FIELD OF THE INVENTION
The present invention relates generally to nonvolatile memory (NVM). More particularly, the present invention relates to methods of and apparatuses for preventing overtunneling in pFET-based NVM memory cells.
BACKGROUND OF THE INVENTION
The demand for embedded nonvolatile memory (NVM) in integrated circuits has grown steadily over the past decade. Desirable characteristics of embedded NVM include low cost, low power, high speed, and high reliability (data retention and program/erase cycling endurance). NVM may be embedded in various integrated circuit (IC) technologies such as, for example, the widely used Complementary Metal Oxide Semiconductor (CMOS) technology. Some embedded CMOS applications include, for example, storing: (1) chip serial numbers, (2) configuration information in ASICs (Application Specific Integrated Circuits), (3) product data in radio frequency identification integrated circuits, (4) code or data in embedded microcontrollers, and (5) analog trim information.
A major barrier for using embedded NVM is cost. An IC fabricator typically requires additional processing steps to manufacture NVM storage transistors. For example, IC fabricators sometimes use two layers of polysilicon for the gate of an NVM storage transistor, rather than one layer as in standard CMOS technology. The additional fabrication step increases the total cost of the IC. Typical embedded EEPROM (electrically erasable programmable read only memory) or Flash NVM uses nFET (n-channel field effect transistor) storage transistors. To ensure charge retention in nFETs, the IC fabricator typically uses a thicker gate oxide than is found in logic transistors, again increasing cost.
To reduce the costs and added complexities of embedding NVM in integrated circuits, efforts have been made to design an NVM that can be integrated with CMOS process technology without introducing additional processing steps. These integration efforts have also involved endeavoring to use pFET-based NVM, rather than the more traditional nFETs-based NVM. The reason for this is that pFET-based memory cells exhibit various performance advantages compared to nFET-based memory cells. pFETs have the following advantages over their nFET-based NVM counterparts: 1) increased program/erase cycle endurance (due to reduced oxide wearout); 2) availability in logic CMOS processes (due to reduced memory leakage arising from more favorable oxide physics); 3) ability to easily store analog as well as digital values (due to precise memory writes); and 4) smaller on-chip charge pumps (due to decreased charge-pump current requirements).
Although using pFETs as NVM transistors affords significant benefits compared to using nFETs as NVM transistors, the possibility of “overtunneling” such cells poses a significant problem. The referred to “overtunneling” problem manifests as follows. pFET-based memory cells use electron tunneling to raise the floating-gate voltage, and impact-ionized hot-electron injection (IHEI) to lower the floating-gate voltage. One characteristic of the IHEI programming method is that the MOSFET channel must be conducting current to allow electrons to inject onto the floating gate. If during a prior tunneling cycle the floating-gate voltage was raised so high that the pFET was turned off, there will be no channel current when a write to the cell is attempted. Effectively, by overtunneling the memory cell, the memory cell becomes “stuck” in an off state, and in the absence of channel current no electron injection can be performed during a programming (i.e. injection) cycle to lower the floating-gate voltage.
The overtunneling problem observed in pFET-based NVM cells detracts their use as reliable memory devices, despite the superior performance advantages they have over nFET-based NVM cells. Accordingly, there is a need for methods and apparatuses for preventing overtunneling in pFET-based NVM cells.
BRIEF DESCRIPTION OF THE INVENTION
Methods and apparatuses to prevent overtunneling in pFET-based nonvolatile floating gate memory (NVM) cells. During a tunneling process, in which charge carriers arc removed from a floating gate of a pFET-based NVM cell, a channel current of a memory cell transistor is monitored and compared to a predetermined minimum channel current required to maintain a conducting channel in an injection transistor of the memory cell. When the monitored channel current drops below the predetermined minimum channel current, charge carriers are injected onto the floating gate by impact-ionized hot-electron injection (IHEI) so that overtunneling is avoided. Other aspects of the inventions are described and claimed below, and a further understanding of the nature and advantages of the inventions may be realized by reference to the remaining portions of the specification and the attached drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows a memory circuit for preventing overtunneling in a pFET-based memory cell, according to an embodiment of the present invention.
FIG. 2 shows another memory circuit for preventing overtunneling in a pFET-based memory cell, according to an embodiment of the present invention.
FIG. 3A is shows yet another memory circuit for preventing overtunneling in a pFET-based memory cell, according to an embodiment of the present invention.
FIG. 3B shows a timing diagram illustrating the operation of the memory circuit in FIG. 3A, according to an embodiment of the present invention.
FIG. 4 shows the memory circuit of FIG. 3A, modified to take advantage of an available negative voltage source, according to an embodiment of the present invention.
FIG. 5 shows the memory circuit of FIG. 1, modified so that it includes a select/bias transistor, according to an embodiment of the present invention.
FIG. 6 shows the memory circuit of FIG. 1, modified so that it includes a capacitor coupled between the floating gate of the memory cell and a voltage source Vdd, according to an embodiment of the present invention.
FIG. 7 shows how the capacitor in the memory circuit shown in FIG. 6 may be formed from a pFET configured as a MOS capacitor (MOSCAP), according to an embodiment of the present invention.
FIG. 8 shows how the capacitor in the memory circuit shown in FIG. 6 may be formed from half of a pFET configured as a MOSCAP, according to an embodiment of the present invention.
FIG. 9 shows the memory circuit of FIG. 6, modified so that the added capacitor is coupled to a control source Vcontrol, rather than Vdd, according to an embodiment of the present invention.
FIG. 10 shows a memory circuit similar to the memory circuit of FIG. 9, where the added capacitor is formed from an nFET, according to an embodiment of the present invention.
FIG. 11 shows a memory circuit that includes both a select/bias transistor similar to that shown in the memory circuit in FIG. 5 and a capacitor similar to that shown in the memory circuit in FIG. 6, according to an embodiment of the present invention.
FIG. 12A shows a memory circuit similar to the memory circuit in FIG. 3A, including a high-voltage switch, according to an embodiment of the present invention.
FIG. 12B shows a timing diagram illustrating the operation of the memory circuit in FIG. 12A, according to an embodiment of the present invention.
FIG. 13 shows a memory circuit similar to the memory circuit in FIG. 12A, including a Vdd switch and a Vwell switch, according to an embodiment of the present invention.
FIG. 14 shows a memory circuit, including capacitor similar to the added capacitor in the memory circuit in FIG. 6, according to an embodiment of the present invention.
FIG. 15A shows a memory circuit employing a source-follower-connected pFET and an additional pulse driver, according to an embodiment of the present invention.
FIG. 15B shows a timing diagram of the operation of the memory circuit in FIG. 15A, according to an embodiment of the present invention.
FIG. 16 shows the memory circuit of FIG. 3A, modified so that it includes a tristate logic gate, according to an embodiment of the present invention.
FIG. 17 shows the memory circuit of FIG. 3A, modified so that it incorporates an nFET in series with a diode, according to an embodiment of the present invention.
FIG. 18 shows the memory circuit of FIG. 3A, modified so that it incorporates a pFET between the current sense amplifier and the drain of the injection transistor, according to an embodiment of the present invention.
FIG. 19 shows an overtunneling prevention memory circuit including a 2×2 array of memory cells and associated overtunneling prevention control circuits, according to an embodiment of the present invention.
FIG. 20 shows an overtunneling prevention memory circuit including a 2×2 array of memory cells and associated overtunneling prevention control circuits, according to an embodiment of the present invention.
FIG. 21 shows an overtunneling prevention memory circuit including a 2×2 array of memory cells with a single shared overtunneling prevention control circuit, according to an embodiment of the present invention.
DETAILED DESCRIPTION
Embodiments of the present invention are described herein in the context of methods and apparatuses for preventing overtunneling in pFET -based nonvolatile memory cells. Those of ordinary skill in the art will realize that the following detailed description of the present invention is illustrative only and is not intended to be in any way limiting. Other embodiments of the present invention will readily suggest themselves to such skilled persons having the benefit of this disclosure.
Reference will now be made in detail to implementations of the present invention as illustrated in the accompanying drawings. The same reference indicators will be used throughout the drawings and the following detailed description to refer to the same or similar parts.
Referring first to FIG. 1, there is shown a memory circuit 1 for preventing overtunneling in a pFET-based memory cell, according to an embodiment of the present invention. Memory circuit 1 comprises a memory cell 10 and an overtunneling prevention control circuit 12. Memory cell 10 comprises an injection transistor 14 and a tunneling capacitor 16. Injection transistor 14 has a floating gate 15, a source coupled to a voltage source Vdd, a body coupled to a well voltage source Vwell, and a drain coupled to overtunneling prevention control circuit 12. As shown, tunneling capacitor 16 is formed from a pFET, with the source, drain and body of the pFET shorted together and coupled to a tunneling voltage source Vtun. However, a tunneling capacitor constructed from other structures such as, for example, an nFET can also be used. The gate of tunneling capacitor 16 is coupled to the floating gate 15 of injection transistor 14.
FIG. 2 shows a memory circuit 2 for preventing overtunneling in a pFET-based memory cell, according to an embodiment of the present invention. Similar to the embodiment in FIG. 1, the memory circuit 2 comprises a memory cell 10 having an injection transistor 14 and a tunneling capacitor 16. Memory circuit 2 also includes an nFET (i.e. an n-channel MOSFET) overtunneling prevention transistor 24 having a drain coupled to the drain of injection transistor 14, a source coupled to a negative supply voltage Vss, and a gate coupled to a reference voltage Vref.
Memory circuit 2 in FIG. 2 operates as follows. Assume the voltage on the floating gate 15 is low, and it is desired to tunnel it up. To tunnel up floating gate 15 a tunnel voltage Vtun of about (Vfg+10V), where Vfg is the floating gate voltage and 10V is typical for a 0.35 μm CMOS process with 75 Å oxides, is applied to the tunneling capacitor 16. Vtun causes electrons to tunnel from floating gate 15, through the tunneling capacitor's dielectric (i.e., the gate oxide, if tunneling capacitor is formed from a pFET or an nFET), to Vtun, thereby raising Vfg. To prevent overtunneling, a reference voltage Vref is applied to the gate of overtunneling prevention transistor 24. Overtunneling prevention transistor 24 operates by sinking a small current Imin (e.g. ˜250 nA) from injection transistor 14. As long as injection transistor 14 is able to source more current than overtunneling prevention transistor 24 sinks, Vdrain remains high, and injection transistor 14 will not inject electrons onto floating gate 15. When, however, Vfg rises so high that injection transistor 14 can no longer source Imin, Vdrain will fall, causing injection transistor to begin injecting electrons onto floating gate 15. Eventually, Vdrain will stabilize at a voltage where the IHEI gate current is equal and opposite to the tunneling gate current. Hence, overtunneling prevention transistor 24 prevents injection transistor 14 from turning off by injecting electrons back onto floating gate 15, thereby forcing the channel current of injection transistor 14 to maintain a value equal to Imin.
In 0.35 μm and smaller CMOS logic processes, a voltage of not more than about 12V can be applied to the body of tunneling capacitor 16, without risking body-to-substrate breakdown. Because a voltage of ˜10V is needed across the gate oxide of tunneling capacitors 16 to cause appreciable electron tunneling, Vfg must be roughly (12V−10V)=2V. To obtain channel currents in the range of 10 nA to 10 μA, Vdd should then be ˜3.3V. To obtain reasonable IHEI in injection transistor 14, Vdrain should be ˜−2V, meaning Vss should be ˜−2.5V. Unfortunately, most modern n-well CMOS processes do not offer nFETs that operate with a Vss of more than a few hundred millivolts below ground, because the nFET's substrate-to-source and substrate-to-drain p-n junctions become forward biased. If such limitations are encountered, other embodiments of the present invention may be used. One alternative embodiment is to use a deep n-well or a dual-well process and fabricate an overtunneling prevention transistor, like transistor 24 shown in FIG. 2, in a p-well that can be biased ˜2.5V below ground. A second alternative embodiment is to provide an overtunneling prevention control circuit to emulate the functions of the overtunneling prevention transistor 24 without having to resort to additional processing steps necessary to create a p-well operating below ground. FIG. 3A shows an example of the latter alternative, in accordance with an embodiment of the present invention.
Referring to FIG. 3A, there is shown a memory circuit 3 for preventing overtunneling in a pFET-based memory cell, according to an embodiment of the present invention. Memory circuit 3 comprises a memory cell 10 having an injection transistor 14 and a tunneling capacitor 16, which may be formed from a pFET transistor as shown. The drain of injection transistor 14 is coupled to an overtunneling prevention control circuit 12, which comprises a current sense amplifier 26, a controller 28 coupled to current sense amplifier 26, a pulse driver 30 coupled to controller 28, a capacitor 32 coupled between pulse driver 30 and the drain of injection transistor 14 and a diode 34 coupled between the drain of injection transistor 14 and ground.
Memory circuit 3 in FIG. 3A operates as follows. During tunneling, current sense amplifier 26 monitors the drain current Idrain of injection transistor 14. Tunneling causes Idrain to gradually decrease, as shown in the timing diagram provided in FIG. 3B. Current sense amplifier 26 is configured to trigger when Idrain decreases to a value of Imin. When current sense amplifier 26 triggers, controller 28 instructs pulse driver 30 to pull Vp from Vdd (nominally 3.3V) down to ground. This is indicated in FIG. 3B as occurring at time t1. Capacitor 32 then pulls the drain voltage Vdrain of injection transistor 14 from 0.7V (the “on” voltage of diode 34) to ˜2.6V, causing electron injection to commence in injection transistor 14, and thereby causing Idrain to increase. After a short period of time, at time t2 controller 28 instructs pulse driver 30 to pull Vp from ground back up to Vdd, and waits for current sense amplifier 26 to trigger again. In this fashion overtunneling prevention control circuit 12 pulses Vdrain as needed to ensure that injection transistor 104 is not overtunneled into an “off” state. Note that, although a “current sense” amplifier is employed to determine when Vdrain must be pulsed low to avoid overtunneling, other sensing or monitoring devices and circuits may be used. For example, the cell current may be supplied to any one of many possible current-to-voltage circuit elements (e.g. resistor, diode, current source, etc) so that a voltage is measured and/or monitored, rather than transistor 14's drain current itself.
FIG. 4 shows the memory circuit of FIG. 3A, modified to take advantage of an available negative voltage source Vminus, which would nominally be about ˜3.3V in a 0.35 μm CMOS process, according to an embodiment of the present invention. Memory circuit 4 in FIG. 4 comprises essentially the same elements as in memory circuit 3 of FIG. 3A, but also includes a source-follower-connected pFET 36 configured to operate as a negative-voltage switch. In one embodiment, Vminus may be provided by an off-chip voltage source. In an alternative embodiment, Vminus may be generated on the same semiconductor chip shared by memory cell 10 by using, for example, a negative-voltage charge pump. Source-follower-connected transistor 36 forms a negative-voltage switch as follows. When pulse driver 30 pulls Vp from Vdd (e.g. 3.3V) to ground and the gate of source-follower-connected transistor 36 is pulled to about ˜2.6V, the source of source-follower-connected transistor 36, and with it Vdrain, gets pulled down to about −2V.
FIG. 5 shows the memory circuit of FIG. 1, modified so that it includes a select/bias transistor 38 comprising a pFET, according to an embodiment of the present invention. Select/bias transistor 38 has a gate selectively coupled to a Vbias/select voltage source, a source coupled to Vdd, a drain coupled to the source of injection transistor 14, and a well coupled to voltage source Vwell. Select/bias transistor 38 in memory circuit 5 may be used to select memory cell 10 for injection (e.g. from among an array of memory cells) by controlling the current in injection transistor 14, and/or to limit the current in injection transistor 14 during other operations such as reading, for example.
FIG. 6 shows the memory circuit of FIG. 1, modified so that it includes a capacitor 40 coupled between floating gate 15 of memory cell 10 and voltage source Vdd, according to an embodiment of the present invention. Capacitor 40 of memory circuit 6 may be used to ensure that, when Vdd is pulled low, the floating gate follows.
FIG. 7 shows how capacitor 40 in memory circuit 6 in FIG. 6 may be formed from a pFET 42 configured as a MOS capacitor (MOSCAP), according to an embodiment of the present invention. pFET 42 of memory circuit 7 has a gate coupled to floating gate 15, a source, a drain shorted to the source and coupled to voltage source Vdd, and a body coupled to the body of injection transistor 14 and to a well voltage source Vwell.
FIG. 8 shows how capacitor 40 in memory circuit 6 in FIG. 6 may be formed from half of a pFET 44 configured as a MOSCAP, according to an embodiment of the present invention. Half pFET 44 may be constructed from, for example, a pFET with either the drain or source terminal left floating, or a pFET having either a drain or source terminal but not both. The latter embodiment saves layout area of the integrated circuit on which memory circuit 8 is formed.
FIG. 9 shows the memory circuit 6 of FIG. 6, modified so that capacitor 40 is coupled to a control source Vcontrol, rather than Vdd, according to an embodiment of the present invention. This alternative connection allows the independent control of Vdd and the floating-gate voltage Vfg. Capacitor 40 can be constructed from a pFET or from half of a pFET, as described in FIGS. 7 and 8. Additionally, capacitor 40 may be formed in the same n-well as injection transistor 104, or in a separate n-well.
FIG. 10 shows a memory circuit 100 similar to the memory circuit 9 of FIG. 9, where capacitor 40 in FIG. 9 is formed from an nFET 42, according to an embodiment of the present invention. As shown in FIG. 10, nFET 42 is configured as a MOSCAP, having a gate coupled to floating gate 15 and shorted source, drain and body terminals coupled to the control source Vcontrol. In various alternative embodiments, nFET 42 may be formed with no drain or source terminal, with either a drain terminal or a source terminal but not both, or with both a drain terminal and a source terminal.
FIG. 11 shows a memory circuit 11 that includes the select/bias transistor 38 of memory circuit 5 in FIG. 5 and the capacitor 40 of memory circuit 6 in FIG. 6, according to an embodiment of the present invention. Capacitor 40 may comprise any of the forms described above and may connect to either Vdd or a separate control input.
FIG. 12A shows a memory circuit 112 similar to the memory circuit in FIG. 3A, including a high-voltage switch 44, according to an embodiment of the present invention. High-voltage switch 44, as controlled by controller 28, is opened to prevent tunneling during times when memory cell 10 is being sensed or when charge carriers are being injected onto floating gate 15. High-voltage switch 44 prevents tunneling during sense and inject operations to avoid capacitive coupling between tunneling capacitor 16 and floating gate 15. Capacitive coupling undesirably causes floating gate 15 to be pulled high during tunneling, artificially decreasing the drain current Idrain of injection transistor 14 as shown in the timing diagram provided in FIG. 12B. To read the drain current Idrain accurately, tunneling must first be terminated. Memory circuit 112, including overtunneling prevention control circuit 12 and high-voltage switch 44 performs the following sequence of operations as illustrated in the timing diagram shown in FIG. 12B. During phase 1 a positive-going tunneling pulse is applied to tunneling capacitor 16. Next, during phase 2 controller 28 causes high-voltage switch 44 to open to turn off tunneling. During phase 2, Idrain is measured by sense amplifier 26. Finally, during phase 3, if Idrain is smaller than a predetermined minimum drain current Imin, controller 28 causes pulse driver 30 to pulse, thereby lowering the drain voltage Vdrain of inject transistor 14. The lowering of Vdrain causes inject transistor 14 to begin injecting charge carriers onto floating gate 15.
FIG. 13 shows a memory circuit 113 similar to the memory circuit in FIG. 12A, including a Vdd switch 46 and a Vwell switch 48, according to an embodiment of the present invention. This alternative embodiment may be used to decouple Vdd and Vwell from memory cell 10 or to set them to a low voltage such as ground during tunneling, to capacitively couple floating gate 15 down by a volt or more and thereby reduce the required tunneling voltage Vtun by this same volt or more.
FIG. 14 shows a memory circuit 114, including the capacitor 40 in memory circuit 6 of FIG. 6, according to an embodiment of the present invention. Because capacitor 40 provides significant capacitive coupling between Vdd and floating gate 15, there is less need to switch Vwell down to ground during tunneling. (Capacitor 40 effectively replaces the parasitic well-to-floating-gate capacitance of injection transistor 14.) Consequently, the Vwell switch in the embodiment shown in FIG. 13 is not required. Those skilled in the art will readily understand that many other combinations of switches and memory cells are possible.
FIG. 15A shows a memory circuit 115 employing a source-follower-connected pFET 50 and an additional pulse driver 52, according to an embodiment of the present invention. Pulse drivers 52 and 54 are controlled by a controller 56, which is coupled to a current sense amplifier 58. A first capacitor 60 is coupled between pulse driver 54 and the drain of injection transistor 14. A second capacitor 62 is coupled between pulse driver 52 and the gate of pFET 50. The source of pFET 50 is coupled to the drain of injection transistor 14 and the source and body of pFET 50 are shorted together and coupled to ground. A diode 64 is coupled between the gate of pFET 50 and ground. The use of source-follower-connected transistor 50 allows a lower drop in the drain voltage Vdrain to be realized, compared to some of the previous embodiments described above. The reason for this is that the saturated drain voltage of source-follower-connected transistor 50 can be 100 millivolts or less, whereas the “on-voltage” of diode 34 in the previous embodiments is closer to 700 millivolts. With this difference, Vdrain transitions from 0.1V to ˜3.2V during an injection cycle rather than from 0.7V to ˜2.6V. Because IHEI increases exponentially with drain-to-gate voltage, memory circuit 115 has more efficient injection compared with, for example, memory circuit 3 in FIG. 3A.
FIG. 15B shows a timing diagram of the operation of memory circuit 115 in FIG. 15A. During phase 1 Vpp transitions from 3.3V to ground, pulling the gate of source-follower-connected transistor 50 to ˜2.6V below ground, thereby turning source-follower-connected transistor 50 on and pulling Vdrain close to ground (limited only by the saturation voltage of source-follower-connected transistor 50). During phase 2 Vpp transition from ground to 3.3V, returning the gate of transistor 50 back to ˜0.7V and discharging any accumulated charge on capacitor 62 through diode 64; also, Vp transitions from 3.3V to ground, pulling Vdrain from ˜0.1V to ˜−3.2V and causing injection transistor 14 to inject. During phase 3 Vp transitions from ground to 3.3V, thereby turning off injection.
FIG. 16 shows the memory circuit of FIG. 3A, modified so that it includes a tristate logic gate 66, according to an embodiment of the present invention. Tristate logic gate 66 of memory circuit 116 is driven by pulse driver logic 68, which is controlled by a controller 70. Similar to the previous embodiments, drain current of injection transistor 14 is monitored by a current sense amplifier 72. A capacitor 74 is coupled between tristate logic gate 66 and the drain of injection transistor 14 and a diode 76 is coupled between the drain of injection transistor 14 and ground. As shown, tristate logic gate 66 is an inverter. However, alternative logic gates with tristate outputs may also be used (e.g., such as NANDs or NORs). A benefit of a tristated output is that it reduces the capacitive load presented by capacitor 74 on the Vdrain line during reading and/or sensing, thereby reducing the read/sense times compared to embodiments using non-tristated pulse drivers.
FIG. 17 shows the memory circuit of FIG. 3A, modified so that it incorporates an nFET 76 in series with diode 34, according to an embodiment of the present invention. nFET 76 of memory circuit 117 has a gate, which is controlled by controller 28, a drain coupled to the cathode of diode 34 and a source coupled to ground. Controller 28 is configured to turn off nFET 76 during reading/sensing. A benefit of using nFET 76 is that, when nFET 76 is turned off, Vdrain can have values greater than 700 mV above ground. Note that diode 34 may comprise a source-follower or diode-connected pFET, rather than a p-n junction as shown, and that nFET 76 may be used in all the other circuit implementations in this disclosure (such as, for example, being used in series with transistor 50 in FIG. 15A).
FIG. 18 shows the memory circuit of FIG. 3A, modified so that it incorporates a pFET 78 between current sense amplifier 26 and the drain of injection transistor 14, according to an embodiment of the present invention. pFET 78 of memory circuit 118 has a gate and body, both coupled to ground, a drain coupled to current sense amplifier 26 and a source coupled to the drain of injection transistor 14. The reason for adding pFET 78 is that, if current sense amplifier 26 has an n-type (i.e. nMOS or NPN) input stage, this stage's substrate-to-drain p-n junction cannot assume values more than about 700 mV below ground without turning on (assuming the chip substrate is grounded). Diode-connected pFET 78, or an alternative structure such as a p-n diode, allows Vdrain to pulse more than 700 mV below ground during injection.
Referring now to FIG. 19, there is shown an overtunneling prevention memory circuit 119 including a 2×2 array of memory cells and associated overtunneling prevention control circuits, according to an embodiment of the present invention. In this embodiment the memory cells 10 in a first row of the array have injection transistors 14 with interconnected sources and interconnected bodies. The interconnected sources are coupled to a voltage source Vdd0 and the bodies are coupled to a well voltage source Vwel10. Tunneling capacitors 16 in the first row of the array are coupled to a tunneling voltage source Vtun0. Memory cells 10 in a second row of the array have injection transistors 14 and tunneling capacitors 16, which are configured similar to the injection transistors 14 and tunneling capacitors 16 in the cells in the first row, except that the various transistor and capacitor terminals are coupled to voltage sources Vdd1, Vwell1 and Vtun1, as shown in the figure. In this embodiment, each column of the array has an associated overtunneling prevention control circuit 12, which may comprise any of the previously described overtunneling prevention control circuits. An overtunneling prevention control circuit 12 of an associated column of the array is coupled to the drains of the injection transistors 14 of the memory cells 10 within the associated column. Additionally, the memory cells 10 may comprise any of the various memory cell embodiments shown in FIGS. 5-11 above (or others extrapolated from them) as appropriate. Those skilled in the art will also readily understand that, although only a 2×2 memory array is shown, the array size could be extended to any m-row by n-column, where m and n are integers both greater than or equal to two.
FIG. 20 shows an overtunneling prevention memory circuit 120 including a 2×2 array of memory cells and associated overtunneling prevention control circuits, according to an embodiment of the present invention. This alternative embodiment modifies the embodiment shown in FIG. 19, by adding a master controller 80 that controls the overtunneling prevention control circuits 12 of the two columns so that one row of the array is tunneled at a time. Master controller 80 may be also configured to control high-voltage switch 44, Vdd switch 46 and Vwell switch 48 in an associated row of the array, in the manner and for the purposes described above. Master controller 80 may also provide a “tunneling done” output for the following reason. As master controller 80 monitors the various controllers 28 of the corresponding overtunneling prevention control circuits 12, it knows which cells have tunneled to the point where they need injection pulses to prevent overtunneling. When every cell in a given row has undergone at least one injection pulse, then the cells have been tunneled to the desired value, so master controller 80 halts further tunneling pulses and issues the “tunneling done” signal. As with the embodiment in FIG. 19, overtunneling prevention control circuits 12 may comprise any of the previously described overtunneling prevention control circuits. Additionally, memory cells 10 may comprise any of the various memory cell embodiments shown in FIGS. 5-11 above (or others extrapolated from them) as appropriate.
Overtunneling prevention memory circuit 120 in FIG. 20 may further include a high-voltage charge pump 82, under the control of master controller 80. Use of charge pump 82 is as follows. Every nonvolatile memory system is faced with the dilemma of how to regulate the high-voltage charge pump's output voltage. If the voltage is too low, then the tunneling rate will be too slow. On the other hand, if the voltage is too high, then the memory cells may tunnel so fast that the overtunneling prevention control circuits, as controlled by master controller 80, cannot correct any overtunneling problems. In an embodiment including charge pump 82, charge pump 82 is unregulated (i.e. not set to a fixed voltage) and is designed to gradually ramp up its output voltage and use the “tunneling done” signal to turn off pumping, as soon as cells are fully tunneled. By this means the tunneling voltage is never too low because it continually ramps upward; the tunneling voltage is also never too high, because the master controller 80 turns off pumping when all cells are done (prior to the tunneling voltage becoming too high).
FIG. 21 shows an overtunneling prevention memory circuit 121 including a 2×2 array of memory cells with a single shared overtunneling prevention control circuit, according to an embodiment of the present invention. In contrast to the memory array embodiments shown in FIGS. 19 and 20, a 2:1 multiplexer 84 is employed to route selected column Vdrain lines into a single overtunneling prevention control circuit 12. This implementation saves circuit area compared to the embodiments shown in FIGS. 19 and 20. Memory circuit 121 functions as follows. A tunneling pulse is applied to one row of the array. Then, sequentially, using multiplexer 84 to select individual columns, each cell's drain current Idrain is measured. If necessary, one or more injection pulses are applied to correct overtunneled cells. As with the embodiments in FIGS. 19 and 20, overtunneling prevention control circuits 12 may comprise any of the previously described overtunneling prevention control circuits. Additionally, memory cells 10 may comprise any of the various memory cell embodiments shown in FIGS. 5-11 above (or others extrapolated from them) as appropriate. Those skilled in the art will also readily understand that, although a 2:1 multiplexer is shown, the multiplexer could be extended to any n:p multiplexer, where n is an integer that is greater than or equal to one and represents the number of columns in the memory array, and p is an integer that is greater than or equal to one and represents the number of overtunneling prevention control circuits coupled to the multiplexer.
While particular embodiments of the present invention have been shown and described, it will be obvious to those skilled in the art that, based upon the teachings herein, changes and modifications may be made without departing from this invention and its broader aspects. Therefore, the appended claims are intended to encompass within their scope all such changes and modifications as are within the true spirit and scope of this invention.
Patent Grant
6853583
