[go: up one dir, main page]

CN101325084A - Method for providing dynamic voltage bias to interleaved memory array and its realization circuit - Google Patents

Method for providing dynamic voltage bias to interleaved memory array and its realization circuit Download PDF

Info

Publication number
CN101325084A
CN101325084A CNA2007100456423A CN200710045642A CN101325084A CN 101325084 A CN101325084 A CN 101325084A CN A2007100456423 A CNA2007100456423 A CN A2007100456423A CN 200710045642 A CN200710045642 A CN 200710045642A CN 101325084 A CN101325084 A CN 101325084A
Authority
CN
China
Prior art keywords
voltage
bit line
veq
dynamic
current
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CNA2007100456423A
Other languages
Chinese (zh)
Other versions
CN101325084B (en
Inventor
林殷茵
丁益青
刘欣
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fudan University
Original Assignee
Fudan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fudan University filed Critical Fudan University
Priority to CN2007100456423A priority Critical patent/CN101325084B/en
Publication of CN101325084A publication Critical patent/CN101325084A/en
Application granted granted Critical
Publication of CN101325084B publication Critical patent/CN101325084B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Semiconductor Memories (AREA)

Abstract

本发明属于集成电路技术领域,具体为对交叉型电阻转换型存储单元阵列提供动态偏置电压的方法。交叉型存储阵列依靠交叉的两条金属导线上施加的电信号进行选择操作。本发明设计了一种能够根据阵列的电阻分布情况以及外加操作信号的不同来动态产生对其他非操作的金属线对上施加的电压偏置,使高低两个阻态有较大比值的电阻转换型存储单元的交叉型阵列能够可靠工作。The invention belongs to the technical field of integrated circuits, in particular to a method for providing a dynamic bias voltage to an array of intersecting resistance conversion memory cells. The cross-type memory array relies on the electric signal applied to the two crossed metal wires to perform the selection operation. The present invention designs a resistance conversion that can dynamically generate voltage bias applied to other non-operating metal wire pairs according to the resistance distribution of the array and the difference of the external operation signal, so that the high and low resistance states have a larger ratio Interleaved arrays of type memory cells work reliably.

Description

对交叉型存储阵列提供动态电压偏置的方法及其实现电路 Method for providing dynamic voltage bias to interleaved memory array and its realization circuit

技术领域 technical field

本发明属于大规模数字集成电路技术领域,具体为一种对交叉型阻性存储阵列提供动态电压偏置的方法及其实现电路。The invention belongs to the technical field of large-scale digital integrated circuits, and specifically relates to a method for providing dynamic voltage bias to a cross-type resistive storage array and a circuit for realizing the same.

背景技术 Background technique

存储器在半导体市场中占有重要的地位,仅DRAM(Dynamic RandomAccess Memory)和FLASH两种就占市场的15%,随着便携式电子设备的不断普及,不挥发存储器市场也越来越大。然而传统的不挥发存储器的尺寸已经接近其物理极限,有报道预测FLASH技术的极限在32nm左右,这就迫使人们寻找性能更为优越的下一代不挥发存储器。最近电阻转换型随机可读取存储器件(RRAM,resistive random access memory)因为其高密度、低功耗、低成本的特点引起高度关注,所使用的材料有相变材料[1]、掺杂的SrZrO3[2]、铁电材料PbZrTiO3[3]、铁磁材料Pr1-xCaxMnO3[4]、二元金属氧化物材料[5]、有机材料[6]等。Memory occupies an important position in the semiconductor market. Only DRAM (Dynamic Random Access Memory) and FLASH account for 15% of the market. With the continuous popularization of portable electronic devices, the non-volatile memory market is also growing. However, the size of traditional non-volatile memory is close to its physical limit. It is reported that the limit of FLASH technology is around 32nm, which forces people to look for the next-generation non-volatile memory with better performance. Recently, resistive random access memory (RRAM) has attracted high attention because of its high density, low power consumption, and low cost. The materials used include phase change materials [1], doped SrZrO3[2], ferroelectric material PbZrTiO3[3], ferromagnetic material Pr1-xCaxMnO3[4], binary metal oxide material[5], organic material[6], etc.

目前电阻转换型存储器件主要采用的结构的1T1R结构,即一个选通器件(如二极管、三极管、场效应晶体管等)和一个存储电阻构成一个存储单元。这种存储单元具有结构简单,单元间干扰小,读取速度快,外围电路设计简单等优点,因而被广泛采用。图1示出了这种1T1R单元所构成的存储阵列。然而对于高密度的海量存储来说,1T1R的单元结构确又限制了存储密度的进一步提高,有以下两个原因:1)通常来说晶体管面积要大于存储电阻的面积,因而存储密度受到晶体管面积的限制;2)由于每个存储单元的选通晶体管都要消耗硅片的面积而无法将存储电阻进行层叠形成三维的存储阵列。At present, resistance switching memory devices mainly adopt a 1T1R structure, that is, a gate device (such as a diode, a triode, a field effect transistor, etc.) and a storage resistor form a memory unit. This kind of storage unit has the advantages of simple structure, small inter-unit interference, fast reading speed, simple peripheral circuit design, etc., so it is widely used. Figure 1 shows a memory array composed of such 1T1R cells. However, for high-density mass storage, the cell structure of 1T1R does limit the further improvement of storage density. There are two reasons: 1) Generally speaking, the transistor area is larger than the area of the storage resistor, so the storage density is limited by the transistor area. 2) Since the gate transistor of each memory cell will consume the area of the silicon wafer, the memory resistors cannot be stacked to form a three-dimensional memory array.

针对以上两个问题,一个比较理想的解决方案就是能够制造可靠的交叉型(cross-point)存储阵列。交叉型(cross-point)存储阵列是指存储单元中没有选通器件,存储电阻位于两条金属线交叉的地方,通过对两条相互交叉的金属线加不同的电压,对交叉处的存储电阻实施操作。如图2所示为交叉型的存储阵列图。For the above two problems, an ideal solution is to be able to manufacture a reliable cross-point storage array. The cross-point memory array means that there is no gating device in the memory cell, and the memory resistor is located at the intersection of two metal lines. By applying different voltages to the two intersecting metal lines, the memory resistor at the intersection Implement the action. As shown in Figure 2, it is a diagram of a cross-type storage array.

设计交叉型(cross-point)存储阵列的关键性问题就是能够通过外围电路的设计来减少漏电流的影响。因为对于交叉型的阻性存储阵列来说,实际是构成了一个电阻性的网络,这个网络有复杂的漏电流通路,因此会产生很大的干扰信号,影响存储阵列操作的功能正确性。The key issue in designing a cross-point memory array is to reduce the influence of leakage current through the design of peripheral circuits. Because for the cross-type resistive memory array, a resistive network is actually formed. This network has a complicated leakage current path, so a large interference signal will be generated, which will affect the functional correctness of the memory array operation.

对于交叉型存储阵列,为了减少漏电流的影响,通常的做法是提供箝位的偏置电压Veq[7],这种偏置电压的作用是使非选择的存储单元两端保持相等的电位,从而限制漏电流的通路。在传统的实现中,这个箝位的偏置电压一般采用一个固定的值,最常见的是V/2和V/3偏置方案[8],固定大小的偏置电压很难适应阵列的适配和所加操作信号的变化,所以无法保证交叉型存储阵列的可靠工作。For interleaved memory arrays, in order to reduce the impact of leakage current, the usual practice is to provide a clamped bias voltage Veq[7], which is used to keep the two ends of the non-selected memory cells at the same potential. Thereby limiting the path of leakage current. In the traditional implementation, the bias voltage of this clamp generally adopts a fixed value, the most common ones are V/2 and V/3 bias schemes [8], and it is difficult to adapt the fixed bias voltage to the array. Coordination and the change of the added operation signal, so the reliable operation of the interleaved memory array cannot be guaranteed.

本发明提出了对交叉型存储阵列提供动态的偏置电压,其特征在于电压偏置可以根据阵列中各个电阻所处的状态和外加的操作信号来进行调节,使偏置电压Veq取得最优值。这样的动态偏置方案可以保证交叉型存储阵列可靠的工作。The present invention proposes to provide a dynamic bias voltage for the interleaved memory array, which is characterized in that the voltage bias can be adjusted according to the state of each resistor in the array and the external operating signal, so that the bias voltage Veq can obtain an optimal value . Such a dynamic bias scheme can ensure reliable operation of the interleaved memory array.

发明内容 Contents of the invention

本发明的目的在于提出一种对交叉型存储阵列提供动态电压偏置的方法。这个动态的偏置电压能够根据阵列电阻分布的情况和外加操作信号进行动态调整,使之能够以最佳的大小来抑止漏电流,提高交叉型存储阵列工作的可靠性。同时,还提供这种动态电压偏置的实现电路。The object of the present invention is to propose a method for providing dynamic voltage bias to an interleaved memory array. This dynamic bias voltage can be dynamically adjusted according to the distribution of the array resistance and the external operating signal, so that it can suppress the leakage current with an optimal value and improve the reliability of the interleaved memory array. At the same time, a circuit for realizing the dynamic voltage bias is also provided.

本发明提出的对交叉型存储阵列提供动态电压偏置的方法,包括采用动态偏置的交叉型存储阵列和产生动态偏置电压的算法两个部分。其中,所述动态偏置的交叉型存储阵列的单元由电阻转换型存储介质构成。在这个阵列中,采用产生动态偏置电压的算法得到动态的偏置电压。算法通过操作时对流入存储阵列选择位线的电流进行采样,根据所采样的信号来产生动态的偏置。这个算法可以有效地寻找动态偏置的最佳值。具体步骤如下:The method for providing dynamic voltage bias to the interleaved memory array proposed by the present invention includes two parts: the dynamically biased interleaved memory array and the algorithm for generating the dynamic bias voltage. Wherein, the cells of the dynamically biased interleaved memory array are composed of resistance-switching memory media. In this array, a dynamic bias voltage is obtained by using an algorithm for generating a dynamic bias voltage. The algorithm samples the current flowing into the selection bit line of the memory array during operation, and generates a dynamic bias according to the sampled signal. This algorithm can efficiently find the optimal value of dynamic bias. Specific steps are as follows:

(1)先在存储阵列选择的位线上施加电压Vread,电压Vread大小为1V~5V;(1) First apply a voltage Vread on the bit line selected by the memory array, and the voltage Vread is 1V to 5V;

(2)在非选中字线和非选中位线上施加动态偏置电压Veq,Veq电压的初始值为零;(2) Apply a dynamic bias voltage Veq on the unselected word line and the unselected bit line, and the initial value of the Veq voltage is zero;

(3)采样流入所选择位线上的电流,如果该电流不为零,则将电压Veq加上一个小的步进值ΔVeq,重复第(2)步;如果该电流接近于零,则继续下一步;这里一般取ΔVeq为0.01V~0.05V;(3) Sample the current flowing into the selected bit line, if the current is not zero, add a small step value ΔVeq to the voltage Veq, and repeat step (2); if the current is close to zero, continue Next step; here, ΔVeq is generally taken as 0.01V ~ 0.05V;

(4)对电压Veq产生一个小的偏移量,施加到非选中字线和非选中位线;这个偏移量的大小一般为0.1V~0.2V;(4) Generate a small offset to the voltage Veq and apply it to the unselected word line and the unselected bit line; the size of this offset is generally 0.1V to 0.2V;

(5)根据流入位线上的电流来判断所选择的存储单元的状态。(5) The state of the selected memory cell is judged from the current flowing in the bit line.

本发明提出了一种设计交叉型电阻转换存储器的方案。这种方案可以应用于两态阻值相差很大的电阻转换存储单元。The invention proposes a scheme for designing a cross-type resistance switching memory. This scheme can be applied to resistance switching memory cells with two-state resistance values that differ greatly.

附图说明 Description of drawings

图1采用1T1R(一个选通器件和一个存储电阻)单元结构的存储阵列。Fig. 1 adopts the memory array of 1T1R (one gating device and one storage resistance) cell structure.

图2交叉型电阻转换存储阵列。Figure 2 Interleaved resistance switching memory array.

图3交叉型电阻转换存储阵列中的寄生漏电流通路。Fig. 3 The parasitic leakage current path in the interleaved resistive switching memory array.

图4交叉型电阻转换存储器的外围电路图。Figure 4 is the peripheral circuit diagram of the cross-type resistance switching memory.

图5交叉型电阻转换存储阵列的读操作箝位偏置方案。Fig. 5 The read operation clamping bias scheme of the interleaved resistance switching memory array.

图6交叉型电阻转换存储阵列的读操作实施图。Fig. 6 is an implementation diagram of a read operation of a cross-type resistance switching memory array.

图7交叉型电阻转换存储阵列读操作信号噪声仿真图。Fig. 7 Simulation diagram of the signal noise of the read operation of the cross-type resistance switching memory array.

图8交叉型电阻转换存储阵列的动态偏置产生算法。Fig. 8 The dynamic bias generation algorithm of the interleaved resistive switching memory array.

图9交叉型电阻转换存储阵列的动态偏置产生电路的逻辑框图。FIG. 9 is a logical block diagram of a dynamic bias generating circuit of a cross-type resistance switching memory array.

图10产生动态偏置电压的一种实施方案。Fig. 10 is an embodiment of generating dynamic bias voltage.

图11使用动态偏置方案的交叉型电阻转换存储阵列的读操作电流仿真。FIG. 11 is a simulation of the read operation current of an interleaved resistive switching memory array using a dynamic bias scheme.

具体实施方式 Detailed ways

下文结合图示及参考实施例更具体地描述本发明,本发明提供优选实施例,但不应该被认为仅限于在此阐述的实施例。The present invention will be described in more detail below with reference to the illustrations and examples. The present invention provides preferred embodiments, but should not be construed as being limited to the embodiments set forth herein.

在此参考图是本发明的理想化实施例的示意图,本发明所示的实施例不应该被认为仅限于图中所示的区域的特定形状。Where the referenced figures are schematic illustrations of idealized embodiments of the invention, the illustrated embodiments of the invention should not be construed as limited to the specific shapes of the regions shown in the figures.

应当理解,当称一个元件在“另一个元件上”或“在另一个元件上延伸”时,这个元件可以直接在“另一个元件上”或直接“在另一个元件上延伸”,或也可能存在插入元件。相反,当称一个元件“直接在另一个元件上”或“直接在另一个元件上延伸”时,不存在插入元件。当称一个元件与“另一个元件连接”或“与另一个元件耦接”时,这个元件可以直接连接或耦接到另一个元件,也可以存在插入元件。相反,当称一个元件直接与“另一个元件连接”或直接“与另一个元件耦接”时,不存在插入元件。It will be understood that when an element is referred to as being "on" or "extending over" another element, the element may be directly "on" or "extend" directly on the other element, or it may also be There is an insert element. In contrast, when an element is referred to as being "directly on" or "directly extending over" another element, there are no intervening elements present. When an element is referred to as being "connected" or "coupled" to another element, the element can be directly connected or coupled to the other element or intervening elements may also be present. In contrast, when an element is referred to as being directly "connected" or "directly coupled" to another element, there are no intervening elements present.

图1示出了采用1T1R单元结构的电阻转换存储阵列。存储单元由一个可编程的电阻120和1个选通器件100构成。选通器件100可以是如图所示的二极管,也可以是MOS管、双极性晶体管等具有开关作用的器件。在这种存储阵列中,存储单元之间通过选通器件可以相互隔离开来,从而减少了单元之间的干扰。例如在操作120时,只有选通器件100是导通的,其余的选通器件都是关断的,那么就只有存储电阻120才能够被外加的电信号所作用。Figure 1 shows a resistance switching memory array using a 1T1R cell structure. The memory cell is composed of a programmable resistor 120 and a gating device 100 . The gating device 100 may be a diode as shown in the figure, or a device with a switching function such as a MOS transistor or a bipolar transistor. In this memory array, memory cells can be isolated from each other by gate devices, thereby reducing interference between cells. For example, in the operation 120, only the gating device 100 is turned on, and the rest of the gating devices are turned off, so only the storage resistor 120 can be affected by the external electrical signal.

图2示出了交叉型(cross-point)的阻性转换存储单元阵列。存储单元120连接在金属导线200和210之间。金属导线(200~201)和金属导线(210~213)交叉的地方定位一个存储单元。以操作存储单元120为例,需要在金属导线200和210上施加不同的电压信号,两个电信号同时作用,就可以对其进行相应的操作。对于阵列中不需要操作的单元(如121),希望其不受电信号的影响,保存原来的状态。然而由于这些阻性的存储单元构成一个电阻性的网络,其中存在着一些寄生漏电流的通路。寄生漏电流会使存储器的性能严重下降,甚至产生错误的操作,可以表现为:1)误读出,因为漏电流的影响,使另灵敏发达器不能正确区分存储电阻的状态;2)误写入,漏电流对未选中的单元进行作用,使其状态发生不希望的翻转。FIG. 2 shows a cross-point resistive switching memory cell array. The memory cell 120 is connected between the metal wires 200 and 210 . A storage unit is positioned where the metal wires (200-201) intersect with the metal wires (210-213). Taking the operation of the storage unit 120 as an example, different voltage signals need to be applied to the metal wires 200 and 210 , and the two electrical signals act simultaneously to perform corresponding operations on it. For the units (such as 121 ) that do not need to be operated in the array, it is hoped that they will not be affected by the electrical signal and keep the original state. However, since these resistive memory cells form a resistive network, there are some paths for parasitic leakage currents. The parasitic leakage current will seriously degrade the performance of the memory, and even cause erroneous operation, which can be manifested as: 1) Misreading, because of the influence of leakage current, other sensitive devices cannot correctly distinguish the state of the storage resistor; 2) Miswriting In, the leakage current acts on unselected cells, causing them to flip their states undesirably.

图3示出了在交叉型的存储阵列中可能会存在的寄生漏电流通路。存储单元300是需要操作的单元,在相应的金属导线311上施加驱动电压330,同时将金属导线321接地。此时,存储单元300两端的电压大小等于驱动电压。然而在这个阻性网络中,存在着从311到321的寄生电流通路,如虚线350所示。这条寄生通路经过电阻300、导线320、电阻302、电阻303一直到通过导线321流向地。假设选择的存储电阻300为高阻,而不幸的是电阻301、302、303均为低阻,那么这个漏电流对操作电流的影响非常大,甚至会影响正常的工作。如果对300施加的是读操作,那么所得到的电流实际上是流过300的电流和漏电流之和,从而不能表示300的电阻状态。如果对300施加的是写操作,但是存储电阻301~302也会通过较大的电流或者电压信号,从而使它们的状态发生偏转。FIG. 3 shows possible parasitic leakage current paths in an interleaved memory array. The storage unit 300 is a unit that needs to be operated, and the driving voltage 330 is applied to the corresponding metal wire 311 while the metal wire 321 is grounded. At this time, the voltage across the memory cell 300 is equal to the driving voltage. However, in this resistive network, there is a parasitic current path from 311 to 321 as shown by dashed line 350 . This parasitic path passes through the resistor 300 , the wire 320 , the resistor 302 , and the resistor 303 until it flows to the ground through the wire 321 . Assuming that the selected storage resistor 300 is high-resistance, but unfortunately the resistors 301, 302, and 303 are all low-resistance, then this leakage current has a great influence on the operating current, and even affects normal operation. If a read operation is applied to 300 , the resulting current is actually the sum of the current flowing through 300 and the leakage current, so it cannot represent the resistance state of 300 . If the write operation is applied to 300 , the storage resistors 301 - 302 will also pass a relatively large current or voltage signal, thereby causing their states to be deflected.

图4示出了交叉型阵列存储器的系统电路图。450~453为构成存储阵列的电阻转换型存储单元,它们受到字线电路(400、401)和位线电路(420、421)的同时作用。其中字线电路(400、401)的作用是根据行地址输入。选择需要操作的行,并且对不需要操作的行提供偏置Veq(410、411)。位线电路(421、422)是根据列地址输入选择需要操作的列,根据数据输入选择需要加在选择位线上的电信号。灵敏放大器440用于读出所选择的电阻存储单元的电阻状态。由于在交叉型存储阵列中会有比较大的漏电流,所以灵敏放大器必须能够消除漏电流的影响。FIG. 4 shows a system circuit diagram of an interleaved array memory. 450-453 are resistance switching memory cells constituting the memory array, and they are simultaneously acted upon by the word line circuit (400, 401) and the bit line circuit (420, 421). The function of the word line circuit (400, 401) is to input according to the row address. Rows that need to be operated are selected, and offset Veq (410, 411) is provided to the rows that do not need to be operated. The bit line circuit (421, 422) selects the column to be operated according to the column address input, and selects the electric signal to be added to the selected bit line according to the data input. Sense amplifier 440 is used to sense the resistance state of the selected resistive memory cell. Since there will be a relatively large leakage current in the interleaved memory array, the sense amplifier must be able to eliminate the influence of the leakage current.

图5示出对于交叉型存储阵列读操作的偏置方案。VBLsel 500是对选择位线施加的电压偏置,VWLsel510是对选择字线的电压偏置,等于地电位。对非选择的位线施加的电压偏置是Veqb(520~524),对非选择的字线施加的电压偏置为Veqw(511~514)。从图中可知,选择操作的存储电阻上所加的电压信号等于Vblsel500,而对于在非选择位线和选中字线交叉点上的存储电阻两端的电压信号等于Veqb。阵列中的其他存储电阻两端的电压信号则等于(Veqb-Veqw)。电阻转换型存储单元的读操作,是通过施加一定的电压在存储单元上,然后区分其电流的大小,所以要尽量减少漏电流对信号电流的影响。理想的一种偏置方式是选择Veqw=Veqb=VBLsel。在这种情况下,理论上从选择位线上流过的电流等于流过选中的存储电阻的电流,在选中字线和非选中字线交叉处的存储电阻也有电流流过,但是不影响读出的信号电流。其他在非选中字线和非选中位线交叉处的存储电阻则没有电流流过,因为它们的两端被箝制在相等的电位上。然而由于在阵列中,寄生通路是随着各个存储单元的阻值变化的,并且由于开关管的参数的不匹配,以及位线电阻等原因,一个固定的Veq并不能很好的起到嵌位的作用,所以在本专利中提出了产生动态Veq的思想,使Veq可以根据电阻网络情况和施加的操作电压进行动态的自我调整,尽可能地减少漏电流的影响。Figure 5 shows a biasing scheme for a read operation of an interleaved memory array. VBLsel 500 is the voltage bias applied to the selected bit line, and VWLsel 510 is the voltage bias applied to the selected word line, which is equal to the ground potential. The voltage bias applied to the unselected bit lines is Veqb (520-524), and the voltage bias applied to the unselected word lines is Veqw (511-514). It can be seen from the figure that the voltage signal applied to the storage resistor in the selection operation is equal to Vblsel500, and the voltage signal at both ends of the storage resistor at the intersection of the non-selected bit line and the selected word line is equal to Veqb. The voltage signals across the other storage resistors in the array are then equal to (Veqb-Veqw). The read operation of the resistance conversion memory cell is to apply a certain voltage to the memory cell, and then distinguish the magnitude of its current, so the influence of the leakage current on the signal current should be minimized. An ideal biasing method is to select Veqw=Veqb=VBLsel. In this case, theoretically, the current flowing from the selected bit line is equal to the current flowing through the selected storage resistor, and the storage resistor at the intersection of the selected word line and the non-selected word line also has a current flowing, but it does not affect the readout. signal current. The other memory resistors at the intersection of the unselected word line and the unselected bit line have no current flow because their ends are clamped at equal potential. However, because in the array, the parasitic path changes with the resistance of each memory cell, and due to the mismatch of the parameters of the switch tube and the resistance of the bit line, a fixed Veq cannot be well clamped. Therefore, the idea of generating dynamic Veq is proposed in this patent, so that Veq can perform dynamic self-adjustment according to the resistance network conditions and the applied operating voltage, and reduce the influence of leakage current as much as possible.

图6示出了交叉型存储器的读出方案。600为箝位电压,它的作用是限制漏电流的大小。610为需要读出的存储电阻,它的两端连接在金属导线632和643上。通过金属导线632和643共同作用来进行选择。金属导线632上加一定的电压,而金属导线643接地,于是就形成了由632到643的通路。读出电流从金属导线632,经过存储电阻610,然后流到地。灵敏放大器通过感知流入632的电流,来判断选择的存储单元的电阻状态。在本方案中的开关(如620),可以用N型的MOS管代替,为了满足其具有较小的导通电阻,选择其具有较大的宽长比。Veq(600)是嵌位电压,它使未选通的字线(如640)和未选通的位线(如630)偏置在相等的电位。如果加在金属导线632上的电压偏置一定,那么流入它的电流是随着Veq变化的。FIG. 6 shows a readout scheme of an interleaved memory. 600 is the clamping voltage, and its function is to limit the magnitude of the leakage current. 610 is a storage resistor to be read, and its two ends are connected to metal wires 632 and 643 . The selection is made by the cooperation of metal wires 632 and 643 . A certain voltage is applied to the metal wire 632, and the metal wire 643 is grounded, so a path from 632 to 643 is formed. The read current flows from the metal wire 632, through the storage resistor 610, and then to ground. The sense amplifier senses the current flowing into 632 to determine the resistance state of the selected memory cell. In this solution, the switch (such as 620) can be replaced by an N-type MOS transistor, and in order to satisfy its smaller on-resistance, it is selected to have a larger width-to-length ratio. Veq (600) is the clamping voltage that biases the ungated word line (eg, 640) and the ungated bit line (eg, 630) at equal potentials. If the voltage bias applied to the metal wire 632 is constant, the current flowing into it varies with Veq.

图7示出了对于图6中256×16的交叉型阵列读操作的仿真结果。在仿真中,我们采用了这样的参数:低阻态电阻为2k,高阻态电阻为100k。由于低阻态的阻值较低,所以会有比较多的寄生漏电流支路。我们仿真了两种情况下的漏电流情况,分别是:1)漏电流最大的情况,即所有未选择的存储单元都处于低阻态;2)漏电流最小的情况,即所有未选择的存储单元都位于高阻态。在这两种情况下。我们又分别仿真了选择的单元是高阻态和第阻态的情况。曲线800和810表示了在情况1下,流入所选择位线和所选择存储电阻的电流随Veq的变化情况。曲线840表示了信号噪声比,即实际流过存储电阻的电流和漏电流之比。从图中可以得到这样的规律,信号噪声比随着Veq的增大而增大,在某一值时达到最大值。这个值接近所选择位线上的读电压。同时我们还发现,流入所选择位线的电流随着Veq的增大而减小,而信噪比最大的Veq的值很接近使流入所选择位线的电流为0时的Veq的值。下面再简单介绍一下其他几条曲线的含义。801、811,841对应的是情况1小所选择存储单元是低阻的情况。802,812,842对应的是情况2下所选择存储单元是高阻的情况,而803,813,843对应的是存储电阻为低阻的情况。FIG. 7 shows the simulation results for the read operation of the 256×16 interleaved array in FIG. 6 . In the simulation, we adopted such parameters: the low resistance state resistor is 2k, and the high resistance state resistance is 100k. Since the resistance value of the low-resistance state is relatively low, there will be more parasitic leakage current branches. We simulated the leakage current in two cases, which are: 1) the case of the largest leakage current, that is, all unselected memory cells are in a low-resistance state; 2) the case of the smallest leakage current, that is, all unselected memory cells cells are in a high-impedance state. In both cases. We also simulated the cases where the selected cells are in the high-impedance state and the second-resistance state. Curves 800 and 810 show the variation of the current flowing in the selected bit line and the selected storage resistor as a function of Veq in case 1. Curve 840 represents the signal-to-noise ratio, that is, the ratio of the current actually flowing through the storage resistor to the leakage current. It can be seen from the figure that the signal-to-noise ratio increases with the increase of Veq, and reaches the maximum value at a certain value. This value approximates the read voltage on the selected bit line. At the same time, we also found that the current flowing into the selected bit line decreases with the increase of Veq, and the value of Veq with the largest signal-to-noise ratio is very close to the value of Veq when the current flowing into the selected bit line is 0. Let's briefly introduce the meaning of the other curves. 801, 811, 841 correspond to case 1 where the selected memory cell is low resistance. 802, 812, 842 correspond to the case that the selected memory cell in case 2 is high resistance, and 803, 813, 843 correspond to the case that the storage resistor is low resistance.

图8示出了产生动态Veq的基本算法。根据图7的仿真结果,我们得出了这样的结论,如果能够提供适合的Veq,能够使信号噪声比达到较大的值,从而可以分辨出高阻和低阻的差别,而最小化漏电流的影响。算法描述如下:Figure 8 shows the basic algorithm for generating dynamic Veq. According to the simulation results in Figure 7, we have come to the conclusion that if a suitable Veq can be provided, the signal-to-noise ratio can reach a larger value, so that the difference between high resistance and low resistance can be distinguished, and the leakage current can be minimized Impact. The algorithm is described as follows:

(1)在存储阵列选择的位线上施加电压Vread,Vread大小位于1V~5V之间;(1) Apply a voltage Vread on the bit line selected by the memory array, and the size of Vread is between 1V and 5V;

(2)在非选中字线和非选中位线上施加动态偏置电压Veq,Veq电压的初始值为零;(2) Apply a dynamic bias voltage Veq on the unselected word line and the unselected bit line, and the initial value of the Veq voltage is zero;

(3)采样流入所选择位线上的电流,如果该电流不为零,则将电压Veq加上一个小的步进值ΔVeq,重复第二步;如果该电流接近于零,则继续下一步;这里一般取ΔVeq为0.01V~0.05V;(3) Sample the current flowing into the selected bit line, if the current is not zero, add a small step value ΔVeq to the voltage Veq, and repeat the second step; if the current is close to zero, continue to the next step ; Here, ΔVeq is generally taken as 0.01V ~ 0.05V;

(4)对电压Veq产生一个小的偏移量,施加到非选中字线和非选中位线;这个偏移量的大小一般为0.1V~0.2V;(4) Generate a small offset to the voltage Veq and apply it to the unselected word line and the unselected bit line; the size of this offset is generally 0.1V to 0.2V;

(5)根据流入位线上的电流来判断所选择的存储单元的状态。(5) The state of the selected memory cell is judged from the current flowing in the bit line.

图9示出了产生动态Veq的电路逻辑框图。一个读电压产生模块900的输出端同电流采样模块901相连,同时后者的输出连接到动态电压Veq产生模块902,通路选择开关S1或S2的一端与动态电压Veq产生模块的输出相连,另一端与电流采样模块901相连。通过下面我们对各个电路模块进行分析。模块900是读电压模块,产生的电压作用于选择的位线。读电压会使电流流入所选择的位线,电流采样模块901对这个电流信号进行采样。采样产生的信号作为模块902的输入,模块902根据这一信号产生偏置电压Veq。偏置电压Veq的变化又会影响流入所选择位线电流的大小。当这个环路达到平衡的时候,偏置电压Veq会使流入所选择位线的电流达到一个很小的值,此时我们称该偏置电压Veq是平衡电压Veq。当找到这个平衡点的时候,我们将偏置电压Veq进行一个小的偏移,然后再读出流入位线的电流值。FIG. 9 shows a logic block diagram of a circuit for generating dynamic Veq. The output terminal of a reading voltage generation module 900 is connected with the current sampling module 901, and the output of the latter is connected with the dynamic voltage Veq generation module 902 at the same time, and one end of the path selection switch S1 or S2 is connected with the output of the dynamic voltage Veq generation module, and the other end It is connected with the current sampling module 901. Through the following we analyze each circuit module. Block 900 is a read voltage block that generates a voltage that is applied to a selected bit line. Reading the voltage causes a current to flow into the selected bit line, and the current sampling module 901 samples this current signal. The signal generated by the sampling is used as the input of the module 902, and the module 902 generates the bias voltage Veq according to this signal. Variations in the bias voltage Veq in turn affect the amount of current flowing into the selected bit line. When the loop is balanced, the bias voltage Veq will cause the current flowing into the selected bit line to reach a very small value. At this time, we call the bias voltage Veq the balanced voltage Veq. When this balance point is found, we shift the bias voltage Veq a small amount, and then read the current value flowing into the bit line.

图10示出了产生动态偏置的一个实施电路图。一个大小为几欧姆到几百欧姆(如5欧姆-500欧姆)的采样电阻1000串联在位线上,其一端与读电压产生模块相连,另一端与选择操作的位线相连。同时,采样电阻1000的两端分别连接到MOS管1040a和1040b的源端(或漏端),MOS管1040a和1040b的漏端(或源端)则连接到放大器1010的正反两个输入端,它们的栅极则连接到控制信号1020。放大器1010的输出连接到PMOS管1030的栅极,PMOS管的漏端同电阻1031相连,同时连接到三极管1032的基极和集电极,三极管1032的发射极连接到电阻1033。MOS管1041a的源端(或漏端)连接到三极管1032的集电极,MOS管1041b的源端(或漏端)连接到三极管1032的发射极。MOS管1041a和1041b的栅极连接到控制信号1021。其工作原理是:读电流通过Rsample1000进行采样,然后Rsample两端的电位通过S1和S2输入到Amp1,该两点的电荷存储在C1(1050_a),C2(1050_b)中,用于在S1(1040_a),S2(1040_b)g关断后保持放大器两个输入端的电压差。在进行偏置电压Veq平衡点寻找的过程中,S3导通。当达到平衡时,产生的Veq使读出电流趋向于0,因为当流过Rsample的电流越大,则放大器Amp1(1010)输出端电位越高,则流过R1的电流越小,那么输出电压Veq也就越小。反之流过Rsample的电流越大,则输出电压Veq就越高。因为放大器的放大倍数很大(100db),那么可以使流过Rsample的电流保持在一个很小的值。当控制信号Φ(1020)为低电平时,S1,S2,S3截至,S4导通,这时候Veq比平衡时候偏移0.1V~0.2V,这个偏移量可以用一个工作于深饱和区的三极管得到。此时动态偏置电压Veq通过S4选通输出。当这个偏置电压Veq施加到存储阵列的时候,读电流必然会增大,但是此时的信号噪声比为最佳值或者接近最佳值,因为可以正确读出选择存储单元的电阻状态。Figure 10 shows an implementation circuit diagram for generating dynamic bias. A sampling resistor 1000 with a size of several ohms to hundreds of ohms (such as 5 ohms-500 ohms) is connected in series on the bit line, one end of which is connected to the read voltage generating module, and the other end is connected to the bit line for selection operation. At the same time, the two ends of the sampling resistor 1000 are respectively connected to the source terminals (or drain terminals) of the MOS transistors 1040a and 1040b, and the drain terminals (or source terminals) of the MOS transistors 1040a and 1040b are connected to the positive and negative input terminals of the amplifier 1010 , and their gates are connected to the control signal 1020 . The output of the amplifier 1010 is connected to the gate of the PMOS transistor 1030 , the drain of the PMOS transistor is connected to the resistor 1031 , and is connected to the base and collector of the transistor 1032 , and the emitter of the transistor 1032 is connected to the resistor 1033 . The source terminal (or drain terminal) of the MOS transistor 1041 a is connected to the collector of the triode 1032 , and the source terminal (or drain terminal) of the MOS transistor 1041 b is connected to the emitter of the triode 1032 . The gates of the MOS transistors 1041 a and 1041 b are connected to the control signal 1021 . Its working principle is: the read current is sampled through Rsample1000, and then the potential at both ends of Rsample is input to Amp1 through S1 and S2, and the charges at these two points are stored in C1 (1050_a), C2 (1050_b) for use in S1 (1040_a) , S2(1040_b)g keeps the voltage difference between the two input terminals of the amplifier after it is turned off. During the process of finding the balance point of the bias voltage Veq, S3 is turned on. When the balance is reached, the generated Veq makes the read current tend to 0, because when the current flowing through Rsample is larger, the output potential of the amplifier Amp1 (1010) is higher, and the current flowing through R1 is smaller, so the output voltage Veq is also smaller. Conversely, the larger the current flowing through Rsample, the higher the output voltage Veq. Because the amplification factor of the amplifier is very large (100db), the current flowing through Rsample can be kept at a small value. When the control signal Φ(1020) is low level, S1, S2, S3 are turned off, and S4 is turned on. At this time, Veq is shifted by 0.1V~0.2V compared with the balanced state. The triode is obtained. At this time, the dynamic bias voltage Veq is gated and output by S4. When the bias voltage Veq is applied to the memory array, the read current will inevitably increase, but the signal-to-noise ratio at this time is the optimum value or close to the optimum value, because the resistance state of the selected memory cell can be read correctly.

图11示出了对实施电路的仿真结果。图11(a)示出了偏置电压Veq的变化,即由平衡点电平(1100)通过电平偏移得到读出操作偏置电平1101。图(b)示出了读出电流1120和流过存储单元的电流1121随Veq的变化情况。当Veq偏移前,两个电流都很小,趋于0。当Veq发生偏移,得到有效的读出电流。可以看到1120为5.3μA,而1121为0.7μA,信号噪声比为7,接近最佳情况,可以正确区分存储电阻的状态。Figure 11 shows the simulation results for the implemented circuit. Fig. 11(a) shows the variation of the bias voltage Veq, that is, the read operation bias level 1101 is obtained from the equilibrium point level (1100) through level shifting. Figure (b) shows the variation of the read current 1120 and the current 1121 flowing through the memory cell with Veq. Before Veq shifts, both currents are very small and tend to 0. When Veq shifts, an effective read current is obtained. It can be seen that 1120 is 5.3μA, while 1121 is 0.7μA, and the signal-to-noise ratio is 7, which is close to the optimal situation, and the state of the storage resistor can be correctly distinguished.

参考文献references

[1]J.Maimon,E.Spall,R.Quinn,S.Schnur,″Chalcogenide-based nonvolatile memory technology″,IEEEProceedings ofAerospace Conference,p.2289,2001.[1] J. Maimon, E. Spall, R. Quinn, S. Schnur, "Chalcogenide-based nonvolatile memory technology", IEEE Proceedings of Aerospace Conference, p.2289, 2001.

[2]C.Y.Liu,P.H.Wu,A.Wang,W.Y.Jang,J.C.Young,K.Y.Chiu,and T.Y Tseng,“Bistable resistiveswitching of a sputter-deposited Cr-doped SrZrO3 memory film”,IEEEEDL vol.26,p.351,2005.[2] C.Y.Liu, P.H.Wu, A.Wang, W.Y.Jang, J.C.Young, K.Y.Chiu, and T.Y Tseng, "Bistable resistive switching of a sputter-deposited Cr-doped SrZrO3 memory film", IEEEEDL vol.26, p.351 , 2005.

[3]J.R.Contreras,H.Kohlstedt,U.Pooppe,R.Waser,C.Buchal,and N.A.Pertsev,“Resistive switching inmetal-ferroelectric-metal junctions”,Appl.Phys.Lett.vol.83,p.4595,2003.[3] J.R.Contreras, H.Kohlstedt, U.Pooppe, R.Waser, C.Buchal, and N.A.Pertsev, "Resistive switching inmetal-ferroelectric-metal junctions", Appl.Phys.Lett.vol.83, p.4595 , 2003.

[4]A.Asamitsu,Y.Tomioka,H.Kuwahara,and Y.Tokura,“Current switching of resistive states inmagnetoresistive manganites”,Nature(London)vol.388,p.50,1997.[4] A.Asamitsu, Y.Tomioka, H.Kuwahara, and Y.Tokura, "Current switching of resistive states inmagnetoresistive manganites", Nature (London) vol.388, p.50, 1997.

[5]I.G.Baek,M.S.Lee,S.Seo,M.J.Lee,D.H.Seo,.S.Suh,J.C.Park,S.O.Park,H.S.Kim,I.K.Yoo,U-InChung,and J.T.Moon,“Highly scalable non-volatile resistive memory using simple binary oxide drivenby asymmetric unipolar voltage pulses”,IEDMTech.Dig.p.587(2004).[5] I.G.Baek, M.S.Lee, S.Seo, M.J.Lee, D.H.Seo, .S.Suh, J.C.Park, S.O.Park, H.S.Kim, I.K.Yoo, U-InChung, and J.T.Moon, “Highly scalable non-volatile resistive memory using simple binary oxide drivenby asymmetric unipolar voltage pulses”, IEDMTech.Dig.p.587(2004).

[6]L.P.Ma,J.Liu,and Y.Yang,“Organic electrical bistable devices and rewriteable memory cells”,Appl.Phys.Lett.vol.80,p.2997,2002;L.D.Bozano,B.W.Kean,V.R.Deline,J.R.Salem,and J.C.Scott,“Mechanism for bistability in organic memory elements”,Appl.Phys.Lett.vol.84,p.607,2004.[6] L.P.Ma, J.Liu, and Y.Yang, "Organic electrical bistable devices and rewriteable memory cells", Appl.Phys.Lett.vol.80, p.2997, 2002; L.D.Bozano, B.W.Kean, V.R.Deline , J.R.Salem, and J.C.Scott, "Mechanism for bistability in organic memory elements", Appl.Phys.Lett.vol.84, p.607, 2004.

[7]W Reohr,H Honigschmid,R Robertazzi ect.“Memories of tomorrow”,IEEE circuits and devicesmagazine[8755-3996]Reohr yr:2002 vol:18iss:5pg:17[7] W Reohr, H Honigschmid, R Robertazzi ect. "Memories of tomorrow", IEEE circuits and devices magazine [8755-3996] Reohr yr: 2002 vol: 18iss: 5pg: 17

[8]YC Chen,CF Chen,SH Chen,CT Chen,ect.“A New Thin Film,Cross-point Non-volatile memory usingthreshold switching properties of phase change Chalcognide”,SSIC 2004.Proceedings.Vol.1,p.685-690[8]YC Chen, CF Chen, SH Chen, CT Chen, ect."A New Thin Film, Cross-point Non-volatile memory usingthreshold switching properties of phase change Chalcognide", SSIC 2004.Proceedings.Vol.1, p. 685-690

Claims (4)

1、一种对交叉型存储阵列提供动态电压偏置的方法,其特征在于包括采用动态偏置的交叉型存储阵列,和产生动态偏置电压算法的两个部分;这个算法通过操作时对流入存储阵列选择位线的电流进行采样,根据所采样的信号来产生动态的偏置。1. A method for providing a dynamic voltage bias to an interleaved memory array, characterized in that it comprises an interleaved memory array adopting a dynamic bias, and two parts for generating a dynamic bias voltage algorithm; The storage array selects the current of the bit line for sampling, and generates a dynamic bias according to the sampled signal. 2、根据权利要求1所述的对交叉型存储阵列的动态电压偏置方法,其特征在于所述的产生动态偏置电压算法的具体步骤如下:2. The dynamic voltage bias method for interleaved memory array according to claim 1, characterized in that the specific steps of the algorithm for generating the dynamic bias voltage are as follows: (1)首先在存储阵列选择的位线上施加电压Vread,电压Vread大小为1V~5V;(1) First apply a voltage Vread on the bit line selected by the memory array, and the voltage Vread is 1V to 5V; (2)在非选中字线和非选中位线上施加动态偏置电压Veq,电压Veq的初始值为零;(2) Apply a dynamic bias voltage Veq on the unselected word line and the unselected bit line, and the initial value of the voltage Veq is zero; (3)采样流入所选择位线上的电流,如果该电流不为零,则将电压Veq加上一个小的步进值ΔVeq,重复第(2)步;如果该电流接近于零,则继续下一步;这里取ΔVeq为0.01V~0.05V;(3) Sample the current flowing into the selected bit line, if the current is not zero, add a small step value ΔVeq to the voltage Veq, and repeat step (2); if the current is close to zero, continue Next step; here take ΔVeq as 0.01V~0.05V; (4)对电压Veq产生一个小的偏移量,施加到非选中字线和非选中位线;这个偏移量的大小为0.1V~0.2V;(4) Generate a small offset to the voltage Veq and apply it to the unselected word line and the unselected bit line; the offset is 0.1V to 0.2V; (5)根据流入位线上的电流来判断所选择的存储单元的状态。(5) The state of the selected memory cell is judged from the current flowing in the bit line. 3、一种如权利要求1所述对交叉型存储阵列的动态电压偏置方法的实现电路,其特征在于由读电压产生模块、电流采样模块、动态电压产生Veq模和通路选择开关S1和S2组成;其中读电压产生模块输出端同电流采样模块相连,同时后者的输出连接到动态电压Veq产生模块,通路选择开关S1或S2的一端与动态电压Veq产生模块的输出相连,另一端与电流采样模块相连。3. A realization circuit of the dynamic voltage biasing method to the interleaved memory array as claimed in claim 1, characterized in that the Veq module and path selection switches S1 and S2 are generated by the read voltage generation module, the current sampling module, and the dynamic voltage The output terminal of the read voltage generation module is connected to the current sampling module, and the output of the latter is connected to the dynamic voltage Veq generation module. One end of the path selection switch S1 or S2 is connected to the output of the dynamic voltage Veq generation module, and the other end is connected to the current sampling module. connected to the sampling module. 4、根据权利要求3所述的实现电路,其特征在于采样电阻串联在位线上,其一端与读电压产生模块相连,另一端与选择操作的位线相连;同时,采样电阻的两端分别连接到两个MOS管的源端或漏端,两个MOS管的漏端或源端则连接到放大器的正反两个输入端,它们的栅极则连接到一个控制信号。放大器的输出连接到一个PMOS管的栅极,PMOS管的漏端同一个电阻相连,同时连接到一个三极管的基极和集电极,三极管的发射极连接到电阻,一个MOS管的源端或漏端连接到上述三极管的集电极,另一个MOS管的源端或漏端连接到上述三极管的发射极;上述两个MOS管的栅极连接到一个控制信号。4. The implementation circuit according to claim 3, characterized in that the sampling resistor is connected in series on the bit line, one end of which is connected to the read voltage generation module, and the other end is connected to the bit line for selection operation; meanwhile, the two ends of the sampling resistor are respectively Connect to the source or drain of two MOS transistors, the drain or source of the two MOS transistors are connected to the positive and negative input terminals of the amplifier, and their gates are connected to a control signal. The output of the amplifier is connected to the gate of a PMOS transistor, the drain of the PMOS transistor is connected to a resistor, and connected to the base and collector of a triode, the emitter of the triode is connected to the resistor, and the source or drain of a MOS transistor One end is connected to the collector of the triode, and the source or drain of the other MOS transistor is connected to the emitter of the triode; the gates of the two MOS transistors are connected to a control signal.
CN2007100456423A 2007-09-06 2007-09-06 Method for providing dynamic voltage off-set for decussation type memory array and implementing circuit thereof Expired - Fee Related CN101325084B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2007100456423A CN101325084B (en) 2007-09-06 2007-09-06 Method for providing dynamic voltage off-set for decussation type memory array and implementing circuit thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2007100456423A CN101325084B (en) 2007-09-06 2007-09-06 Method for providing dynamic voltage off-set for decussation type memory array and implementing circuit thereof

Publications (2)

Publication Number Publication Date
CN101325084A true CN101325084A (en) 2008-12-17
CN101325084B CN101325084B (en) 2011-05-04

Family

ID=40188572

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2007100456423A Expired - Fee Related CN101325084B (en) 2007-09-06 2007-09-06 Method for providing dynamic voltage off-set for decussation type memory array and implementing circuit thereof

Country Status (1)

Country Link
CN (1) CN101325084B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110648697A (en) * 2018-06-27 2020-01-03 台湾积体电路制造股份有限公司 Selection circuit, latch-up prevention circuit for memory storage system and method

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN2338914Y (en) * 1998-03-20 1999-09-15 戴俊中 Dynamic class A power amplifier
US6614309B1 (en) * 2002-02-21 2003-09-02 Ericsson Inc. Dynamic bias controller for power amplifier circuits
US7236017B2 (en) * 2005-01-05 2007-06-26 The Boeing Company High speed sample-and-hold circuit

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110648697A (en) * 2018-06-27 2020-01-03 台湾积体电路制造股份有限公司 Selection circuit, latch-up prevention circuit for memory storage system and method
US11145335B2 (en) 2018-06-27 2021-10-12 Taiwan Semiconductor Manufacturing Co., Ltd. Latch-up prevention circuit for memory storage system
CN110648697B (en) * 2018-06-27 2021-12-31 台湾积体电路制造股份有限公司 Selection circuit, latch-up prevention circuit for memory storage system and method

Also Published As

Publication number Publication date
CN101325084B (en) 2011-05-04

Similar Documents

Publication Publication Date Title
CN108092658B (en) Logic circuit operation method
Liang et al. Cross-point memory array without cell selectors—Device characteristics and data storage pattern dependencies
US8194433B2 (en) Method and apparatus for accessing a bidirectional memory
CN100409363C (en) Method for spanning memory array, executing equipotential reading out and eliminating leakage current and its system
Sheu et al. Fast-write resistive RAM (RRAM) for embedded applications
US8675384B2 (en) Circuit for concurrent read operation and method therefor
CN101872647B (en) One-time programming resistance random access memory cell, array, memory and method of operation thereof
US7646633B2 (en) Method for programming phase-change memory and method for reading date from the same
US20090079009A1 (en) Memory device, memory circuit and semiconductor integrated circuit having variable resistance
CN103858172B (en) Selection equipment for cross point memory structure
CN101763891A (en) Phase-change memory unit and operation method thereof
CN101042933B (en) Non-volatile SRAM cell, array, method of operation and application thereof
US7242174B1 (en) Digital sensing circuit
Park et al. Analysis of resistance variations and variance-aware read circuit for cross-point ReRAM
CN109817269B (en) Test circuit block, variable resistance memory device and method of forming same
CN101325084B (en) Method for providing dynamic voltage off-set for decussation type memory array and implementing circuit thereof
JP2011204287A (en) Storage device
CN102473448B (en) Non-volatile memory array with resistive sense element block erase and uni-directional write
CN112310145A (en) memory cell array
JP2012533176A (en) Nonvolatile memory array having resistance sensing elements for block erase and unidirectional writing
CN114388022A (en) Signal amplification in MRAM during read
JP2012230725A (en) Memory data readout circuit
CN108630252B (en) Detect the sensing amplifier of memory cell current conversion
EP3772064A1 (en) A memory device and a method for configuring a memory device
CN115312097A (en) Multi-bit resistive random access memory writing circuit, method and memory device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20110504

Termination date: 20160906