 9, the clamp MOS transistor Q5, and further the clamp MOS transistors Q6 and Q7 of the input protection circuit 6. it can.

  Second, the voltage applied to the varistor 11 to apply a specified pulse current by applying a voltage higher than the rated voltage, for example, the breakdown voltage Vb or a voltage in the vicinity thereof is a diode of the input protection circuit 6. For the transistors 7, 8, the thyristor 9, the clamp MOS transistor Q5, and further the clamp MOS transistors Q6, Q7, if the voltage is not destroyed, it is a voltage that allows a current smaller than the prescribed pulse current to flow.

  Third, the varistor 11 is a variable resistance element having a breakdown voltage larger than that of the diodes 7 and 8, the thyristor 9, the clamp MOS transistor Q 5, and further the clamp MOS transistors Q 6 and Q 7 of the input protection circuit 6.

  Fourth, the varistor 11 is a variable resistance element having a stray capacitance larger than that of the diodes 7 and 8 of the input protection circuit 6, the thyristor 9, the clamp MOS transistor Q5, and further the clamp MOS transistors Q6 and Q7. As understood from the structure of FIG. 2, when used for a power supply terminal, it is clear that semiconductor ceramics is not a dielectric but has a relatively large capacitive component. Since such a stray capacitance component acts so as to mitigate changes in the transient voltage, the larger one is useful for preventing electrostatic breakdown. When used as a signal terminal, it is necessary to reduce the capacity within an allowable range so that it can respond to a high-speed signal.

  Fifth, the breakdown voltage of the varistor 11 is smaller than the breakdown voltage of the diodes 7 and 8, the thyristor 9, the clamp MOS transistor Q5, and further the clamp MOS transistors Q6 and Q7 of the input protection circuit 6. The varistor 11 can break down before the input protection circuit 6 is broken, and the overvoltage can be released.

  Sixth, the breakdown voltage of the varistor 11 is a circuit protected by the diodes 7 and 8, the thyristor 9, the clamp MOS transistor Q5, and further the clamp MOS transistors Q6 and Q7 of the input protection circuit 6, for example, the MOS transistors Q1 and Q2. It is smaller than the breakdown voltage of the CMOS inverter circuit consisting of

  As described above, the varistor 11 takes into account the relationship between the characteristics and capabilities of the overvoltage protection elements such as the diodes 7 and 8 that constitute the circuit 6 of the input protection built in the semiconductor integrated circuit chip 2. Thus, the electrostatic breakdown preventing effect by the varistor 11 can be made effective.

  The signal propagation distance from the connection terminal 3 to the varistor 11 is shorter than the signal propagation distance from the connection terminal 3 to the corresponding external terminal 4 of the semiconductor integrated circuit chip 2. Thereby, it is possible to prevent the semiconductor integrated circuit chip 2 from being directly affected by the destructive effect due to the overvoltage before the varistor 11 functions due to the overvoltage.

  Since the varistor 11 is a surface mount type varistor mainly composed of semiconductor ceramics, the mounting area or occupied area of the varistor 11 can be reduced. This surface mounting can reduce the manufacturing cost of the IC card.

<Application to multimedia card>
Next, a specific example in which an IC card using the varistor 11 is applied to a multimedia card will be described.

  FIG. 4 illustrates the connection mode of the varistor to the connection terminal of the multimedia card. The multimedia card has a card size of 24 mm × 32 mm × 1.4 mm according to the specifications by its standardization organization. On the card substrate 1, as connection terminals, a connection terminal 3a for inputting a chip select signal CS, a connection terminal 3b for inputting a command CMD, a connection terminal 3c for inputting a clock signal CLK, a connection terminal 3d for inputting / outputting data DAT, It has a connection terminal 3e to which a power supply voltage Vcc is supplied and two connection terminals 3f and 3g to which a ground voltage Vss is supplied. These connection terminals 3a to 3g are connected to a controller chip and a non-volatile memory chip (not shown) mounted in the region indicated by 30 in the figure. Note that the arrangement of the connection terminals 3a to 3g in FIG. 4 is different from that of an actual multimedia card.

  Varistors 11a to 11e are mounted on the card substrate 1 between the connection terminals 3a to 3e and the connection terminals 3e and 3g, respectively. In FIG. 4, one varistor is provided for each corresponding terminal, but a plurality of varistors may be connected in series.

  In particular, since the varistor 11e disposed at the connection terminal 3e that receives the power supply voltage Vcc functions as a bypass capacitor, the varistor 11e may be replaced with a bypass capacitor, or a bypass capacitor 31 may be further provided in parallel as shown in FIG. .

  FIG. 5 is a plan view illustrating the configuration of the multimedia card mainly in the circuit element mounting state. FIG. 6 is a longitudinal sectional view thereof. The card substrate 1 is made of glass epoxy resin or the like, and the connection terminals 3a to 3g are formed on the back surface of the card substrate 1 with conductive patterns. The varistors 11a to 11e, the controller chip 33, and the non-volatile memory chips 34a and 34b are mounted on the surface of the card substrate 1 through wiring patterns and conductive patterns. In the figure, 36 is a conductive pattern connected to the corresponding connection terminals 3a to 3g through the through hole 40, and 35 is a wiring pattern for connecting one end of the varistors 11a to 11e to the ground voltage Vss. The varistors 11a to 11e are mounted on the surface of the wiring pattern 35 and the connection terminals 3a to 3e.

  In the figure, 38 and 39 are bonding patterns, and 37 is a wiring pattern for connecting the corresponding bonding pattern 38 and the conductive pattern 36. The conductive pattern 38 and the corresponding external terminal 50 of the controller chip 33 are connected by a bonding wire 41, and the bonding pattern 39 corresponding to the external terminal 51 of the controller chip 33 is connected by a bonding wire 42. The connection between the bonding pattern 39 and the corresponding external terminal 52a of one nonvolatile memory chip 34a is connected by a bonding wire 43a, and the connection between the bonding pattern 39 and the corresponding external terminal 52b of the other nonvolatile memory chip 34b is a bonding wire 43b. Connected with. The semiconductor integrated circuit chip is a so-called bare chip, and the external terminals 50, 51, 52a, and 52b are bonding pads made of aluminum, aluminum alloy, copper, or the like.

  The nonvolatile memory chips 34a and 34b are, for example, electrically rewritable flash memory chips. The flash memory chip has a memory cell array in which nonvolatile memory cell transistors having a control gate, a floating gate, a source and a drain are arranged in a matrix, for example, and according to commands and addresses supplied from the outside, data read, erase, write, Operations such as verification are performed. The non-volatile memory chips 34a and 34b formed of the flash memory chip have, as external terminals 52a and 52b, a chip enable signal (also referred to as a chip selection signal) for instructing chip selection / CE input terminal, and a write enable for instructing a write operation. Command data indicating whether the input terminal of signal / WE, input / output terminals I / O0 to I / O7, input / output terminals I / O0 to I / O7 are used for input / output of command data or address input Enable signal / CDE input terminal, output enable signal / OE input terminal for instructing output operation, clock signal / SC input terminal for instructing data latch timing, ready / busy signal for instructing externally whether write operation is in progress It has an R / B output terminal and an input terminal for a reset signal / RES.

  The controller chip 33 controls the read / write operation with respect to the nonvolatile memory chips 34a and 34b in accordance with an instruction from the outside, and further considers data security or copyright protection to the nonvolatile memory chips 34a and 34b. It has a security function for encrypting data to be written and decrypting data read from the nonvolatile memory chips 34a and 34b.

  The external terminal 50 of the controller chip 33 corresponds to the input / output functions of the connection terminals 3a to 3g, and serially inputs a select signal CS input terminal for instructing the selection operation of the multimedia card and a command CMD instructing the operation of the multimedia card. An input terminal for input, an input terminal for a clock signal CLK regarded as a synchronization signal for signal input / output operation of the external terminal 50, a terminal for serially inputting / outputting data DAT, and an input terminal for the power supply voltage Vcc and the ground voltage Vss. . In the controller chip 33, the input protection circuit 6 and the clamp MOS transistors Q6 and Q7 described with reference to FIG.

  As external terminals 51 for memory access in the controller chip 33, there are output terminals for a chip selection signal / CE0 for the nonvolatile memory chip 34a, and output terminals for a chip selection signal / CE1 for the nonvolatile memory chip 34b. The external memory chip 34a, 34b has an external terminal corresponding to the external terminal and whose input / output direction is reversed.

《Bonding wire connection》
As described above, the bonding wires 41 are used to connect the connection terminals 3a to 3g and the external terminals 50 of the controller chip 33, and the bonding wires 43a and 43b are used to connect the controller chip 33 and the nonvolatile memory chips 34a and 34b. Therefore, a large number of wiring patterns having the same function as the connection using the bonding wires need not be formed densely on the card substrate 1. The space above the controller chip 33 and the non-volatile memory chips 34a and 34b can be used for wiring. In short, the substrate wiring can be simplified by the aerial wiring of the bonding wires. Therefore, it can contribute to the cost reduction of the card substrate 1.

《Overlay implementation》
In the configuration of FIG. 5, two nonvolatile memory chips 34a and 34b are connected in parallel to the controller chip 33 by bonding wires. At this time, the non-volatile memory chips 34a and 34b are mounted on the card substrate 1 in a state where the non-volatile memory chips 34a and 34b are stacked with their positions shifted so that the external terminals 52a and 52b are exposed. Thereby, compared with the case where each non-volatile memory chip 34a, 34b is arrange | positioned without overlapping, distance with the controller chip 33 becomes short, and the routing length of bonding wire 43a, 43b becomes short. Therefore, the possibility of undesired contact and disconnection of the bonding wire can be reduced. The amount of shift when stacking a plurality of nonvolatile memory chips may be determined within a range in which one lower layer chip can exist under the bonding external terminal of the upper layer chip. This is because if there is no lower layer chip under the bonding external terminal, there is a risk of chip damage due to mechanical force during bonding.

  In particular, at this time, the condition that the surface area of the one surface of the card substrate 1 is larger than the total area of the nonvolatile memory chips 34a and 34b and the controller chip 33 is satisfied. This is a consideration for ensuring that the card board 1 can have a sufficient margin enough to cope with the constraint that the wiring layer is formed only on one surface of the card board. The idea is different from simply stacking non-volatile memory chips in order to reduce the area of the card substrate 1.

<Linear layout>
In the example of FIG. 5, a row-like arrangement is adopted for the arrangement of the nonvolatile memory chips 34 a and 34 b and the controller chip 33 mounted on the card substrate 1. That is, in the order of the varistors 11a to 11e, the controller chip 33, and the plurality of memory chips 34a and 34b, the distance away from the connection terminals 3a to 3g of the multimedia card is increased, and these are opposed from one side of the card substrate 1. It is arranged in a row toward the side. With this row arrangement, the second varistors 11a to 11e for finally releasing the overvoltage are closest to the connection terminals 3a to 3g which are overvoltage application terminals, and the nonvolatile memory chips 34a and 34b storing the data are farthest away. Therefore, since it is effective in absorbing surge of high-speed pulses, it is high in terms of strengthening prevention of electrostatic breakdown of the controller chip 33 by the varistors 11a to 11e and protecting stored data of the nonvolatile memory chips 34a and 34b. Reliability can be obtained.

《Through hole deviation with respect to connection terminal》
As shown in FIG. 5, the through hole 40 is provided at a position deviated with respect to the connection terminals 3 a to 3 g. That is, as illustrated in detail in FIG. 7A, when the through hole 40 is formed in the connection terminal 3a exposed from the IC card, the through hole 40 is placed on the sliding surface of the connection terminal 3a. To form a biased position. The position to be biased may be (B) in FIG. Thereby, even if the IC card is removed from the mounting slot, the terminal 40A of the slot does not slide into the through hole 40, and no mechanical force is applied to the through hole 40. It is possible to prevent the pattern from being cracked from the through hole 40 or from being damaged by wear around the through hole hole.

《Through hole formation outside the mold area》
In FIG. 6, the controller chip 33 and the non-volatile memory chips 34a and 34b are molded with a thermosetting resin 55 as a whole. The varistor element can be brought into the mold or provided outside the mold. At this time, the through hole 40 is not included in the mold region of the thermosetting resin 55. Therefore, when molding is performed by applying pressure, it is possible to eliminate the possibility that the mold resin 55 leaks to the back side of the card substrate 1 through the through hole 40 and causes a molding defect.

《Metal cap》
In FIG. 6, the surface of the card substrate 1 on which the varistors 11 a to 11 e, the controller chip 33, and the nonvolatile memory chips 34 a and 34 b when the varistor is provided outside the mold is covered with a metal cap 56. Thereby, compared with a resin cap, it becomes a measure against EMI (Electro Magnetic Interference), and sealing by mechanical tightening and high temperature cap sealing are also possible.

  FIG. 25 shows several types of structures of the metal cap 56. (A) shows the case where it separates one by one and is manufactured by forging, and a slight step portion for labeling is also formed. (B) shows a case in which after forging, punching is performed to separate one by one. (C) shows the case of manufacturing with sheet metal drawing. (D) is the perspective view which looked at the metal cap manufactured by sheet metal narrowing down of (C) from the back. Since the corner portion is wrinkled during the drawing process, a notch is formed in advance.

《Stitch bonding》
FIG. 8 partially shows a multimedia card in which stitch bonding is applied to the connection of the nonvolatile memory chip. FIG. 9 is a longitudinal sectional view of a stitch bonding portion. Similar to FIG. 5, the non-volatile memory chips 34a and 34b are mounted on the card substrate 1 in a state where a plurality of the non-volatile memory chips 34a and 34b are stacked so as to expose the external terminals 52a and 52b. The external terminals 52 a and 52 b of the nonvolatile memory chip that receive the same signal from the controller chip 33 are sequentially connected in series by bonding wires 57. A bonding technique such as so-called stitch sewing, that is, stitch bonding is employed. As shown in FIG. 5, the bonding wire can be shortened as a whole as compared with the case where the controller chip 33 is connected to the external terminals 52a and 52b separately by the bonding wires 43a and 43b. The number of bonding wires can be reduced, and also in this respect, the possibility of undesired contact and disconnection due to the dense bonding wires can be reduced. Since chip selection for the non-volatile memory chips 34a and 34b must be performed separately, stitch bonding cannot be used for the bonding wires 43a and 43b for transmitting the chip selection signals / CE0 and / CE1, as shown in FIG. Has been left in the same bonding format.

  When performing stitch bonding, the bonding method for the bonding pad 52a differs depending on the bonding type of the wire bonder to be used. (A) in FIG. 10 shows a case where nail head bonding is used. At this time, because of the structure of the wire bonder, the end of the bonding wire is cut into a crescent shape, so the next bonding base point is set at a position different from the end point. It must be. Therefore, stitch bonding is completed by the bonding wires 57 and 57 that are necessarily divided into a plurality of wires. On the other hand, FIG. 10B shows a case where wedge bonding is used. When a wire bonder that supports this is used, bonding can be performed one after another without cutting the bonding wire halfway. Therefore, according to this method, stitch bonding can be performed with one bonding wire 57.

  FIG. 11 is a plan view showing the configuration of a multi-media card having a four-stack structure of nonvolatile memory chips. FIG. 12 is a longitudinal sectional view thereof. Even in the case of four stacks, the non-volatile memory chips 34a to 34d are mounted on the card substrate 1 in such a state that the non-volatile memory chips 34a to 34d are stacked while being shifted in position so as to expose the external terminals 52a to 52d. At this time, the external terminals 52a to 52d of the nonvolatile memory chips 34a to 34d that receive the same signal from the controller chip 33 are sequentially connected in series by the bonding wires 60 as in the case of the stitch bonding. Since chip selection for the nonvolatile memory chips 34a to 34d must be performed separately, stitch bonding is not used for the bonding wires 43a to 43d for transmitting the chip selection signals / CE0 to / CE3. Has been left in the same bonding format. However, if the chip selection signal is converted into an ID command, stitch bonding can be performed.

《Shield pattern》
In the configuration of FIG. 11, the conductive shield pattern 61 shown in FIG. 12 is adopted in the card substrate 1 in order to further alleviate the influence of electrostatic discharge generated in the vicinity of the card substrate 1. That is, a wide conductive shield pattern 61 is formed on the exposed surface of the connection terminals 3 a to 3 g on the card substrate 1. Although the conductive shield pattern 61 is not particularly limited, it may be connected to the connection terminals 3f and 3g for supplying the ground power source Vss or may be left floating because it has a relatively large surface area. The conductive shield pattern 61 can disperse static electricity.

《CS input terminal at chip end》
As shown in FIG. 11, in a structure in which a plurality of nonvolatile memory chips 34a to 34d are stacked while being shifted, chip selection signals / CE0 to / CE3 among the external terminals 52a to 52d of the nonvolatile memory chips 34a to 34d. Are connected to the external terminals 51 of the controller chip 33 by bonding wires 43a to 43d, respectively, at the ends of the external terminal arrays of the nonvolatile memory chips 34a to 34d. The stack structure in FIG. 8 is exactly the same. As shown in FIGS. 8 and 11, in a configuration in which a plurality of nonvolatile memory chips are to be selected separately, an external terminal for inputting a chip selection signal among external terminals of the nonvolatile memory chip is a controller chip 33. The chip selection signal output external terminal 51 must be separately connected, and even if a technique such as the stitch bonding cannot be adopted for this portion, the chip selection signal output external terminal is used as described above. Since the terminals are arranged at the end of the nonvolatile memory chip, it is easy to take necessary connections without being obstructed by other bonding wires. The effect becomes more prominent as the number of stacks of nonvolatile memory chips increases. As shown in FIG. 5, when the number of stacks of the nonvolatile memory chips is two, it is easy to draw out two bonding wires in parallel from one bonding pattern by using a highly accurate wire bonder. In this case, even when stitch bonding is not employed, there is no problem in adopting a configuration in which an external terminal for chip selection signal input is disposed at the chip end.

<Multiple grouped implementation of stacked nonvolatile memory chips>
FIG. 13 shows still another example of the multimedia card. FIG. 14 is a partial longitudinal sectional view thereof. The multimedia card shown in FIG. 13 is a single-phase wiring in which two sets of non-volatile memory chips are stacked on the card substrate 1 and a wiring pattern and a bonding pattern are formed on only one surface together with the connection terminals. The card substrate 1 is configured to be used. In this structure, a so-called COB (Chip On Board) structure in which a semiconductor bare chip is mounted on a substrate is applied.

  In FIG. 13, the connection terminals 3a to 3g, the wiring patterns 35 and 37, the bonding pattern 38, the bonding patterns 39A and 39C, and the wiring pattern 39B are all formed on the mounting side of the card substrate 1. The connection terminals 3a to 3g and the wiring pattern 35 are exposed to the surface from the opening formed in the card substrate 1 so that the varistors 11a to 11e can be connected. Similarly, the bonding patterns 38, 39A and 39C are also exposed to the surface from the openings formed in the card substrate 1, and are external terminals 50 and 51 of the controller chip 33 and external terminals 52a to 52d of the nonvolatile memory chips 34a to 34d. Bonding is possible. In FIG. 13, stitch bonding is performed to bond the bonding pattern 39A to the external terminals 52a and 52b of the nonvolatile memory chips 34a and 34b and to bond the bonding pattern 39C to the external terminals 52c and 52d of the nonvolatile memory chips 34c and 34d. However, stitch bonding similar to that shown in FIG. 8 may be employed except for the chip selection signal.

  When two sets of two non-volatile memory chips stacked as shown in FIG. 13 are mounted on the card substrate 1, the thickness dimension can be reduced as compared with the four-stack structure as shown in FIG. Accordingly, if two sets of four non-volatile memory chips stacked on the card substrate 1 are mounted on the card substrate 1, it is possible to obtain twice the storage capacity with the same thickness as the four-stack structure as shown in FIG.

  Further, when a plurality of connection terminals 3a to 3g are arranged along one side of the card substrate 1, a card controller 33 is arranged along the longitudinal direction along the adjacent side, and the connection terminals 3a to 3g are connected to each other. If the nonvolatile memory chips are arranged in parallel in a direction substantially perpendicular to the arrangement direction, the nonvolatile memory chips can be efficiently mounted on the plate surface of the card substrate 1.

  Each of the divided stack structure and the structure in which the connection terminals 3a to 3g and the controller chip 33 are arranged close to the two sides of the card substrate 1 increase the density of mounting the non-volatile memory chip on the card substrate of a prescribed size. Or, it becomes easy to increase the number of mounting.

  FIG. 15 shows an example of another multimedia card to which the divided stack structure and a structure in which connection terminals and a controller chip are arranged close to two adjacent sides of the card substrate are applied. The example shown in the figure is different from FIG. 13 in that a wiring pattern and a bonding pattern are formed on both sides of the card substrate 1 together with the connection terminals, and the direction of the stacked nonvolatile memory chips is aligned.

  In FIG. 15, the connection terminals 3a to 3g and the wiring pattern 39B are formed on the back surface of the card substrate, and the wiring patterns 35 and 37, the bonding pattern 38, and the bonding patterns 39A and 39C are formed on the surface of the card substrate 1. A through hole 40A is used to connect the wiring pattern 39B to the bonding patterns 39A and 39C. 15 does not employ the stitch bonding as in FIG. 13, but may employ the same stitch bonding as in FIG. 8 except for the chip selection signal.

  Similarly to FIG. 13, the multimedia card of FIG. 15 has a predetermined size because of the divided stack structure and the structure in which the connection terminals 3a to 3g and the controller chip 33 are arranged close to the two sides of the card substrate 1. It is easy to increase the density of mounting the nonvolatile memory chip on the card substrate or increase the number of mounting.

  At this time, the through hole in the mold may be structured such that the hole is filled with a conductive paste, solder resist, or the like to prevent mold resin leakage.

《Stack structure of memory chip and controller chip》
16 and 17 show an example in which a controller chip is placed on a memory chip and both are stacked. In FIG. 16, the external terminal 51 of the controller chip 33 is directly connected to the external terminal 52 of the nonvolatile memory chip 34 by bonding between the chips by the bonding wire 70, but the operation power sources Vss and Vcc to the nonvolatile memory chip 34 are connected. In order to reduce the feeding resistance, power wiring patterns 71A and 72A are formed on the back surface of the card substrate 1, and the bonding patterns 71B and 72B connected by the through holes 71D and 72D and the nonvolatile memory chip 34 are bonded to the bonding wires 71C and 72C. Connected with. However, when the power supply resistance to the nonvolatile memory chip 34 is sufficiently low, Vcc and Vss may be supplied from the terminal 51 and the terminal 52. The mounting structure and the like of the varistors 11a to 11e are the same as those described above, and circuit elements having the same functions as those in FIG.

  FIG. 17 illustrates a structure in which a memory chip and a controller chip are stacked using LOC (Lead On Chip). Reference numerals 73a to 73g respectively indicate part of leads of the lead frame for LOC. The leads 73e and 73f for the power supplies Vcc and Vss, for example, extend in a bowl shape to form pass bars 74A and 74B, respectively. The nonvolatile memory chip 34 is fixed to the path bars 74A and 74B, and the controller chip 33 is fixed to the nonvolatile memory chip 34. The external terminal 50 of the controller chip 33 is connected to the leads 73 a to 73 g by the bonding wire 4. The external terminal 51 of the controller chip 33 is directly connected to the external terminal 52 of the non-volatile memory chip 34 by bonding between the chips by bonding wires 70. The operation power supplies Vss and Vcc are supplied to the non-volatile memory chip 34. In order to reduce the feeding resistance, the path bars 74A and 74B and the nonvolatile memory chip 34 are connected by bonding wires 75 and 75, respectively. The varistors 11a to 11e are surface-mounted with a conductive paste such as an Ag paste between corresponding leads.

  FIG. 18 shows another example of an IC card to which the COB structure is applied. FIG. 19 is a longitudinal sectional view of the IC card, and FIG. 20 is an explanatory view of a conductive pattern formed on the bottom surface of the card substrate of the IC card of FIG. Conductive patterns 80a to 80g are formed on the bottom surface of the card substrate 84, and openings 81a to 81g are formed in the card substrate 84 correspondingly. The conductive patterns 80a to 80f constitute connection terminals exposed from the IC card. The semiconductor integrated circuit chip 83 is connected to the conductive pattern 80f through the opening 81g, and a ground voltage Vss is supplied as a substrate potential. Bonding pads 85a to 85f constituting external terminals of the semiconductor integrated circuit chip 83 are connected to connection electrodes 80a to 80f by bonding wires 86 through openings 81a to 81f. In the same manner as described above, the varistors 82a to 82e mainly composed of semiconductor ceramics are connected to the connection electrodes 80a to 80g and the conductive patterns 80g through the openings 81a to 81g in order to reinforce prevention of electrostatic breakdown of the semiconductor integrated circuit chip 83. In between, surface mounting is performed with a conductive paste such as an Ag paste.

《Notes》
FIG. 21 illustrates an IC card or the like having a warning for preventing electrostatic breakdown. As shown in FIG. 21A, the surface of an IC card such as the multimedia card on which a plurality of connection terminals are exposed and a semiconductor integrated circuit chip is mounted should not touch the connection terminals 3a to 3g. A warning 90 is provided, for example, “Do not touch the connection terminal”. In addition, a manufacturing management code may be entered in this area. Further, as illustrated in FIG. 21B, a display for clearly indicating the position of the IC card with the finger, for example, a caution display 91 in the form of a finger printed at the position of the finger when attaching / detaching is provided. . Furthermore, as illustrated in FIG. 21B, a cautionary note that prompts the packaging material 92 such as a laminated film, paper box, or plastic case that wraps the IC card not to touch the connection terminals 3a to 3g of the IC card. 93 may be provided.

  The notices 90 and 93 and the display 91 are useful for preventing the IC card from being electrostatically damaged due to unexpected handling due to ignorance of the IC card handler.

《IC card assembly method》
FIG. 22 shows a method of assembling the IC card shown in FIGS. First, a varistor is mounted on a predetermined conductive pattern of a card substrate such as a PCB substrate or a tape substrate (S1). Solder paste or silver paste is used for mounting. Thereafter, the paste is cured (baked) (S2), and a required number of semiconductor integrated circuit chips are die-bonded (die-bonded) to the conductive pattern on the card substrate (S3). Then, the surface of the card substrate is cleaned by plasma cleaning (S4). Thereafter, the bonding pad of the die-bonded semiconductor integrated circuit chip and the conductive pattern are bonded by thermal ultrasonic waves using a gold bonding wire (S5). Then, resin potting sealing is performed on the semiconductor integrated circuit chip and the bonding wire (S6), and the resin is cured by resin baking (S7). Finally, a metal cap is bonded and fixed to the card substrate from above, or Secure with plastic insert mold.

  As described above, since the varistor is first mounted on the card substrate and then die bonding or wire bonding of the semiconductor integrated circuit chip is performed, it can be protected by the varistor when the IC card is assembled, and the yield of the IC card can be increased. Can be improved. However, the varistor may be mounted later for the convenience of manufacturing such as temperature conditions.

<Data recovery terminal>
FIG. 23 illustrates an IC card that focuses on the viewpoint of data recovery. The basic configuration is the same as that in FIG. 5, and the difference is that it has a plurality of data recovery terminals. In FIG. 23, in order to emphasize the connection state of the data recovery terminal, the connection state between the controller chip 33 and the nonvolatile memory chips 34a and 34b is simplified. In FIG. 23, circuit elements having the same functions as those in FIG.

  Although not shown in FIG. 5, the controller chip 33 has an input terminal of a test signal / TEST pulled up internally as one of the external terminals 51 (also simply referred to as a test terminal / TEST). When the low level is input, the test terminal / TEST has an interface terminal with the nonvolatile memory chips 34a and 34b, in particular, an output terminal and an input / output terminal in a high output impedance state or an input / output operation disabled state. Control.

  The card substrate 1 is formed with data recovery terminals 92 connected to all external terminals 51 on the memory interface side of the controller chip 33 in a one-to-one correspondence by wiring 91. Similarly to the data recovery ground terminal 96 connected to the external terminal for ground power supply Vss by the wiring 95 among the external terminals 50 on the card interface side of the controller chip 33, similarly to the card interface side of the controller chip 33. A data recovery power supply terminal 94 connected to the external terminal for the power supply Vcc among the external terminals 50 by a wiring 93 is provided. In FIG. 23, reference numeral 90 denotes a guard ring added to the card substrate 1 for preventing electrostatic breakdown. The guard ring 90 circulates around the card substrate 1 and is connected to the ground power supply terminal of the circuit.

  Since the data evaluation terminals 92, 94, and 96 are formed on the card substrate 1, when the controller chip 33 is disabled from the memory control operation due to electrostatic breakdown or the like, the data evaluation terminals 92, 94 are externally provided. , 96, direct access control of the nonvolatile memory chips 34a, 34b can be performed. As a result, even if the controller chip 33 is destroyed, if data remains in the nonvolatile memory chips 34a and 34b, this can be easily recovered.

  When the controller chip 33 has a security function for encrypting data to be written to the nonvolatile memory and decrypting the data read from the nonvolatile memory, the data recovery is performed by an IC card. The manufacturer or a person who has obtained permission from the manufacturer decrypts the data read from the nonvolatile memory chip and recovers the data.

<Data recovery method>
FIG. 24 illustrates a data recovery processing procedure for an IC card having the data evaluation terminal.

  A multimedia card (MMC) or the like in which the controller has malfunctioned due to failure to avoid electrostatic breakdown by the input protection circuit or the varistor is set as a data recovery target (S10). In addition, it is possible to make an MMC data recovery target whose connection terminal is physically destroyed. First, the cap 56 is removed from the target MMC (S11), and a probe such as a tester is applied to the data evaluation terminals 92, 94, 96 (S12). Then, the input terminal of the test signal / TEST is fixed at a low level, and the memory interface terminal of the controller chip 33 is controlled to a high impedance state (a state where input / output operation is impossible) (S13). As a result, the nonvolatile memory chip with built-in MMC is released from the control of the controller chip, and can be directly accessed from the data evaluation terminals 92, 94, 96. In this state, data is read from the nonvolatile memory chip (S14). Here, the controller chip 33 has a security function for encrypting data to be written to the nonvolatile memory chip and decrypting data read from the nonvolatile memory. The data read without going through is decrypted. The decrypted data as described above is written into the new MMC through the connection terminals 3a to 3g as usual (S15). As a result, a new MMC whose data has been recovered is provided to the user (S16). At this time, the encryption specification of the controller chip can be determined by the manufacturing trace code of the card or the manufacturing code written in the nonvolatile memory.

  As a result, even if the input circuit of the controller chip 33 is electrostatically destroyed, if the data in the nonvolatile memory chips 34a and 34b is safe, the data in the memory card can be easily recovered.

<Flash memory chip>
Here, the flash memory chip will be described. FIG. 26 shows an example of a flash memory chip. In the figure, reference numeral 103 denotes a memory array having a memory mat, a data latch circuit, and a sense latch circuit. The memory mat 103 has many nonvolatile memory cell transistors that can be electrically erased and written. For example, as illustrated in FIG. 27, the memory cell transistor includes a source S and a drain D formed in a semiconductor substrate or a memory well SUB, a floating gate FG formed in a channel region via a tunnel oxide film, and a floating The control gate CG is configured to overlap the gate via an interlayer insulating film. The control gate CG is connected to the word line 106, the drain D is connected to the bit line 105, and the source S is connected to a source line (not shown).

  The external input / output terminals I / O0 to I / O7 are also used as address input terminals, data input terminals, data output terminals, and command input terminals. The X address signal input from the external input / output terminals I / O 0 to I / O 7 is supplied to the X address buffer 108 via the multiplexer 107. The X address decoder 109 decodes the internal complementary address signal output from the X address buffer 108 and drives the word line.

  A sense latch circuit (not shown) is provided at one end of the bit line 105, and a data latch circuit (not shown) is provided at the other end. The bit line 105 is selected by the Y gate array circuit 113 based on the selection signal output from the Y address decoder 111. The Y address signal input from the external input / output terminals I / O0 to I / O7 is preset in the Y address counter 112, and the address signal sequentially incremented from the preset value is supplied to the Y address decoder 111.

  The bit line selected by the Y gate array circuit 113 is conducted to the input terminal of the output buffer 115 during the data output operation, and is conducted to the output terminal of the input buffer 117 via the data control circuit 116 during the data input operation. Connection of the output buffer 115, the input buffer 117, and the input / output terminals I / O0 to I / O7 is controlled by the multiplexer 107. Commands supplied from the input / output terminals I / O 0 to I / O 7 are given to the mode control circuit 118 via the multiplexer 107 and the input buffer 117. The data control circuit 116 makes it possible to supply the memory array 103 with logical value data in accordance with the control of the mode control circuit 118 in addition to the data supplied from the input / output terminals I / O 0 to I / O 7.

  In the control signal buffer circuit 119, the chip enable signal / CE, the output enable signal / OE, the write enable signal / WE, the signal / SC for instructing the data latch timing, the reset signal / RES, and the command data are provided as access control signals. An enable signal / CDE is supplied. The mode control circuit 118 controls the signal interface function with the outside according to the state of these signals, and controls the internal operation according to the command code. In the case of a command or data input to the input / output terminals I / O0 to I / O7, the signal / CDE is asserted. If the command is input, the signal / WE is further asserted, and if it is data, the signal / WE is negated. If it is an address input, the signal / CDE is negated and the signal / WE is asserted. Thereby, the mode control circuit 118 can distinguish commands, data, and addresses that are multiplexed from the external input / output terminals I / O0 to I / O7. The mode control circuit 118 can assert the ready / busy signal R / B during an erase or write operation to notify the state to the outside.

  The internal power supply circuit 120 generates various operation power supplies 121 for writing, erasing, verifying, reading, etc., and supplies them to the X address decoder 109 and the memory cell array 103.

  The mode control circuit 118 controls the flash memory as a whole according to the command. The operation of the flash memory is basically determined by commands. Commands assigned to the flash memory are commands such as read, erase, and write.

  The flash memory has a status register 122 to indicate its internal state, and the contents can be read from the input / output terminals I / O0 to I / O7 by asserting the signal / OE.

  Although the invention made by the present inventor has been specifically described based on the embodiments, it is needless to say that the present invention is not limited thereto and can be variously modified without departing from the gist thereof.

  For example, the present invention can be applied to a memory card other than a multimedia card, such as a compact flash memory. In addition, a structure in which memory chips are shifted and stacked, a structure in which a through hole is biased with respect to a connection terminal of an IC card, a structure in which a through hole is formed outside a mold area, stitch bonding to a stacked semiconductor integrated circuit chip, The structure in which the CS input terminal at the end is arranged, the IC card in which the stacked nonvolatile memory is divided into a plurality of groups, the precautionary note, and the IC card having the data recovery terminal are not necessarily applied to a configuration having a varistor. Not. The memory mounted on the IC card of the present invention is not limited to a non-volatile memory, and may be a volatile memory (SRAM, DRAM, etc.). Further, an IC card on which both a nonvolatile memory and a volatile memory are mounted may be used.

  In the above description, the case where the invention made mainly by the present inventor is applied to the memory card which is the field of use that has been the background has been described. However, the present invention is not limited to this, and a passbook, a credit card, an ID card, etc. It can also be applied to the use of an IC card.

FIG. 3 is a circuit diagram illustrating an example of an IC card according to the present invention with respect to one connection terminal. It is an axial sectional view showing an example of a sectional structure of a varistor. It is an IV diagram which shows the characteristic of a varistor. It is explanatory drawing which illustrates the connection aspect of the varistor with respect to the connection terminal of a multimedia card. It is explanatory drawing which illustrated the structure of the multimedia card mainly in the circuit element mounting state. FIG. 6 is a longitudinal sectional view of the multimedia card in FIG. 5. It is explanatory drawing which shows the condition which biased the through hole with respect to the connection terminal of a multimedia card. It is a top view which shows partially the multimedia card which applied the stitch bonding to the connection of a non-volatile memory chip. It is a longitudinal cross-sectional view of a stitch bonding part. It is explanatory drawing which shows the wire bonding condition in the case of using a nail head bonding, and a case of using wedge bonding. It is explanatory drawing which illustrated planarly the structure of the multimedia card of the four-stack structure of a non-volatile memory chip. It is a longitudinal cross-sectional view which illustrates the cross-section of the multimedia card of FIG. 1 is a plan view illustrating a multimedia card to which a divided stack structure of memory chips and a structure in which connection terminals and a controller chip are arranged close to two adjacent sides of a card substrate are applied. FIG. It is a fragmentary longitudinal cross-sectional view of the multimedia card of FIG. FIG. 5 is a plan view illustrating another multimedia card to which a divided stack structure of memory chips and a structure in which connection terminals and a controller chip are arranged close to two adjacent sides of a card substrate are applied. It is a top view which illustrates the memory card which mounted the controller chip on the memory chip and stacked both. It is a top view which illustrates the memory card which stacked the memory chip and the controller chip using LOC. It is a top view which shows another example of the IC card to which the COB structure is applied. It is a longitudinal cross-sectional view of the IC card of FIG. It is explanatory drawing of the conductive pattern currently formed in the card substrate bottom face of the IC card of FIG. It is explanatory drawing which illustrates the IC card which has a cautionary note and caution display for electrostatic destruction prevention. It is a flowchart which illustrates the assembly method of the IC card which mounted the varistor. It is a top view of IC card which paid its attention to the viewpoint of data recovery. It is a flowchart which illustrates the procedure of the data recovery process with respect to IC card provided with the terminal for data evaluation. It is explanatory drawing which illustrates several types of structures of a metal cap. 1 is a block diagram illustrating a configuration of a flash memory chip. 1 is a cross-sectional view schematically showing a structure of a nonvolatile memory cell transistor for a flash memory chip.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Card substrate 2 Semiconductor integrated circuit chip 3 Connection terminal 3a-3g Connection terminal 4 External terminal 6 Input protection circuit 7, 8 Diode 9 Thyristor Q5, Q6, Q7 Clamp MOS transistor Vss Ground terminal Vcc Power supply terminal 11 Varistor 11a-11e Varistor 33 Controller chip 34a, 34b, 34c, 34d Non-volatile memory chip 38, 39 Bonding pattern 39A Bonding pattern 39B Wiring pattern 39C Bonding pattern 40 Through hole 41, 42 Bonding wire 43a, 43b, 43c, 43d Bonding wire 50, 51, 52a, 52b External terminal (binding pad)
55 Thermosetting resin 56 Metal cap 57 Bonding wire 60 Bonding wire 61 Conductive shield pattern 90 Precautionary note 91 Caution indication 92 Packaging material 92, 94, 96 Data recovery terminal

Claims (5)

A first surface of the card substrate having a first semiconductor integrated circuit chip and a second semiconductor integrated circuit chip;
A plurality of terminal electrodes on the second surface of the card substrate facing the first surface;
The plurality of terminal electrodes include a terminal used for supplying a first potential power supply, a terminal used for supplying a second potential power supply, a terminal used for data input / output, a terminal used for clock input, and a command input Including terminals used for
The first semiconductor integrated circuit chip is connected to each of the plurality of terminal electrodes, and performs access control to the second semiconductor integrated circuit chip according to a command input from a terminal used for the command input,
The second semiconductor integrated circuit chip is a plurality of nonvolatile memory chips;
Each of the plurality of nonvolatile memory chips has a pair of sides, and a plurality of external terminals are arranged along one side thereof,
The plurality of external terminals are connected to a terminal used for supplying the first potential power supply and a terminal used for supplying the second potential power supply, and to the first semiconductor integrated circuit chip. And a terminal used for the data input / output, a terminal used for the clock input, and a terminal used for the command input,
The first semiconductor integrated circuit chip has a plurality of chip selection signal output terminals for outputting a selection signal to a chip for access control among the plurality of nonvolatile memory chips,
The plurality of nonvolatile memory chips have a chip selection signal input terminal for inputting the selection signal arranged at an end of the arrangement of the plurality of external terminals, and the one side on which the external terminals are arranged And a non-volatile memory device in which the plurality of external terminals are respectively exposed for wire bonding by being stacked with their positions being shifted from each other in one direction of each of the sides intersecting the same.   2. The nonvolatile memory device according to claim 1, wherein the plurality of nonvolatile memory chips have the same terminals of the plurality of external terminals excluding the chip selection signal input terminals connected by a continuous bonding wire. The plurality of terminal electrodes are connected to the conductive pattern of the first surface through through holes,
The first semiconductor integrated circuit chip is connected to the plurality of terminal electrodes by being connected to the conductive pattern by a bonding wire,
3. The nonvolatile memory according to claim 1, wherein the through hole is disposed outside a mold region that covers the first surface of the card substrate together with the first semiconductor integrated circuit chip and the second semiconductor integrated circuit chip. apparatus.   The non-volatile memory device according to claim 3, wherein the through hole is formed at a position biased with respect to a sliding surface of the plurality of terminal electrodes. A first semiconductor integrated circuit chip mounted on the first surface of the card substrate;
A conductive pattern provided on the first surface of the card substrate and connected to the first semiconductor integrated circuit chip;
A second semiconductor integrated circuit chip mounted on the first surface of the card substrate and connected to the first semiconductor integrated circuit chip;
A plurality of terminal electrodes connected to the second surface facing the first surface via the conductive pattern and through holes;
The non-volatile memory device, wherein the first semiconductor integrated circuit chip is disposed so as to be sandwiched between the conductive pattern and the second semiconductor integrated circuit chip.

Images