Method for programming a reference cell

Information

  • Patent Grant
  • 7512009
  • Patent Number
    7,512,009
  • Date Filed
    Thursday, April 27, 2006
    18 years ago
  • Date Issued
    Tuesday, March 31, 2009
    15 years ago
Abstract
A method for programming one or more reference cells is described. The reference cell is programmed a predetermined amount, its program state is sensed relative to a prescribed cell on the same die (e.g., a memory cell or a golden bit cell), and the programming process continues until the reference cell fails a preselected read operation. In one preferred embodiment, the memory cell used during the reference cell programming process is the cell in the memory array having the highest native threshold value. In another preferred embodiment, the memory cell used during the reference cell programming process is a native cell that is on-board the die containing the memory array, but not a cell within the memory array.
Description
FIELD OF THE INVENTION

The present invention relates to sensing schemes for read operations on semiconductor devices, and, more particularly, to a method for programming a reference cell for use in a read operation.


BACKGROUND OF THE INVENTION

Memory devices, such as random access memory (RAM), read-only memory (ROM), non-volatile memory (NVM) and like, are known in the art. These devices provide an indication of the data which is stored therein by providing an output electrical signal. A device called a sense amplifier is used for detecting the signal and determining the logical content thereof. U.S. Pat. No. 4,916,671 to Ichiguchi describes one such sense amplifier.


In general, prior art sense amplifiers determine the logical value stored in a cell by comparing the output of the cell with a fixed reference voltage level. The aforementioned U.S. Pat. Nos. 6,134,156 and 6,128,226 describe an alternative circuit architecture in which a reference cell is used to generate a reference signal in lieu of a fixed reference voltage value.


When a memory cell is programmed or erased, the signal it generates differs from the reference signal by some margin. Since reading the cell's state should always result the same (i.e., either programmed or erased depending on the cell's state), introducing such margin is necessary to overcome imperfections in the reading process and to compensate for drifts in the cell's threshold voltage (e.g., caused by retention loss or disturbs). A reduction in the original margin due to imperfections in the reading process (e.g., due to operation at different operational conditions) is referred to as “margin loss.”


It is well understood that the placement of a reference signal to which an array cell signal can be compared during sensing can be achieved in a number of ways. When close to ground signals are sensed as in the aforementioned U.S. Pat. Nos. 6,134,156 and 6,128,226 patents, the reference cell signal develops at an intermediate rate between that of a programmed cell and an erased cell. When set this way, the array cells' signals on one side of the reference signal are determined to be programmed cells, while signals on the other side of the reference signal are determined to be erased cells. For example, array cells generating signals smaller than the reference signal are considered to be programmed and array cells generating signals larger than the reference signal are considered to be erased. Conventionally, such placement is achieved by using a reference cell whose current is between the erased and programmed cells' current levels. The reference cell's current level can be controlled by the reference cell's size, its programming level, or its gate voltage level. Furthermore, if voltage signals are used to detect the cells' contents, then the reference signal placement can be controlled by providing a different load capacitance on the reference cell compared to that of the array cells. However, if the array and the reference cells differ in their sizes, in their operating gate voltages, or in their loads then some margin loss will be introduced to the sensing scheme. On the other hand, placing the reference cells' signals by properly programming the reference cells (while operating the array and reference cells at identical conditions) minimizes the sensing scheme sensitivity to operating conditions, process parameters and environmental conditions variations, thereby minimizing the margin loss, if any, that is introduced to the sensing system.


When reference cell placement is by programming, it must be programmed a precise amount in order to achieve its intended purpose. There are difficulties attendant with reliable programming of a reference cell so as to minimize operating margin loss, as well as accurate placement of a programmed reference cell relative to the memory array cells. The present invention provides a method for programming reference cells to minimize margin loss and maximize cycling performance.


SUMMARY OF THE PRESENT INVENTION

The present invention provides a method for programming one or more reference cells, with the programming being performed relative to a prescribed cell on the same die as the reference cell (e.g., a memory cell or a golden bit cell).


According to one aspect of the invention, a method for programming a reference cell for use in an integrated circuit memory having an array of memory cells each exhibiting a native threshold voltage value is described. That method comprising the steps of first locating an address for the memory cell in the array that has the highest native threshold voltage value (VTNH). A reference cell is programmed a predetermined amount and its program state is sensed relative to the VTNH memory cell. The programming and sensing steps are repeated until the sensing step indicates that the reference cell has been programmed an amount sufficient to fail a first preselected read operation.


In a more particular methodology in accordance with this aspect of the invention, the locating step can include the steps of iteratively increasing a gate voltage applied to the memory cells and performing the first preselected read operation at each such applied gate voltage until a final gate voltage is identified at which all the memory cells in the array pass the first preselected read operation. Further, the first preselected read operation can exclude memory cells that have already passed the first preselected read operation at a previously applied gate voltage.


According to another aspect of the invention, a method for programming a reference cell for use in an integrated circuit memory having a plurality of memory cells each exhibiting a native threshold voltage value is described. That method locates an address for the VTNH cell by applying a first gate voltage value at which at least one memory cell fails a first preselected read operation and increasing the applied gate voltage until a final gate voltage value is reached at which each of the memory cells can just pass the first preselected read operation. The reference cell is programmed a predetermined amount and the program state of the reference cell relative to the VTNH memory cell is sensed by performing a second preselected read operation on the reference cell. The programming and sensing steps are repeated until the sensing step indicates that the reference cell has been programmed an amount sufficient to fail the second preselected read operation.


According to still another aspect of the invention, a method for programming a set of reference cells for use in performing respective read operations on an integrated circuit memory having a plurality of memory cells is described. That method locates the VTNH cell, determines a placement for a reference voltage read signal relative to the VTNH cell, places a reference voltage erase verify signal relative to the reference voltage read signal, and places a reference voltage program verify signal relative to the reference voltage read signal.


The foregoing methods can have their sensing steps performed relative to the VTNH memory cell and also relative to a native cell (a golden bit cell) on-board the same die.


The inventive method can be utilized to program a reference cell used with a memory array, a sliced array having one or more columns of memory cells, and redundant or auxiliary arrays.


These and other more specific aspects and features of the present invention can be appreciated from the acompanying Drawing Figures and Detailed Description of Certain Preferred Embodiments.





BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be understood and appreciated more fully from the following detailed description taken in conjunction with the drawings in which:



FIG. 1 illustrates a spread of native threshold voltage values for a number of memory cells on a die.



FIG. 2 illustrates target threshold voltage values for programming reference cells in order to achieve exemplary and desired operating margins.



FIG. 2A illustrates a practical example of operating margins and a preferred sequence of steps to establish the operating margins relative to a particular cell within the memory array.



FIG. 3A plots signal development through an array cell and a reference cell when both are driven by the same VCCR voltage supply.



FIG. 3B is the plot of FIG. 3A, showing the array cell being driven by a trimmed external supply voltage;



FIG. 3C is the plot of FIG. 3B, showing the reference cell at various program states and the results of the read operations.



FIG. 4 illustrates an overall process for programming a selected reference cell according to a first preferred method of the present invention.



FIG. 4A illustrates a process for locating the memory cell within a memory array having the highest native threshold voltage value (denoted “VTNH”) among the memory cells in the array.



FIG. 5 illustrates a process for programming a selected reference cell to a target value in accordance with the first preferred method.



FIG. 6 illustrates a process for more accurately determining the threshold voltage of a reference cell placed in accordance with the embodiments of the present invention.



FIG. 7 illustrates a process for programming a reference cell relative to another reference cell.



FIG. 8A plots signal development for a “golden bit” cell when driven by a trimmable external supply voltage, together with the VTNH cell and an exemplary reference cell.



FIG. 8B is the plot of FIG. 8A, showing the reference cell at various program states relative to a golden bit's signal development that has been placed using the trimmable external supply voltage and the results of various read operations.





DETAILED DESCRIPTION OF CERTAIN PREFERRED EMBODIMENTS

By way of overview and introduction, the present invention is described in connection with a methodology for programming a reference cell to enable sensing of the contents of a memory cell from close to ground level. Such a memory array is described in the aforementioned U.S. Pat. No. 6,134,156 and U.S. application Ser. No. 09/827,756, filed on even date herewith, entitled “Architecture And Scheme For A Non-Strobed Read Sequence.” By using a reference cell instead of a fixed threshold for comparison, a low voltage signal can be reliably processed irrespective of any changes in temperature or power supply level. The present invention has applicability to other sensing schemes, including A.C. and D.C. sensing techniques, and with read operations from either the source or drain side of a transistor, as can be appreciated by those of skill in the art.


Reference is made to FIG. 1 which illustrates the memory array cells native threshold voltage distribution. The memory cells on the die include both array cells that are addressed in a conventional manner, and additional cells such as auxiliary and reference cells that are used for a variety of purposes including quality control during manufacturing and read operations when the array cells are put into service. The native reference cells can lie anywhere along the threshold voltage (VT) axis. By way of example, the native threshold values for two reference cells are indicated as lying within the distribution curve, though that is not required.


The methods of the present invention enable precision programming of reference cells relative to a memory cell on the die, for example, relative to an array cell, another reference cell, or to a golden bit. Each of these techniques defines a related but different method described herein with reference to a flow diagram of a preferred embodiment for each such technique. These techniques are performed after manufacturing the devices, prior to placing the memory array into service. Generally, there are a number of target threshold voltage values that are set in a corresponding number of reference cells as a result of the reference cell programming process of the invention. That is to say, different types of read operations can be performed on the memory cell such as program verify, erase verify and operations in which temporary states of a cell in the progress of an erase or a program operation should be detected. The basic difference presented by these in-progress read cycles is the placement of the reference cell signal. Since the reference cell's state is not changed according to the type of the read operation, different reference cells should be used for each of these types of operations.


As illustrated in FIG. 2, target threshold voltage values are spaced relative to one another by predetermined margins M1, M2, M3 and M4 which are established to ensure reliable reads, read verifies, erase verifies, program verifies, or other read operations on the array memory cells. To minimize margin loss and maximize product endurance and reliability, it is critical that the margins M1 . . . M4 provide a reliable buffer and a tight operating window. As well, it is essential that the placement of the programmed reference cells is defined (at least for one of the reference cells) relative to the array threshold voltages (VT) distribution.



FIG. 2 illustrates an example of the margins M1 . . . M4 established and positioned relative to the memory cell in the array having the highest native threshold value (“VTNH”). As will be appreciated from the discussion below, the margins can be established and positioned relative to any cell, such as the VTNH cell, a golden bit cell, or another reference cell.



FIG. 2A illustrates a practical example of the margins M1 . . . M4 that can be well defined using the techniques of the present invention. Four margins are illustrated: cycling margin (“CM”), erase margin (“EM”), retention margin (“XM”) and refresh margin (“FM”). These margins are maintained between the reference cells, once their threshold voltage value have been programmed. In other words, the margins are controlled based on the response of each reference cell when driven by a standard voltage source VCCR. Since the reference programming procedure results in reference threshold voltages that may not be exactly as the original target levels, it is necessary that the actual reference cells threshold voltages maintain the following margins:

VT_RD_REF_actual−VTNH>=CM+EM
VT_RD_REF_actual−VT_EV_REF_actual>=EM
VT_RV_REF_actual−VT_RD_REF_actual>=XM
VT_PV_REF_actual−VT_RV_REF_actual>=FM

where VT_RD_REF_actual is the threshold voltage value of a reference cell programmed to implement a cell-contents read operation, VT_EV_REF_actual is the threshold voltage value of a reference cell programmed to implement an erase verify operation concerning the contents of a cell, VT_RV_REF_actual is the threshold voltage value of a reference cell programmed to implement a refresh verify operation concerning the contents of a cell, and VT_PV_REF_actual is the threshold voltage value of a reference cell programmed to implement a program verify operation concerning the contents of a cell.



FIG. 2A also illustrates a preferred sequence of steps to establish the operating margins relative to a particular cell within the memory array. The steps in FIG. 2A are shown relative to the VTNH cell, though other cells can be used such as the golden bit cell, as described below. Use of the VTNH cell facilitates discussion because it clearly illustrates the relationship between the required threshold voltages to be programmed into the reference cells and the required operating margins. Thus, with reference to the VTNH cell, a reference cell is programmed to be the RD_REF cell by programming it to have a margin M of CM+EM above VTNH. If a golden bit cell were used as the base line threshold voltage, then the margin M would most likely be greater by the difference between the threshold voltage of the golden bit cell (V_OTPG) and the VTNH cell.


Next, another reference cell is programmed to be the erase verify cell by programming that cell to have a margin EM below the actual threshold voltage of the RD_REF cell, as can be appreciated from FIG. 2A. Likewise, the RV_REF cell can be programmed to be XM above the actual threshold voltage of the RD_REF cell, and another cell is programmed to be the PV_REF cell by programming it to have a margin FM above the actual threshold voltage of the RV_REF. It will be appreciated that other arrangements of the reference cells' threshold voltages and of the sequence of programming the reference cells are possible and are part of the present invention. Furthermore, the placement of each reference cell can be made relative to an array cell, a “golden” cell, or another reference cell.


When put into service, it is desirable to drive both the array cells' and the reference cells' gates with the same voltage supply VCCR during all standard read operations. However, during reference programming, a trimmable external voltage supply (EXT_VCCR) driving the array cells' gates can be used to facilitate the reference cell programming procedure. The effect of trimming the external supply is illustrated in FIGS. 3A, 3B and 3C. In FIG. 3A, the development over time of voltage signals from a given, native reference cell is plotted together with the VTNH cell when both are driven by VCCR. We can define a “pass read ‘1’” operation to be when the array cell has a greater signal than the reference cell and a “fail read ‘1.’” operation to be when the array cell signal is lower than the reference signal. For example, the result of a read operation between the native VTNH cell and the native reference cell shown in FIG. 3A will be “fail read ‘1’”.


In FIG. 3B, the gate voltage applied to the VTNH cell has been reduced by a margin M, and the lower gate voltage results in slower signal development in that cell. Consequently, when compared to VTNH driven at the trimmed supply voltage level, the reference cell REF1 (driven by the standard gate voltage VCCR) must be programmed further in order to fail the read ‘1’ operation. FIG. 3C illustrates the reference cell REF1 at various program states along the process of being programmed including its native state in which it fails the read ‘1’ operation, an interim state in which the cell REF1 has been partially programmed, yet still fails the read ‘1’ operation, and a state in which the further programmed reference cell signal is smaller than the array cell's signal and so it passes the read ‘1’ test. At this last state, it is ensured that the threshold voltage of the reference cell has been raised by at least the M margin.


A read operation or “sensing” of a cell can be performed as described in the aforementioned, co-pending U.S. Pat. No. 6,535,434, issued Mar. 18, 2003, entitled “Architecture And Scheme For A Non-Strobed Read Sequence,” which describes the steps taken to sense a close to ground signal and other sensing methods (e.g., DC current sensing).



FIG. 4 illustrates the overall process for placing a selected reference cell with regard to a first preferred method of the present invention. At step 410, the VTNH memory cell within the memory cell array is located. That cell is then used as a reference for verifying the programming of the selected reference cell, as indicated at step 420. All margins are then established with respect to the threshold voltage for that cell, which is the highest threshold voltage of any cell in the memory array. This method places the VT for the selected reference cell by programming the cell in accordance with a prescribed criterion at step 430. Generally, the criterion is that the VT for the reference cell is VTNH+M, where M is a prescribed margin. As a specific example, if the selected reference cell is the RD_REF cell, then M is CM+EM and target reference cell VT is set to be greater than or equal to VTNH+CM+EM. The reference cell is preferably programmed a predetermined amount until placed, as described in detail with regard to the flow diagrams of FIGS. 5-8.


With reference now to FIG. 4A, step 410 of FIG. 4 for locating the VTNH cell proceeds by first setting an initial, low voltage level (EXT_VCCR) that is externally applied to the gates of respective memory cells, in succession, as the array is parsed, as shown at step 450. Meanwhile, a predetermined gate voltage level such as the standard gate voltage level at read, VCCR, is applied to the gate of the reference cell. It will be recalled that the reference cell has a similar structure (e.g., size, access path, environment, etc.) to the array cells and has a matched load; thus, as was illustrated in FIG. 3B, the comparatively low voltage applied to the gate of the still native array cells by EXT_VCCR at step 452 causes the array cell to develop a smaller signal than the reference cell. The memory array is then parsed starting with a first array cell address, as shown at step 454.


At a prescribed moment after applying these gate voltages, for example, as described in the aforementioned patent application entitled “Architecture And Scheme For A Non-Strobed Read Sequence,” a read operation is performed on consecutive memory cells in the array against the reference cell (see step 456) until a fail-to-read-1 condition is detected. Assuming that EXT_VCCR has been initially set to a low enough value (e.g., 2 Volts), then the first read operation is expected to fail a read ‘1’ operation, as tested at step 458, since both cells are native and are expected to have similar characteristics. If the read does not fail “read ‘1’” (i.e., it passes), then if the entire array has not already been parsed, as tested at step 460, then the next array cell is selected at step 462 and the process is repeated, starting again at step 456, until a fail read ‘1’ is detected. The address of the array cell that failed the test is stored as well as the EXT_VCCR level, at step 464. The voltage EXT_VCCR is increased at step 466, preferably by a fixed small increment J (typically smaller than 100 mV), and, if the array has not been completely parsed, as tested at step 468, this read operation is repeated starting from the next array cell address at step 470 until the fail read ‘1’ is detected (at step 458). On the other hand, if the array has been completely parsed yet the present address has failed to read ‘1,’ then the die is considered as a bad die, and the operator is advised of this at step 469. A bad die can also be detected if the level of EXT_VCCR has been incremented beyond a prescribed maximum, as shown at step 471. The maximum level for EXT_VCCR can be established based on empirical data for processing similar devices.


After an array cell address is stored at step 464 and a valid next address is obtained at step 470, further addresses are examined to determine whether EXT_VCCR must be incremented and to store a new array cell address. As additional array cells, if any, fail the read ‘1’ test at step 458, the already stored address and EXT_VCCR level can be replaced by the present address and EXT_VCCR level for the presently read array cell, and the EXT_VCCR level can be increased again, with the process continuing from the next array cell address, assuming there are more addresses to parse (as tested at step 468).


As described above, if the read operation repeats starting at the last stored cell address and continues to pass the read ‘1’ operation at step 458 then the last address of the array will eventually be reached, as tested at step 460. Once the entire array has already parsed, the flow proceeds to step 480 at which point a test is made to ensure that the die is not defective by examining the current setting for EXT_VCCR. If EXT_VCCR is above an established minimum setting, then the last stored address is the VTNH memory cell address because it was the last cell that failed to be read as a ‘1’. Also, the last stored level for EXT_VCCR is the V_VTNH level. These values are preserved for use in reference cell programming, as indicated at step 482. On the other hand, if EXT_VCCR is not above an established minimum level, the operator is advised that the die is defective, as indicated at step 484.


This process enables each cell to be read only once, in other words, a single parse of the array locates the VTNH memory cell address, as indicated at step 462. This is a preferred method although other flows for parsing the array can be used although they will probably result in a longer process due to the larger number of read operations (e.g., for each EXT_VCCR level start the read operations at a specific first address and parse the array till a fail read ‘1’ is detected).


The increment step of the EXT_VCCR supply in this process determines the accuracy of the VTNH cell identification. Accuracy of identification means that there is no cell in the array whose threshold voltage is larger than VTNH plus the EXT_VCCR increment step level, where VTNH is the threshold voltage of the cell identified as the VTNH cell. Assume for example that a 100 mV increment step is used. At the end of the process, the stored address indicates the cell that failed read ‘1’ at the stored EXT_VCCR level. The remaining cells that were read in the next loop of read cycles, applying to their gates the stored EXT_VCCR level plus the 100 mV increment step, pass the read ‘1’ test (otherwise the address of the first failing one would have been stored). However, if a smaller increment step was used, for example 50 mV, then one of those remaining cells could have failed the read ‘1’ test. Thus, if a more accurate identification of the VTNH cell is required, then either one of the following two options can be adopted:

    • 1. Use a smaller EXT_VCCR increment step for the whole process.
    • 2. Use a coarse EXT_VCCR increment step for a first pass of the whole array and a finer EXT_VCCR increment step for a second pass of the cells starting at the stored VTNH cell address.


A separate aspect of this procedure is that a final voltage value for EXT_VCCR which is arrived at once the array has been parsed provides an indication as to whether the threshold voltage of the reference cell or an array cell is outside of a standard distribution of values. This value is therefore useful in determining the quality of the memory array. If at the end of the process the final EXT_VCCR level is the starting level (EXT_VCCR_min) then this indicates that the reference cell threshold voltage is significantly larger than the array cells' threshold voltages. As well, by setting a maximum level for the EXT_VCCR supply (EXT_VCCR_max), if this level is reached then it indicates that the reference cell threshold voltage is significantly lower than at least one of the array cells' threshold voltages. Any indication of such non-standard distribution of threshold voltages can be used as a quality check and as a basis for a decision such as to reject the part (steps 469 and 484 in FIG. 4A).


Moreover, while the foregoing steps are operative to locate the VTNH memory cell, they also can serve as a blank test for the memory array, in which all cells are read once, and so the locating process does not increase the sort time during manufacturing.


Optionally, the final value of EXT_VCCR is stored, and, if desired, the value of EXT_VCCR at each cell at which a read ‘1’ operation fails can be stored.


With reference now to FIG. 5, a process for programming a selected reference cell REF1 to a target value in accordance with the first preferred method is described. At step 510, the VTNH cell, once located, is driven by a trimmable voltage source EXT_VCCR at a voltage value of

EXT_VCCR=VCCR−M

where M is a prescribed margin suitable for the selected reference cell. For purposes of illustration, assume that the reference cell REF1 is the read reference cell RD_REF shown in FIG. 2A for which

M=CM+EM.

Meanwhile, as indicated at step 520, the reference cell REF1 is driven with a standard supply voltage having a supply voltage value of VCCR. As should be appreciated, the trimmable voltage source EXT_VCCR permits the VTNH array cell signal to be temporarily placed at the target location of the reference cell REF1 signal, by applying a reduced gate voltage during this reference cell programming phase test as compared to the VCCR voltage ordinarily applied to the VTNH (and other array cells) when the memory cell is put into service.


At step 530, a program pulse is applied to the reference cell REF1. A test is then made at step 540 to determine whether the result of a read operation of the VTNH cell (driven by the trimmed gate voltage EXT_VCCR ) against the reference cell REF1 (driven by the standard gate voltage at read, VCCR) is PASS read ‘1’ or FAIL. This test is performed by sensing the program state of the reference cell relative to the VTNH cell (driven by the trimmed gate voltage EXT_VCCR ) preferably using the standard sensing scheme that is used when the device is put into service. If it does not pass “read ‘1’”, then the reference cell has not been programmed to the location at which the VTNH cell signal has been temporarily placed (by the trimmed gate voltage EXT_VCCR), and further programming pulses are required. The process loops back to step 530 so that an additional program pulse can be applied, and then a read test as at step 540 is again performed. The program pulses can be applied in a fixed increment (i.e., a predetermined amount), or an algorithm can be employed to enhance the process. For one suitable algorithm, see U.S. application Ser. No. 10/300,924, filed on even date herewith, entitled “Programming and Erasing Methods for A Reference Cell of An NROM Array,” which application is a continuation-in-part application of U.S. application Ser. No. 09/730,586 , filed Dec. 7, 2000, which is a continuation-in-part application of U.S. Ser. No. 09/563,923, filed May 4, 2000. The process flow ends when the test at step 540 indicates that the reference cell REF1 has been placed at the target location, that is, when it has been programmed an amount sufficient to pass the read ‘1’ test.


The incremental programming of the reference cell REF1 is illustrated in FIG. 3C, in which the VTNH cell is driven by the trimmed gate voltage EXT_VCCR (step 510) while an initially native reference cell REF1 is driven by the standard gate voltage at read (step 520). The initially native reference cell REF1 exhibits signal development that is much more rapid than that of the VTNH cell, as shown by the curve “VREF1 (native)”, when the cells are driven with such applied gate voltages. That curve fails to read ‘1,’ and so the process of FIG. 5 continues by applying one or more program pulses to the reference cell REF1 at step 530. These program pulses have the effect of partially programming the reference cell and reducing its rate of signal development. That is reflected in the curve “VREF1 (partial PGM'ing)”, which also fails to read ‘1.’ Consequently, additional program pulses are applied at step 530 and the test at step 540 is repeated, and this loop continues until a PASS read ‘1’ is detected, namely, when the signal development at the reference cell REF1 is akin to the curve “VREF1 (done PGM'ing)”. Of course, different tests can be arranged to provide a boundary condition to gauge when to stop programming the reference cell (e.g., external read of the cells currents, or other tests).


The process flow of FIG. 5 permits the reference cell REF1 to be positioned in the vicinity of the target; however, since each program pulse increments the VT of the reference cell by a predetermined amount, the actual VT of the reference cell after being programmed by the process flow of FIG. 5 is typically above the target level. Consequently, if further reference cells REF2 . . . REFx are to be programmed relative to the VT of that reference cell REF1, e.g., relative to the RD_REF reference cell, then knowledge of the actual level of the REF1 threshold voltage, e.g., VT_RD_REF_actual, may be desired. The VT level of the reference cell REF1 relative to the VTNH cell can be refined by iteratively decreasing EXT_VCCR starting from EXT_VCCR=VCCR−M by a controlled amount which is a smaller value than the VT increment as a result of a programming pulse.



FIG. 6 illustrates a process flow for refining the position of reference cell REF1. It should be understood that the just-programmed reference cell itself is not being partially erased, but rather the trimmable voltage source EXT_VCCR is adjusted so that the actual VT of the reference cell REF1 can be more accurately determined.


At step 610, the VTNH cell is driven with EXT_VCCR, that is with the voltage value that was applied at step 510 described above. Namely,

EXT_VCCR=VCCR−M

Meanwhile, at step 620, the reference cell REF1 is driven with a standard gate voltage VCCR. At step 630, a counter N is set to zero for use in flow of FIG. 6 to monitor the number of times that the trimmable voltage source has been trimmed a prescribed amount. Other techniques can be used as will be appreciated for tracking the number of adjustments or the overall adjustment to the voltage source EXT_VCCR.


At step 640, the voltage value of EXT_VCCR is decreased by a prescribed amount K, which amount is preferably 50 mV or less, and at step 650 the counter N is incremented. At step 660 the VTNH array cell is read against the reference cell REF1 and a test is made to determine whether the result is PASS read ‘1’. It is expected that in the first loop that the VTNH cell will pass this test, with the process flow looping to step 640 to again decrease the voltage value of EXT_VCCR and then increase the counter N (at step 650). Ultimately, however, EXT_VCCR will be reduced such that the VTNH cell no longer passes the read ‘1’ test, and, at that point, the actual VT of the reference cell REF1 is determined at step 670 to be:

VT_REF1_actual=VTNH+VCCR−EXT—VCCR  (1)

Expressed another way, the actual VT of the reference cell REF1 is:

VT_REF1_actual=VTNH+M+N*K,

where N is the number of loops in which the voltage value emanating from the EXT_VCCR source was decreased by the constant amount K.


A further process flow can be used in like manner to increment EXT_VCCR from its new-found value to a higher value by selecting a new voltage interval E, where E<K (e.g., E is 25 mV or less). Yet a further process flow can again decrement EXT_VCCR, and so on, to more accurately place the threshold voltage of the reference cell REF1.


With reference now to the flow diagram of FIG. 7, the steps taken for programming further reference cells REF2 . . . REFx relative to the reference cell REF1 is described. Preferably, the actual threshold voltage of the reference cell REF1 has been identified through the process flow of FIG. 6. In this first preferred embodiment, the reference cells are located relative to the VTNH cell, which was located in connection with the process flow of FIG. 4, described above.


The threshold voltage for the reference cell REF2 is to be placed at:

VT_REF2=VT_REF1_actual+M2,  (2)

where M2 is the margin between REF1 and REF2. M2 can be positive or negative. If the threshold voltage of the reference cell REF2 is lower than that of the reference cell REF1, then M2 is a positive voltage value; otherwise, M2 is a negative voltage value. For example, if REF1 is the RD_REF cell and REF2 is the RV_REF cell, then the margin M2 is XM, as shown in FIG. 2A. For the EV_REF cell in FIG. 2A a negative M2 margin (EM) should be used. Using equation (1) to solve equation (2) for EXT_VCCR, we see that

EXT_VCCR=VCCR−VT_REF1_actual+VTNH−M2
EXT_VCCR=VCCR−M1−N*K−M2

Where, M1 is the original target margin of the REF1 reference cell, N*K is the amount of over programming introduced to the REF1 cell past the original target, and M2 is a positive or negative target margin between the REF1 and REF2 threshold voltages.


For example, if REF2 to be programmed relative to REF1 is the EV_REF cell shown in FIG. 2A then M2=−EM, i.e., a negative margin. Similarly, if REF2 is the RV_REF cell shown in FIG. 2A then M2=XM, i.e., a positive margin.


Thus, to place REF2 relative to REF1, the gate of the VTNH cell is driven by EXT_VCCR while the gate of REF2 is driven by the standard voltage VCCR, as indicated at steps 710 and 720, respectively. At step 730, a program pulse of a predetermined amount is applied to the reference cell REF2 to move it from its native state and place it at its target programmed state relative to REF1. The cell REF2 is being placed relative to REF1 by sensing its value relative to the VTNH cell, while the VTNH cell is driven by a gate voltage which places the VTNH signal taking into account the actual placement of REF1 and the required margin between the respective VTs of REF1 and REF2. At step 740, the partially programmed reference cell REF2 is tested to see if it still passes a read ‘1’ test, in which case it has not been sufficiently programmed to place it M2 from the reference cell REF1. If the cell fails the read ‘1’ test, as is initially expected since both cells are essentially native, then a further program pulse is applied at step 730 and the test at step 740 is repeated until the reference cell REF2 fails the read ‘1’ test. At that point, the target placement for REF2 has been achieved.


As described above in connection with the target programming of the reference cell REF1, a more accurate location of the actual placement of the reference cell REF2 can be achieved using a process flow as in FIG. 6, this time applied to the reference cell REF2 and this time setting the actual VT at step 650 to be:

VT_REF2=VTNH+M1+N1*K+M2+N2*K.

This refinement may be necessary if the reference cell REF2 is to be used for placing yet a further reference cell, or for other purposes, such as accurate monitoring of the reference cell retention loss after bake.


Further programming of reference cells will follow one of the flows described above, depending if they are programmed relative to the VTNH cell or relative to another, already programmed, reference cell.


It is know in the art that, after manufacturing, memory devices are tested to detect malfuctioning or marginal devices. As part of these tests the array cells may be programmed and then the device is introduced to a relatively high temperature cycle. After this high temperature cycle, the threshold voltage of the array cells may drop by some amount. This threshold voltage drop is known as the “retention loss.” Since the reference cells are also programmed, their threshold voltage may also drop by some amount (which may be different than that of the array cells due to differences in the programming levels of the array cells and the reference cells). Since the threshold voltages of all the programmed memory cells are affected, whether a reference cell or an array cell (including the VTNH cell), there is no relative way to determine the cell's state. A non-relative way to determine the cells threshold voltage state, such as by an external measurement of the cells currents, is very expensive in terms of test time. Thus, an internal relative measurement of the array cells and the reference cells threshold voltages is desired.


A native cell which has never been programmed and which is on-board the same die is utilized to provide an internal relative measurement, in accordance with another aspect of the present invention. This native cell, referred to herein as a “golden cell” or “golden bit,” permits the internal read mechanism to be used not only for the reference cells' programming flow as described above in connection with FIGS. 4-7, but also for detection of array cell and reference cell threshold voltage changes after a high temperature cycle. The process flows described above can each be modified such that the reference cells' signal placements are performed relative to this golden bit in conjunction with the native cell having the maximum threshold voltage (that is, the VTNH cell). Now, even after a high temperature cycle, the reference cells' threshold voltages can be re-estimated against the golden bit because the threshold voltage of this unprogrammed (native) cell does not change as a result of any high temperature cycle(s).


The golden bit is a memory cell having the same size, the same environment, similar loads and matched or similar access paths as an array memory cell. However, the golden bit is usually not among the memory cells in the memory array, but rather is typically included in an auxiliary array. Referring back to FIG. 4A, the process flow for locating the VTNH cell, the V_VTNH voltage level found at step 462 is the EXT VCCR level at which the VTNH cell signal just passed a read ‘1’ (i.e., became higher than) the reference cell signal. FIG. 8A plots signal development for the golden bit cell when driven by the trimmable external supply EXT_VCCR, together with the VTNH cell and an exemplary reference cell. In particular, the native golden bit signal can be placed by trimming the voltage applied to its gate using a method similar to that of FIG. 4A for finding V_VTNH. The voltage value of a trimmable gate voltage source EXT_VCCR is adjusted until the golden bit just passes a read ‘1’ test, and that defines the value of V_GB. The difference D between the gate voltages applied to the VTNH cell and the golden bit cell,

DVTNH-GB=V_VTNH−V—GB

can be used to program the reference cells and to now determine the actual threshold voltages of the programmed reference cells relative to the golden bit instead of the VTNH cell. As shown in FIG. 8B, reference cells can be programmed relative to the golden bit using a target threshold voltage level of:

VT_REF1=VTNH+M1.

The method consists of placing the golden bit signal at the target place of the programmed reference cell, applying a programming pulse to the reference cell, and reading the contents of the cell to determine its contents. These steps are repeated until a PASS read ‘1’ is detected, as described above in connection with FIGS. 4-7. Thus, the flow is as described above for programming the reference cell using the VTNH cell, except that in the previous discussion the VTNH cell signal was placed at the target place of the programmed reference cell whereas now the golden bit cell's signal is placed at the target place of the programmed reference signal. The golden bit cell's signal is placed by driving its gate with

EXT_VCCR=VCCR−M1−DVTNH-GB.

Preferably, the reference cells are contained in a reference unit which includes a selector that can select which of several reference cells to use for a given operation, wherein each reference cell is programmed as described above to a prescribed level.


If the reference cells are NROM cells, they can be used as either single or two bit cells. When used as single bit cells one of the two bits in the cell is programmed as explained above while the other bit can be maintained native, programmed or erased. On the other hand, if the two bits of the NROM reference cell are used as references, then each of them is programmed as explained above.


Persons skilled in the art will appreciate that the present invention is not limited to what has been particularly shown and described hereinabove. Rather the scope of the present invention is defined only by the claims which follow.

Claims
  • 1. A method for programming a set of reference cells for use in performing respective read operations on an integrated circuit memory having a plurality of memory cells, comprising the steps of: i) driving a golden cell on the die with a predetermined external gate voltage value;ii) determining a placement for a reference voltage read signal relative to the golden cell;iii) placing a reference voltage erase verify signal relative to the reference voltage read signal; andiv) placing a reference voltage program verify signal relative to the reference voltage read signal.
  • 2. The method as in claim 1, wherein the predetermined external gate voltage value is the difference D between a first external gate voltage value at which the memory cell in the array which has the highest native threshold voltage value (VTNH) just passes the first preselected read operation and a second external gate voltage value at which a golden cell on the die cell just passes the first preselected read operation (V_GB).
CROSS-REFERENCE TO RELATED APPLICATIONS

The present application is a divisional application of U.S. Ser. No. 10/454,820, filed Jun. 5, 2003 now U.S. Pat. No. 7,064,983, which is herein incorporated by reference, which application is a divisional application of U.S. Ser. No. 09/827,757, filed Apr. 5, 2001, entitled “Reference Programming” (now issued as U.S. Pat. No. 6,584,017), which is herein incorporated by reference. This application is related to U.S. Pat. No. 6,128,226, issued Oct. 3, 2000 entitled “Method and Apparatus for Operating with a Close to Ground Signal,” to U.S. Pat. No. 6,134,156, issued Oct. 17, 2000 entitled “Method for Initiating a Retrieval Procedure in Virtual Ground Arrays,” to U.S. Pat. No. 6,535,434, issued Mar. 18, 2003 entitled “Architecture And Scheme For A Non-Strobed Read Sequence,” and U.S. Pat. No. 6,490,204, issued Dec. 3, 2002, entitled “Programming and Erasing Methods For A Reference Cell Of An NROM Array,” the foregoing patents and patent applications being incorporated by reference in their entireties as if set forth herein.

US Referenced Citations (323)
Number Name Date Kind
3881180 Gosney, Jr. Apr 1975 A
3952325 Beale et al. Apr 1976 A
4145703 Blanchard et al. Mar 1979 A
4247861 Hsu et al. Jan 1981 A
4373248 McElroy Feb 1983 A
4386417 Collins et al. May 1983 A
4404747 Collins Sep 1983 A
4435786 Tickle Mar 1984 A
4494016 Ransom et al. Jan 1985 A
4507673 Aoyama Mar 1985 A
4613956 Paterson et al. Sep 1986 A
4663645 Komori et al. May 1987 A
4665426 Allen et al. May 1987 A
4672409 Takei et al. Jun 1987 A
4733105 Shin et al. Mar 1988 A
4758869 Eitan et al. Jul 1988 A
4760555 Gelsomini et al. Jul 1988 A
4780424 Holler et al. Oct 1988 A
4839705 Tigelaar et al. Jun 1989 A
4857770 Partovi et al. Aug 1989 A
4916671 Ichiguchi Apr 1990 A
4961010 Davis Oct 1990 A
4992391 Wang Feb 1991 A
5029063 Lingstaedt et al. Jul 1991 A
5081371 Wong Jan 1992 A
5086325 Schumann et al. Feb 1992 A
5094968 Schumann et al. Mar 1992 A
5117389 Yiu May 1992 A
5120672 Mitchell et al. Jun 1992 A
5142495 Canepa Aug 1992 A
5168334 Mitchell et al. Dec 1992 A
5172338 Mehrotra et al. Dec 1992 A
5204835 Eitan Apr 1993 A
5237213 Tanoi Aug 1993 A
5260593 Lee Nov 1993 A
5268861 Hotta Dec 1993 A
5280420 Rapp Jan 1994 A
5295092 Hotta Mar 1994 A
5327383 Merchant et al. Jul 1994 A
5334555 Sugiyama et al. Aug 1994 A
5338954 Shimoji Aug 1994 A
5352620 Komori et al. Oct 1994 A
5357134 Shimoji Oct 1994 A
5366915 Kodama Nov 1994 A
5375094 Naruke Dec 1994 A
5386388 Atwood et al. Jan 1995 A
5400286 Chu et al. Mar 1995 A
5402374 Tsuruta et al. Mar 1995 A
5436478 Bergemont Jul 1995 A
5463586 Chao et al. Oct 1995 A
5477499 Van Buskirk et al. Dec 1995 A
5490110 Sawada et al. Feb 1996 A
5495440 Asakura Feb 1996 A
5523972 Rashid et al. Jun 1996 A
5532962 Auclair et al. Jul 1996 A
5534804 Woo Jul 1996 A
5553030 Tedrow et al. Sep 1996 A
5557221 Taguchi et al. Sep 1996 A
5557570 Iwahashi Sep 1996 A
5559687 Nicollini et al. Sep 1996 A
5568419 Atsumi et al. Oct 1996 A
5590068 Bergemont Dec 1996 A
5590074 Akaogi et al. Dec 1996 A
5592417 Mirabel Jan 1997 A
5596527 Tomioka et al. Jan 1997 A
5599727 Hakozaki et al. Feb 1997 A
5600586 Lee Feb 1997 A
5608679 Mi et al. Mar 1997 A
5612642 McClinyock Mar 1997 A
5617357 Haddad et al. Apr 1997 A
5636288 Bonneville et al. Jun 1997 A
5638320 Wong et al. Jun 1997 A
5661060 Gill et al. Aug 1997 A
5663907 Frayer et al. Sep 1997 A
5672959 Der Sep 1997 A
5675280 Nomura Oct 1997 A
5675537 Bill et al. Oct 1997 A
5683925 Irani et al. Nov 1997 A
5696929 Hasbun et al. Dec 1997 A
5708608 Park et al. Jan 1998 A
5715193 Norman Feb 1998 A
5717581 Canclini Feb 1998 A
5726946 Yamagata et al. Mar 1998 A
5751037 Aozasa et al. May 1998 A
5760634 Fu Jun 1998 A
5768184 Hayashi et al. Jun 1998 A
5768192 Eitan Jun 1998 A
5774395 Richart et al. Jun 1998 A
5781476 Seki et al. Jul 1998 A
5781478 Takeuchi et al. Jul 1998 A
5784314 Sali et al. Jul 1998 A
5808506 Tran Sep 1998 A
5815435 Van Tran Sep 1998 A
5822256 Bauer et al. Oct 1998 A
5825683 Chang Oct 1998 A
5828601 Hollmer et al. Oct 1998 A
5835935 Estakhri et al. Nov 1998 A
5861771 Matsuda et al. Jan 1999 A
5862076 Eitan Jan 1999 A
5875128 Ishizuka Feb 1999 A
5877537 Aoki Mar 1999 A
5880620 Gitlin et al. Mar 1999 A
RE36179 Shimoda Apr 1999 E
5892710 Fazio et al. Apr 1999 A
5903031 Yamada et al. May 1999 A
5910924 Tanaka et al. Jun 1999 A
5920503 Lee et al. Jul 1999 A
5920507 Takeuchi et al. Jul 1999 A
5930195 Komatsu Jul 1999 A
5933366 Yoshikawa Aug 1999 A
5933367 Matsuo et al. Aug 1999 A
5946258 Evertt et al. Aug 1999 A
5966603 Eitan Oct 1999 A
5990526 Bez et al. Nov 1999 A
5999494 Holzrichter Dec 1999 A
6000006 Bruce et al. Dec 1999 A
6002614 Banks Dec 1999 A
6005423 Schultz Dec 1999 A
6005805 Takeuchi Dec 1999 A
6037627 Kitamura et al. Mar 2000 A
6040610 Noguchi et al. Mar 2000 A
6064226 Earl May 2000 A
6064251 Park May 2000 A
6074916 Cappelletti Jun 2000 A
6075402 Ghilardelli et al. Jun 2000 A
6078518 Chevallier Jun 2000 A
6081456 Dadashev Jun 2000 A
6094095 Murray et al. Jul 2000 A
6107862 Mukainakano et al. Aug 2000 A
6117714 Beatty Sep 2000 A
6118207 Ormerod et al. Sep 2000 A
6122198 Haddad et al. Sep 2000 A
6128226 Eitan et al. Oct 2000 A
6128227 Kim Oct 2000 A
6130572 Ghilardelli et al. Oct 2000 A
6130574 Bloch et al. Oct 2000 A
6133095 Eitan et al. Oct 2000 A
6134156 Eitan Oct 2000 A
6137718 Reisinger Oct 2000 A
6150800 Kinoshita et al. Nov 2000 A
6154081 Pakkala et al. Nov 2000 A
6156149 Cheung et al. Dec 2000 A
6157242 Fukui Dec 2000 A
6163484 Uekubo Dec 2000 A
6175523 Yang et al. Jan 2001 B1
6181597 Nachumovsky Jan 2001 B1
6188211 Rincon-Mora et al. Feb 2001 B1
6195196 Kimura et al. Feb 2001 B1
6198342 Kawai Mar 2001 B1
6205059 Gutala et al. Mar 2001 B1
6208200 Arakawa Mar 2001 B1
6208557 Bergemont et al. Mar 2001 B1
6214666 Mehta Apr 2001 B1
6215697 Lu et al. Apr 2001 B1
6215702 Derhacobian et al. Apr 2001 B1
6218695 Nachumovsky Apr 2001 B1
6219290 Chang et al. Apr 2001 B1
6222762 Guterman et al. Apr 2001 B1
6229734 Watanabe May 2001 B1
6246555 Tham Jun 2001 B1
6252442 Malherbe Jun 2001 B1
6256231 Lavi et al. Jul 2001 B1
6261904 Pham et al. Jul 2001 B1
6265268 Halliyal et al. Jul 2001 B1
6266281 Derhacobian et al. Jul 2001 B1
6272047 Mihnea et al. Aug 2001 B1
6275414 Randolph et al. Aug 2001 B1
6281545 Liang et al. Aug 2001 B1
6285246 Basu Sep 2001 B1
6285614 Mulatti et al. Sep 2001 B1
6297096 Boaz Oct 2001 B1
6297974 Ganesan et al. Oct 2001 B1
6304485 Harari et al. Oct 2001 B1
6307784 Hamilton et al. Oct 2001 B1
6324094 Chevallier Nov 2001 B1
6326265 Liu et al. Dec 2001 B1
6330192 Ohba et al. Dec 2001 B1
6335874 Eitan Jan 2002 B1
6337502 Eitan et al. Jan 2002 B1
6339556 Watanabe Jan 2002 B1
6343033 Parker Jan 2002 B1
6346442 Aloni et al. Feb 2002 B1
6348381 Jong Feb 2002 B1
6353356 Liu Mar 2002 B1
6353555 Jeong Mar 2002 B1
6356469 Roohparvar et al. Mar 2002 B1
6359501 Lin et al. Mar 2002 B2
6374337 Estakhri Apr 2002 B1
6385086 Mihara et al. May 2002 B1
6396741 Bloom et al. May 2002 B1
6400209 Matsuyama et al. Jun 2002 B1
6400607 Pasotti et al. Jun 2002 B1
6407537 Antheunis Jun 2002 B2
6410388 Kluth et al. Jun 2002 B1
6417081 Thurgate Jul 2002 B1
6418506 Pashley et al. Jul 2002 B1
6421277 Tsunesada Jul 2002 B2
6426898 Mihnea et al. Jul 2002 B1
6429063 Eitan Aug 2002 B1
6433624 Grossnickle et al. Aug 2002 B1
6436766 Rangarajan et al. Aug 2002 B1
6436768 Yang et al. Aug 2002 B1
6438035 Yamamoto et al. Aug 2002 B2
6440797 Wu et al. Aug 2002 B1
6445030 Wu et al. Sep 2002 B1
6449190 Bill Sep 2002 B1
6452438 Li Sep 2002 B1
6456528 Chen Sep 2002 B1
6456533 Hamilton et al. Sep 2002 B1
6458656 Park et al. Oct 2002 B1
6458677 Hopper et al. Oct 2002 B1
6472706 Widdershoven et al. Oct 2002 B2
6490204 Bloom et al. Dec 2002 B2
6493266 Yachareni et al. Dec 2002 B1
6512701 Hamilton et al. Jan 2003 B1
6519180 Tran et al. Feb 2003 B2
6528390 Komori et al. Mar 2003 B2
6532173 Lioka et al. Mar 2003 B2
6535020 Yin Mar 2003 B1
6535434 Maayan et al. Mar 2003 B2
6537881 Rangarjan et al. Mar 2003 B1
6538270 Randolph et al. Mar 2003 B1
6541816 Ramsbey et al. Apr 2003 B2
6552387 Eitan Apr 2003 B1
6555436 Ramsbey et al. Apr 2003 B2
6559500 Torii May 2003 B2
6562683 Wang et al. May 2003 B1
6566194 Ramsbey et al. May 2003 B1
6566699 Eitan May 2003 B2
6577514 Shor et al. Jun 2003 B2
6577532 Chevallier Jun 2003 B1
6577547 Ukon Jun 2003 B2
6583005 Hashimoto et al. Jun 2003 B2
6583479 Fastow et al. Jun 2003 B1
6584017 Maayan et al. Jun 2003 B2
6593606 Randolph et al. Jul 2003 B1
6608526 Sauer Aug 2003 B1
6614295 Tsuchi Sep 2003 B2
6614686 Kawamura Sep 2003 B1
6614692 Maayan et al. Sep 2003 B2
6617179 Kim Sep 2003 B1
6617215 Halliyal et al. Sep 2003 B1
6624672 Confaloneri et al. Sep 2003 B2
6627555 Eitan et al. Sep 2003 B2
6630384 Sun et al. Oct 2003 B1
6633496 Maayan et al. Oct 2003 B2
6633956 Mitani Oct 2003 B1
6639271 Zheng et al. Oct 2003 B1
6639849 Takahashi et al. Oct 2003 B2
6642148 Ghandehari et al. Nov 2003 B1
6642573 Halliyal et al. Nov 2003 B1
6642586 Takahashi Nov 2003 B2
6643170 Huang et al. Nov 2003 B2
6643181 Sofer et al. Nov 2003 B2
6645801 Ramsbey et al. Nov 2003 B1
6649972 Eitan Nov 2003 B2
6653190 Yang et al. Nov 2003 B1
6653191 Yang et al. Nov 2003 B1
6654296 Jang et al. Nov 2003 B2
6664588 Eitan Dec 2003 B2
6670241 Kamal et al. Dec 2003 B1
6670669 Kawamura Dec 2003 B1
6674138 Halliyal et al. Jan 2004 B1
6677805 Shor et al. Jan 2004 B2
6680509 Wu et al. Jan 2004 B1
6686242 Willer et al. Feb 2004 B2
6717207 Kato Apr 2004 B2
6723518 Papsidero et al. Apr 2004 B2
6731542 Le et al. May 2004 B1
6738289 Gongwer et al. May 2004 B2
6744692 Shiota et al. Jun 2004 B2
6765259 Kim Jul 2004 B2
6768165 Eitan Jul 2004 B1
6788579 Gregori et al. Sep 2004 B2
6791396 Shor et al. Sep 2004 B2
6794249 Palm et al. Sep 2004 B2
6831872 Matsuoka Dec 2004 B2
6836431 Chang Dec 2004 B2
6871258 Micheloni et al. Mar 2005 B2
6885585 Maayan et al. Apr 2005 B2
6912160 Yamada Jun 2005 B2
6917544 Maayan et al. Jul 2005 B2
6928001 Avni et al. Aug 2005 B2
6937523 Eshel Aug 2005 B2
6967872 Quader et al. Nov 2005 B2
6996692 Kouno Feb 2006 B2
7064983 Maayan et al. Jun 2006 B2
7079420 Shappir et al. Jul 2006 B2
20010006477 Banks Jul 2001 A1
20020004921 Muranaka et al. Jan 2002 A1
20020064911 Eitan May 2002 A1
20020140109 Keshavarzi et al. Oct 2002 A1
20020145465 Shor et al. Oct 2002 A1
20020145911 Maayan et al. Oct 2002 A1
20020145918 Maayan et al. Oct 2002 A1
20020199065 Subramoney et al. Dec 2002 A1
20030001213 Lal Jan 2003 A1
20030002348 Chen et al. Jan 2003 A1
20030076710 Sofer et al. Apr 2003 A1
20030134476 Rolzin et al. Jul 2003 A1
20030145176 Dvir et al. Jul 2003 A1
20030145188 Cohen et al. Jul 2003 A1
20030155659 Verma et al. Aug 2003 A1
20030190786 Ramsbey et al. Oct 2003 A1
20030197221 Shinozaki et al. Oct 2003 A1
20030202411 Yamada Oct 2003 A1
20030206435 Takahashi Nov 2003 A1
20030209767 Takahashi et al. Nov 2003 A1
20030218207 Hashimoto et al. Nov 2003 A1
20030218913 Le et al. Nov 2003 A1
20030222303 Fukuda et al. Dec 2003 A1
20030227796 Miki et al. Dec 2003 A1
20040014290 Yang et al. Jan 2004 A1
20040021172 Zheng et al. Feb 2004 A1
20040027858 Takahashi et al. Feb 2004 A1
20040047198 Lusky et al. Mar 2004 A1
20040151034 Shor et al. Aug 2004 A1
20040153621 Polansky et al. Aug 2004 A1
20040157393 Hwang Aug 2004 A1
20040222437 Avni et al. Nov 2004 A1
20050117395 Maayan et al. Jun 2005 A1
20050140405 Do et al. Jun 2005 A1
20060084219 Lusky et al. Apr 2006 A1
Foreign Referenced Citations (41)
Number Date Country
0 822 557 Feb 1998 EP
0 843 398 May 1998 EP
0580467 Sep 1998 EP
0461764 Jul 2000 EP
1 091 418 Apr 2001 EP
1 207 552 May 2002 EP
1 223 586 Jul 2002 EP
1 365 452 Nov 2003 EP
001217744 Mar 2004 EP
54-053929 Apr 1979 JP
60-200566 Oct 1985 JP
60201594 Oct 1985 JP
63-249375 Oct 1988 JP
3-285358 Dec 1991 JP
04-226071 Aug 1992 JP
05-326893 Dec 1993 JP
06151833 May 1994 JP
06-232416 Aug 1994 JP
09-017981 Jan 1997 JP
11-162182 Jun 1999 JP
11-354758 Dec 1999 JP
2001-085646 Mar 2001 JP
2001-118392 Apr 2001 JP
2001-156189 Jun 2001 JP
3358663 Oct 2002 JP
WO 8100790 Mar 1981 WO
WO 9615553 May 1996 WO
WO 9625741 Aug 1996 WO
WO 9803977 Jan 1998 WO
WO 9931670 Jun 1999 WO
WO 9957728 Nov 1999 WO
WO 0165566 Sep 2001 WO
WO 0165567 Sep 2001 WO
WO 0243073 May 2002 WO
WO 03032393 Apr 2003 WO
WO 03054964 Jul 2003 WO
WO 03079370 Sep 2003 WO
WO 03079446 Sep 2003 WO
WO 03083916 Oct 2003 WO
WO 03088258 Oct 2003 WO
WO 03088353 Oct 2003 WO
Related Publications (1)
Number Date Country
20060268621 A1 Nov 2006 US
Divisions (2)
Number Date Country
Parent 10454820 Jun 2003 US
Child 11413962 US
Parent 09827757 Apr 2001 US
Child 10454820 US