Apparatus and method for improving sensing margin of electrically programmable fuses

The present invention relates generally to semiconductor devices and, more particularly, to an apparatus and method for improving sensing margin of electrically programmable fuse type devices in semiconductor devices.

Electrically programmable fuses (eFUSEs) are widely used to implement memory redundancy functionality in dynamic random access memory (DRAM), static random access memory (SRAM) and embedded memory devices. Programmable fuses can also be utilized in applications such as electronic chip identification, product feature personalization, and thermal diode calibration, to name a few.

In a redundancy application, for example, chips returned from fabrication are usually tested and a certain percentage of them are usually found to be bad (e.g., from random contingencies that may occur in the fabrication process). The percentage of good, usable chips is commonly referred to as the “yield.” Accordingly, redundancy may enable programming a chip at the testing phase so that flawed portions of the chip are not used in favor of unflawed redundant portions, thereby increasing the yield and decreasing the percentage of chips that must be thrown away as unusable. Many of the applications for electronic fuses combine the electronic fuse with the use of a memory in some way, for example, as part of a memory cell.

Additionally, in electronic chip identification or part number identification, an eFUSE may be used to “program” or write information into a non-volatile memory, which may be comprised of individual memory cells, and the information can then be read from the memory as a unique chip identifier or part number identifier. Furthermore, for thermal diode calibration, a non-volatile memory programmed using electronic fuses can be used to hold a test temperature and the thermal diode voltage read at the test temperature.

Regardless of the specific application environment, an eFUSE is typically programmed by passing a sufficient current through the structure such that its resistance is significantly altered from its initially fabricated state. In order to determine whether a particular fuse has been programmed or not, a sense circuit may be used to detect one of two possible “states” of the fuse. More specifically, the sense circuit holds one of two latched values therein, which is driven by a comparison between an applied voltage (by the sense circuit) across the fuse and a reference voltage generated within the sense circuit. The reference voltage is designed to be between a fuse voltage corresponding to the programmed state and a fuse voltage corresponding to the unprogrammed state.

However, as is typically the case with semiconductor circuit devices, the devices that make up the sense circuit are susceptible to parameter changes (e.g., voltage threshold changes in the case of field effect transistors) during the lifetime of the product in the field. Furthermore, as devices continually move to smaller geometries and lower operating voltages, the structural differences between programmed and unprogrammed fuses become less, thus making it more difficult to sense the state of the fuse. In order to compensate for such variability, sense circuits have been designed using statistical simulations of the expected operating conditions, including power supply voltage, temperature, manufacturing process variations, and fuse resistance. The circuit devices are then adjusted as needed to achieve a sufficient sense margin (V FUSE −V REF ) so as to tolerate the simulated variability.

Alternatively and/or additionally, fuse resistances are measured at the time of fuse programming (e.g., at wafer final test (WFT)) to assure that sufficient margin exists to successfully sense both intact and programmed fuses during the lifetime of the product. Unfortunately, the time required for implementing comprehensive fuse resistance measurement can be prohibitively long at WFT, particularly in chip designs employing thousands of fuses. It is desirable to enable a different sense margin at WFT with a digital circuit to quickly detect weak fuse cells. Accordingly, there remains a need for improving the sensing margin of electrically programmable fuses in semiconductor devices.

The foregoing discussed drawbacks and deficiencies of the prior art are overcome or alleviated by an apparatus for sensing the state of a programmable resistive memory element device. In an exemplary embodiment, a latch device is coupled to a fuse node and a reference node, the fuse node included within a fuse leg and the reference node configured within a reference resistance leg. The latch device is configured to detect a differential signal developed between the reference node and the fuse node as the result of sense current passed through the fuse leg and the reference resistance leg. The fuse and reference resistance legs are further configured for first and second sensing modes, wherein the second sensing mode utilizes a different level of current than the first sensing mode.

In another embodiment, a method for boosting the sense margin of a sense circuit for a programmable resistive memory element device includes coupling a latch device to a fuse node and a reference node, the fuse node included within a fuse leg and the reference node configured within a reference resistance leg. The latch device is configured to detect a differential signal developed between the reference node and the fuse node as the result of sense current passed through the fuse leg and said reference resistance leg. The fuse and reference resistance legs are configured for first and second sensing modes, wherein the second sensing mode utilizes a different level of current than the first sensing mode.

In still another embodiment, a method for determining the state of a programmable resistive memory element of a semiconductor device includes implementing, during a wafer final test stage, a first test mode to determine the state of the programmable resistive memory element. The first test mode includes detecting a differential signal developed between a reference node and a fuse node of a test circuit as the result of sense current passed through a fuse leg and a reference resistance leg of the test circuit. A second test mode is enabled to determine, during the lifetime of the semiconductor device, the state of the programmable resistive memory element. The second test mode includes detecting a differential signal developed between the reference node and fuse node of the test circuit as the result of sense current passed through the fuse leg and reference resistance leg of the test circuit. The second test mode utilizes a different level of current than the first test mode.

Referring to the exemplary drawings wherein like elements are numbered alike in the several Figures:

FIG. 1 is a schematic diagram of a differential sense circuit for determining the state of a programmable fuse type device;

FIG. 2( a ) is a schematic diagram of a differential sense circuit for determining the state of a programmable fuse type device, adapted for increased sense margin, in accordance with an embodiment of the invention;

FIG. 2( b ) is a schematic diagram of a differential sense circuit for determining the state of a programmable fuse type device, adapted for increased sense margin, in accordance with an alternative embodiment of the invention;

FIG. 3 is a process flow diagram illustrating a method for testing electronic fuse type devices, with increased sensing margin used after testing during product lifetime, in accordance with a further embodiment of the invention;

FIG. 4 is a timing diagram illustrating the operation of the sense circuit of FIG. 2 and the flow diagram of FIG. 3 in a first mode of testing during product fabrication; and

FIG. 5 is a timing diagram illustrating the operation of the sense circuit of FIG. 2 and the flow diagram of FIG. 3 in a second mode during product lifetime.

Disclosed herein is an apparatus and method for improving sensing margin of electrically programmable fuse type devices in semiconductor devices. Briefly stated, a sensing circuit is configured so as to be able to provide at least two separate modes of fuse sensing in which a different level of current is utilized during the second mode of sensing. Optionally, further test modes with different levels of current may be included. In so doing, a first mode of sensing uses a different level of current such that the trip point resistance range (over simulated device parameter shifts) is changed during device final testing. By applying a higher current with respect to the first mode, a second mode provides more sense margin and ensures sensing operation in field even if the voltage threshold shifts during product lifetime. Accordingly, individual resistance measurements of fuse devices during wafer final testing may be avoided. It should be noted that although the exemplary embodiments described herein are presented in terms of electronic fuse devices, it will be appreciated that the same principles are equally applicable to other current mode sensed memory elements and programmable devices such as, but not limited to, antifuses, EPROMs, mask ROM and flash memory. The following exemplary embodiments apply a higher current in a second mode than that in a first mode for the sensing scheme discussed herein, however it should be appreciated that the opposite can also be implemented (i.e., the second mode applies a lower current than the first mode).

Referring initially to FIG. 1 , there is shown a schematic diagram of a differential sense circuit 100 suitable for determining the state of a programmable fuse type device, generally designated by “F.” As is shown, a CMOS latch 102 (transistors N 1 , P 1 , N 2 and P 2 ) is coupled to a fuse node 104 and a reference node 106 , and is configured to latch a data bit representative of whether the fuse F is in a high resistance state (e.g., programmed) or a low resistance state (e.g., unprogrammed), based on a comparison of the fuse's resistance and the resistance of reference resistance R. PFET P 0 and NFET N 0 , controlled by FSET_N and FSET_P, respectively, are used to selectively couple the latch 102 the power supply rails (e.g., V DD and ground).

In addition, PFETs P 3 and P 4 , located within the fuse and reference resistance legs, respectively, are controlled by signal PRCHG_N and used to precharge the fuse and reference nodes prior to fuse sensing, as well as to supply reference current through fuse F and resistor R during fuse sensing. During the precharge phase, the fuse node 104 isolated from the fuse F through NFET N 3 , while the reference node 106 is isolated from the reference resistor R through NFET N 4 . Finally, a pair of output buffers (CMOS) inverters P 5 /N 5 and P 6 /N 6 is included on the fuse and reference nodes to drive the sensed state of the fuse stored in latch 102 to the appropriate output circuitry (not shown).

Prior to sensing of the state of the fuse F, the power supply connections to the latch 102 are disengaged; i.e., PFET P 0 and NFET N 0 are in a nonconducting state. Control signal PRCHG_N is active (low) to activate P 3 and P 4 and precharge the fuse and reference nodes to V DD . Control signal SIGDEV_P is initially low, but then goes active high to couple the fuse F to the fuse node 104 and to couple the reference resistor R to the reference node 106 . Depending upon whether the fuse F is in a high resistance state or a low resistance state, the amount of current through the fuse path will be either lower or higher than the amount of current through the reference path. Consequently, the voltage of the fuse node 104 (V FUSE ) will be lower or higher than the voltage of the reference node (V REF ).

However, as indicated above, the simulated variability of the various sense circuit parameters demonstrates a shift in the resistance trip point range over the lifetime of the device. Because the sense circuit of FIG. 1 is only capable of fuse state sensing using a single nominal level of test current (i.e., has a single test mode), the prospect of sensing the correct state of a fuse during the product's lifetime becomes less certain, particularly where the trip point resistance increases with increased negative bias thermal instability (NBTI) of PFET devices, for example.

Therefore, in accordance with an embodiment of the invention, FIG. 2( a ) is a schematic diagram of a differential sense circuit 200 for determining the state of a programmable fuse type device, adapted for increased sense margin. The device shown therein depicts an exemplary modification that allows the sense circuit 200 to implement a second mode of fuse sensing, wherein a larger level of current is used with respect to a first mode of fuse sensing. In so doing, the effective trip point range of the sense circuits are at a lower resistance value with respect to the first mode, thereby facilitating successful fuse state sensing throughout the life of the product. In the embodiment shown, sense circuit 200 is provided with an additional pair of PFETs, P 7 and P 8 in parallel with fuse leg PFET P 3 and resistance leg PFET P 4 , respectively. The additional PFETs P 7 , P 8 are controlled by a separate control signal PRECHG 2 _N, which is activated in the second mode of sensing. Because the additional PFETs provide an extra source of current through each differential leg, the overall trip resistance of the sense circuit 200 is lower in the second mode of operation. Exemplary timing diagrams comparing the first and second modes of operation are discussed hereinafter.

FIG. 2( b ) illustrates an alternative embodiment of a differential sense circuit 220 for determining the state of a programmable fuse type device. In this embodiment, another pair of parallel PFETs P 9 , P 10 are added and controlled by a third control signal PRECHG 3 _N, which is also activated in an optional third mode of sensing. A third mode of sensing may be desirable, for example, to enable the determination of a different sense margin to detect a weak fuse element. Still additional PFET devices can be included for even more modes of sensing, wherein each mode provides a different level of test current than the other modes. Alternatively, a third mode of sensing could also be implemented using the circuit embodiment of FIG. 2( a ), by activating PFETs P 7 and P 8 without activating P 3 and P 4 . In this instance, the device strength of P 7 is different from that of P 3 , and the device strength of P 8 is different from that of P 4 . It will thus be appreciated that even further modes of sensing are possible by providing additional (but distinct) levels of current during fuse testing.

FIG. 3 is a flow diagram illustrating a method for testing electronic fuse type devices with nominal sense margin prior to operation in the field with increased sensing margin, in accordance with a further embodiment of the invention. As initially indicated in block 302 , the state of each fuse of a device is tested (prior to fuse programming) during wafer final testing. In this instance, the first mode (lower current) is used for such testing in order to determine that the fuses are intact. For purposes of description, the “first mode” of testing at lower current is used synonymously with the term “test mode.” Because the fuses are not programmed at this point, it is expected that each fuse should have the unprogrammed low resistance. If the fuses are not intact, as determined at decision block, 304 , then the wafer (or chip within the wafer) is rejected at this point. Assuming, however, that the fuses are intact, they are then programmed in accordance with the device particulars, as shown at block 308 .

Once again, the “test mode” of the applicable sense circuitry is used to determine the state of the programmed fuses, as shown in block 310 . If the sensed data does not match the intended programmed data, as determined at decision block 312 , then the device is rejected as shown at block 314 . However, if the sensed fuse states using the test mode are correct, then “mission” (second) mode sensing is enabled thereafter for use during the product lifetime, as shown in block 316 . This may be carried out in a number of ways such as, for example, preventing the test mode (lower current) from being used again following completion of final testing, providing the customer instructions to wire the device for mission mode (higher current) sensing, or by programming another fuse to force mission mode sensing. Other enablement techniques, however, are also possible.

Finally, FIGS. 4 and 5 are timing diagrams illustrating the operation of the sense circuit 200 of FIG. 2 and the flow diagram of FIG. 3 in the first (test) and second (mission) modes of operation, respectively. Beginning with FIG. 4 , it will be seen that prior to fuse sensing, control signal SIDDEV_P is low, therefore isolating the fuse F and reference resistor R. The power connections P 0 , N 0 , to the latch 102 are also disabled through control signals FSET_N and FSET_P prior to sensing in the test mode. However, in this precharge phase of the test mode, it will further be seen that control signal PRECHG_N is active to hold nodes 104 and 106 at the supply rail voltage, V DD . As the circuit 200 of FIG. 2 also has the second “mission” mode capability, which is not utilized during wafer final test, the control signal PRCHG 2 _N is constantly maintained at a deactivated high state to keep PFETs P 7 and P 8 off.

Continuing with FIG. 4 , during a signal develop portion (t S ) of the test mode sense cycle, control signal SIGDEV_P has activated by transitioning from low to high. This couples the fuse F and reference resistor R to the precharged nodes 104 , 106 , respectively, thereby causing current to flow through each leg. Depending on the state of the fuse, the current flowing through the fuse leg of the differential sense circuit 200 should either be less than or greater than the current flowing through the reference leg. As a result, a differential voltage will be developed between V FUSE and V REF (nodes 104 and 106 ). Then, power to the latch 102 is provided as indicated by the activation of control signals FSET_N and FSET_P at the end of the signal develop portion (t S ) of the test mode cycle.

The next portion of the test mode cycle is depicted in FIG. 4 as a “hold” portion (t H ), in which the differential signal developed across the fuse and reference nodes in latched into the latch 102 . Optionally, an overlap portion (t OL ) maintains the PRECHG_N signal activated during the initial stage of the hold portion (t H ). Shortly thereafter, PRECHG_N is deactivated, allowing the latch 102 to fully drive V FUSE and V REF to the complementary logical values, again depending on the state of the fuse F. As this latched fuse state is realized, the data on the nodes begins to drive output buffers P 5 /N 5 and P 6 /N 6 . In addition, control signal SIGDEV_P is deactivated to decouple the fuse F and reference resistor R from the latch 102 . A final portion of the sensing stage prior to the sensed data being outputted and deemed valid is depicted as t DO in FIG. 4 .

In comparison, FIG. 5 illustrates the operation of the second “mission” mode of sensing during product lifetime. As can be seen, the operation is similar to the first “test” mode of sensing, except that the control PRECHG 2 _N (unlike PRECHG_N) is not active low the entire time during the precharge phase of the sense operation, but is instead activated for a brief time (t P ) just prior to the SIGDEV_P being activated. This prevents the parallel PFETs P 7 and P 8 from unnecessary degradation during product lifetime. However, P 7 and P 8 are preferably activated a sufficient time prior to coupling the fuse and reference resistor to the sensing nodes. Moreover, during this enhanced mode of current sensing, it is seen from FIG. 5 that PRCHG_N and PRCHG 2 _N are simultaneously deactivated after initial signal development.

While the invention has been described with reference to a preferred embodiment or embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted for elements thereof without departing from the scope of the invention. In addition, many modifications may be made to adapt a particular situation or material to the teachings of the invention without departing from the essential scope thereof. Therefore, it is intended that the invention not be limited to the particular embodiment disclosed as the best mode contemplated for carrying out this invention, but that the invention will include all embodiments falling within the scope of the appended claims.