Device and Method for Supplying Current to a Semiconductor Memory to Support a Boosted Voltage Within the Memory During Testing

1. Technical Field

This invention relates in general to semiconductor memories, such as Dynamic Random Access Memories (DRAMs), and, more specifically, to devices and methods for supplying current to semiconductor memories from external sources to support boosted voltages, such as wordline voltages and isolation gate voltages, within such memories while they are tested.

2. State of the Art

Dynamic Random Access Memories (DRAMs) typically include various circuitry that will only operate properly when supplied with a voltage (denoted "V.sub.CCP ") that is "boosted" above the supply voltage (denoted "V.sub.cc "). Such circuitry includes, for example, wordlines, which require a boosted voltage V.sub.CCP to store a full V.sub.cc level in a memory cell, and isolation gates, which require a boosted voltage V.sub.CCP to pass a full V.sub.cc level along a digit line.

In order to supply the boosted voltage V.sub.CCP, DRAMs typically include an internal charge pump that generates the boosted voltage V.sub.CCP on one or more capacitors. These capacitors are typically relatively large so they can supply sufficient current I.sub.CCP to meet any demands that may be made on the charge pump by the DRAM circuitry.

During DRAM compression-mode testing, the demand for current I.sub.CCP from the charge pump may be many times the demand for current I.sub.CCP during normal memory operations. This is because many more wordlines and isolation gates may be operated at the same time during compression-mode testing than during normal memory operations.

Consequently, DRAM designers typically find it necessary to provide a DRAM with a charge pump having capacitors of sufficient size to meet the increased demand for current I.sub.CCP experienced during compression-mode testing, despite the fact that much smaller capacitors would suffice for normal memory operations. As a result, DRAMs shipped to customers typically include charge pumps with capacitors many times the size required for even the most rigorous field applications. These over-sized capacitors unnecessarily occupy integrated circuit (IC) die "real estate," and thus can either limit the functional circuitry that can be provided in a DRAM, or necessitate a larger die than is desirable for a DRAM.

Therefore, there is a need in the art for a device and method for providing current I.sub.CCP to a DRAM or other semiconductor memory during testing without having to use a charge pump with over-sized capacitors.

A semiconductor device, such as a DRAM or other semiconductor memory, in accordance with this invention includes a conductor, such as a voltage bus, that distributes a boosted voltage (e.g., V.sub.CCP) within the semiconductor device. Internal boosting circuitry, such as a voltage regulator, a ring oscillator, and a charge pump, boosts a voltage level on the conductor upon sensing that the voltage level has fallen below a minimum level, such as a preset minimum. A terminal of the semiconductor device, such as a bond pad, receives current from a current source external to the device, and a switching circuit conducts current received through the terminal to the conductor in response to the internal boosting circuitry sensing that the voltage level on the conductor has fallen below the minimum level. As a result, the external current augments the efforts of the internal boosting circuitry to boost the voltage level on the conductor, thereby providing the necessary support for the boosted voltage during times of peak demand, such as during testing, without the need to provide oversized capacitors, for example, in the internal boosting circuitry. The switching circuit itself may be based on one or more pump circuits controlling one or more NMOS transistor switches that conduct the external current to the conductor when activated.

In other embodiments of this invention, the semiconductor device described above may be incorporated into an electronic device, or may be fabricated on the surface of a semiconductor substrate, such as a semiconductor wafer.

In a further embodiment of this invention, a boosted voltage in a semiconductor device is supported by boosting the boosted voltage using an externally generated current when the boosted voltage falls below a minimum level. The boosted voltage may be boosted by, for example, sensing that the boosted voltage has fallen below the minimum level. A charge pump in the semiconductor device can then be driven to boost the boosted voltage above the minimum level, and a switching circuit in the semiconductor device can be driven to conduct the externally generated current to augment the boosting of the boosted voltage by the charge pump.

FIG. 1 is a block diagram of a Dynamic Random Access Memory (DRAM) device in accordance with this invention;

FIG. 2 is a schematic and block diagram of a switching circuit of the DRAM device of FIG. 1;

FIG. 3 is a block diagram of an electronic system incorporating the DRAM of FIG. 1; and

FIG. 4 is a diagram of a semiconductor wafer having a surface on which the DRAM of FIG. 1 is fabricated.

As shown in FIG. 1, a Dynamic Random Access Memory (DRAM) device 10 in accordance with this invention includes a boosted voltage V.sub.CCP bus 12 for distributing the boosted voltage V.sub.CCP within the DRAM device 10. Although this invention will be described with reference to the DRAM device 10, it will be understood by those having skill in the field of the invention that the invention includes a wide variety of semiconductor devices within its scope, and is not limited to DRAM devices.

While the DRAM device 10 is operating, a conventional voltage regulator 14 senses the level of the boosted voltage V.sub.CCP on the bus 12 and outputs an oscillator activation signal REGDIS* when the level of the boosted voltage V.sub.CCP drops below a preset minimum. In response to the oscillator activation signal REGDIS*, a conventional ring oscillator 16 outputs a clock signal CLK that activates a conventional charge pump 18. This causes the charge pump 18 to "boost" the level of the boosted voltage V.sub.CCP above the preset minimum until it reaches a preset maximum, at which point the voltage regulator 14 deactivates the oscillator activation signal REGDIS*, causing the ring oscillator 16 to deactivate the clock signal CLK and thereby deactivating the charge pump 18.

During testing, in particular compression-mode testing, a test signal TEST deactivates an input buffer 20 through which signals entering a bond pad 22 normally pass during non-test mode operations of the DRAM device 10, and enables a switching circuit 24. The bond pad 22 may be any bond pad that is not needed during testing, such as an unused address signal pad or a no-connection (NC) pad. Once enabled, the switching circuit 24 operates in response to the clock signal CLK by passing current I.sub.CCP from an external current source 26 attached to the bond pad 22 to the bus 12, thereby boosting the boosted voltage V.sub.CCP above its preset minimum.

Thus, the switching circuit 24 supports the increased demand on the boosted voltage V.sub.CCP during testing by providing the current I.sub.CCP from an external source. This allows the capacitors (not shown) of the charge pump 18 to be sized for the lesser demand on the boosted voltage V.sub.CCP experienced during normal memory operations rather than the increased demand experienced during test operations. As a result, less die "real estate" is used for the capacitors of the charge pump 18, so the DRAM device 10 can be manufactured on a smaller integrated circuit (IC) die, or more functional circuitry can be provided in the DRAM device 10.

As shown in detail in FIG. 2, during testing, an active (i.e., high) test signal TEST enables the switching circuit 24 by causing an inverter 27 to output a low. Assuming, for the moment, a steady-state condition in which the clock signal CLK has not been activated by the ring oscillator 16 (FIG. 1), the low output by the inverter 27 causes a NOR gate 28 to output a high which, in turn, causes an inverter 30 to output a low and an inverter 32 to output a high. The high from the inverter 32 causes a NOR gate 34 to output a low which, when combined with the low from the inverter 30, causes a NOR gate 36 to output a high. The low from the NOR gate 34 causes an inverter 38 to output a high which, in turn, causes another inverter 40 to output a low, and the high from the NOR gate 36 causes an inverter 42 to output a low which, in turn, causes yet another inverter 44 to output a high. The high from the inverter 44 causes a NOR gate 46 and a NOR gate 48 to both output a low. The low from the NOR gate 48 and the low from the inverter 40 cause a NOR gate 50 to output a high which, in turn, causes a NOR gate 52 to output a low.

The lows output by the NOR gates 46 and 52 ground node "A" of capacitors 54 and 56. Meanwhile, the high test signal TEST activates NMOS transistors 58 and 60, thereby storing the supply voltage V.sub.CC, less the threshold voltage V.sub.T of the transistors 58 and 60, on node "B" of the capacitors 54 and 56. Helper NMOS transistors 62 and 64 help to pull node B of the capacitors 54 and 56 up to V.sub.CC -V.sub.T.

Once the clock signal CLK is activated by the ring oscillator 16 (FIG. 1), a rising edge of the clock signal CLK flips the output of the NOR gate 28 to a low, causing the inverter 30 to output a high and the inverter 32 to output a low. The high from the inverter 30 causes the NOR gate 36 to output a low which, when combined with the low from the inverter 32, causes the NOR gate 34 to output a high. The low from the NOR gate 36 causes the inverter 42 to output a high and the inverter 44 to output a low, and the high from the NOR gate 34 causes the inverter 38 to output a low and the inverter 40 to output a high.

The low from the inverter 44 causes the NOR gate 46 to begin to output a high pulse. This high pulse lasts until the high output of the inverter 40 causes the NOR gate 50 to output a low, causing the NOR gate 48 to output a high and thereby driving the output of the NOR gate 46 low again. The output of the NOR gate 52 remains low on the rising edge of the clock signal CLK.

The high pulse from the NOR gate 46 raises node A of the capacitor 54 up to the supply voltage V.sub.CC during the pulse, which "boosts" the voltage on node B of the capacitor 54 up to 2V.sub.CC -V.sub.T (because V.sub.CC -V.sub.T is already stored across the capacitor 54). This "boosted" voltage causes the NMOS transistors 58 and 62 to turn off, and turns on an NMOS transistor 66, allowing the transistor 66 to pass the external current I.sub.CCP through to the V.sub.CCP bus 12.

A conventional clamp 68 keeps the voltage on node B of the capacitor 54 from exceeding V.sub.CC +4V.sub.T, but it can, of course, be constructed to limit the node to any desired maximum voltage. Also, the boosted voltage on node B of the capacitor 54 causes the transistor 64 to pass a full supply voltage V.sub.CC level through to node B of the capacitor 56 for storage thereon.

With the clock signal CLK still activated, a falling edge of the clock signal CLK flips the output of the NOR gate 28 to a high, causing the inverter 30 to output a low and the inverter 32 to output a high. The high from the inverter 32 causes the NOR gate 34 to output a low which, when combined with the low from the inverter 30, causes the NOR gate 36 to output a high. The low from the NOR gate 34 causes the inverter 38 to output a high and the inverter 40 to output a low, and the high from the NOR gate 36 causes the inverter 42 to output a low and the inverter 44 to output a high.

The low from the inverter 40 causes the NOR gate 52 to begin to output a high pulse. This high pulse lasts until the high output of the inverter 44 causes the NOR gate 48 to output a low, causing the NOR gate 50 to output a high and thereby driving the output of the NOR gate 52 low again. The output of the NOR gate 46 remains low on the falling edge of the clock signal CLK.

The high pulse from the NOR gate 52 raises node A of the capacitor 56 up to the supply voltage V.sub.CC during the pulse, which "boosts" the voltage on node B of the capacitor 56 up to 2V.sub.CC (because V.sub.CC is already stored across the capacitor 56). This "boosted" voltage causes the NMOS transistors 60 and 64 to turn off, and turns on an NMOS transistor 70, allowing the transistor 70 to pass the external current I.sub.CCP through to the V.sub.CCP bus 12.

A conventional clamp 72 keeps the voltage on node B of the capacitor 56 from exceeding V.sub.CC +4V.sub.T, but it can, of course, be constructed to limit the node to any desired maximum voltage. Also, the boosted voltage on node B of the capacitor 56 causes the transistor 62 to pass a full supply voltage V.sub.CC level through to node B of the capacitor 54 for storage thereon. As a result, subsequent operation of the switching circuit 24 periodically boosts node B of the capacitor 54 up to 2V.sub.CC, rather than 2V.sub.CC -V.sub.T as described above, because a full V.sub.CC level is now stored across the capacitor 54.

During normal operations of the DRAM device 10 (FIG. 1), the test signal TEST is low, which causes the inverter 27 to output a high, thereby turning on NMOS transistors 74 and 76, and pulling node B of the capacitors 54 and 56 to ground through NMOS transistors 74, 76, 78, and 80. As a result, the NMOS transistors 66 and 70 are off, and the V.sub.CCP bus 12 is isolated from the bond pad 22 (FIG. 1).

It should be understood that, as a group, the inverters 27, 30, 32, 38, 40, 42, and 44, and the NOR gates 28, 34, 36, 46, 48, 50, and 52, may be referred to generally as "pump control circuitry." It should also be understood that, as a group, the capacitor 54 and the NMOS transistor 58 may be referred to as a "pump circuit," and that, as a group, the capacitor 56 and the NMOS transistor 60 may also be referred to as a "pump circuit."Further, it should be understood that the NMOS transistors 66 and 70 may be referred to as "switches."

As shown in FIG. 3, an electronic system 82 includes an input device 84, an output device 86, a processor device 88, and a memory device 90 incorporating the DRAM device 10 of FIG. 1. Of course, any one of the input, output, and processor devices 84, 86, and 88 can also incorporate the DRAM device 10.

As shown in FIG. 4, the DRAM device 10 of FIG. 1 is fabricated on the surface of a semiconductor wafer 92. However, the DRAM device 10 may also be manufactured on a wide variety of semiconductor substrates other than a semiconductor wafer including, for example, a Silicon on Sapphire (SOS) substrate, a Silicon on Glass (SOG) substrate, and a Silicon on Insulator (SOI) substrate.

Although this invention has been described with reference to particular embodiments, the invention is not limited to these described embodiments. Rather, the invention is limited only by the appended claims, which include within their scope all equivalent devices and methods that operate according to the principles of the invention as described.