TWI378456B - Method and system for reducing program disturb in non-volatile storage - Google Patents
Method and system for reducing program disturb in non-volatile storage Download PDFInfo
- Publication number
- TWI378456B TWI378456B TW96135141A TW96135141A TWI378456B TW I378456 B TWI378456 B TW I378456B TW 96135141 A TW96135141 A TW 96135141A TW 96135141 A TW96135141 A TW 96135141A TW I378456 B TWI378456 B TW I378456B
- Authority
- TW
- Taiwan
- Prior art keywords
- volatile storage
- storage element
- voltage
- volatile
- adjacent
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C11/00—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C11/56—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using storage elements with more than two stable states represented by steps, e.g. of voltage, current, phase, frequency
- G11C11/5621—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using storage elements with more than two stable states represented by steps, e.g. of voltage, current, phase, frequency using charge storage in a floating gate
- G11C11/5628—Programming or writing circuits; Data input circuits
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C16/00—Erasable programmable read-only memories
- G11C16/02—Erasable programmable read-only memories electrically programmable
- G11C16/04—Erasable programmable read-only memories electrically programmable using variable threshold transistors, e.g. FAMOS
- G11C16/0483—Erasable programmable read-only memories electrically programmable using variable threshold transistors, e.g. FAMOS comprising cells having several storage transistors connected in series
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C16/00—Erasable programmable read-only memories
- G11C16/02—Erasable programmable read-only memories electrically programmable
- G11C16/06—Auxiliary circuits, e.g. for writing into memory
- G11C16/34—Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
- G11C16/3418—Disturbance prevention or evaluation; Refreshing of disturbed memory data
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C16/00—Erasable programmable read-only memories
- G11C16/02—Erasable programmable read-only memories electrically programmable
- G11C16/06—Auxiliary circuits, e.g. for writing into memory
- G11C16/34—Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
- G11C16/3418—Disturbance prevention or evaluation; Refreshing of disturbed memory data
- G11C16/3427—Circuits or methods to prevent or reduce disturbance of the state of a memory cell when neighbouring cells are read or written
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C16/00—Erasable programmable read-only memories
- G11C16/02—Erasable programmable read-only memories electrically programmable
- G11C16/06—Auxiliary circuits, e.g. for writing into memory
- G11C16/10—Programming or data input circuits
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C2211/00—Indexing scheme relating to digital stores characterized by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C2211/56—Indexing scheme relating to G11C11/56 and sub-groups for features not covered by these groups
- G11C2211/562—Multilevel memory programming aspects
- G11C2211/5621—Multilevel programming verification
Landscapes
- Engineering & Computer Science (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Computer Hardware Design (AREA)
- Read Only Memory (AREA)
- Non-Volatile Memory (AREA)
- Semiconductor Memories (AREA)
Description
九、發明說明: 【發明所屬之技術領域】 本發明係關於非揮發性健存之技術。 【先前技術】 半導體記憶體裝置已經越來越普遍地用於各種電子裝 置。例如’非揮發性半導體記憶體用於蜂巢式電話、數位 相機、個人數位助理、行動計算裝置、非行動計算裝置及 其他裝置。電可抹除可程式化唯讀記憶體(EEPR0M)及快 閃記憶體係最流行的非揮發性半導體記憶體。 許多類型的EEPROM與快閃記憶體利用一浮動閘極,其 位於一半導體基板内的一通道區上方並與之絕緣。該浮動 閘極位於源極與汲極區之間。一控制閘極係提供在該浮動 閘極上面並與之絕緣。電晶體之臨限電壓受在浮動閘極上 所保持之電荷數量控制。即,在電晶體接通之前必須向該 控制閘極施加以允許其源極與汲極之間傳導的最小電壓數 量係受浮動閘極上的電荷位準控制。 一種快閃記憶體系統之範例使用NAND結構,其包括串 聯配置多個電晶體,夾置於二個選擇閘極之間。串聯電晶 體及選擇閘極稱為NAND串。圖1係顯示一 NAND串之一俯 視圖。圖2係其一等效電路。圖i及2所示NAND串包括串聯 並夾置在一第一(或汲極)選擇閘極12〇與一第二(或源極)選 擇閘極122之間的四個電晶體1〇〇、1〇2、1〇4及1〇6。選擇 閘極120經由位元線接點126將NAND串連接至一位元線。 選擇閘極122將NAND串連接至源極線128。選擇閘極12〇係 I24908.doc 1378456 藉由施加適當電壓至選擇線SGD來控制。選擇閘極122係 藉由施加適當電壓至選擇線SGS來控制。各電晶體100、 102、104及106具有一控制閘極及一浮動閘極。例如,電 晶體100具有控制閘極100CG及浮動閘極100FG。電晶體 102包括控制閘極102CG及一浮動閘極102FG。電晶體104 包括控制閘極104CG及浮動閘極104FG。電晶體106包括一 控制閘極106CG及一浮動閘極106FG。控制閘極100CG係 連接至字線WL3,控制閘極102CG係連接至字線WL2,控 制閘極104CG係連接至字線WL1及控制閘極106CG係連接 至字線WL0。 應注意雖然圖1及2顯示NAND串内的四個記憶體單元, 不過使用四個電晶體僅作為一範例。一 NAND串可具有小 於四個記憶體單元或四個以上的記憶體單元。例如,某些 NAND串包括8個記憶體單元、1 6個記憶體單元、32個記憶 體單元等。本文中的論述不限於在一 NAND串内的任一特 定數量的記憶體單元。 一使用一 NAND結構之快閃記憶體系統之典型架構包括 數個NAND串。各NAND串係藉由其受一選擇線SGS控制之 源極選擇閘極而連接至源極線並藉由其受一選擇線SGD控 制之汲極選擇閘極而連接至其相關聯位元線。各位元線及 經由一位元線接點連接至該位元線連接之個別NAND串包 含該等記憶體單元陣列之行。多個NAND串共用位元線。 一般而言,該位元線在一垂直於該等字線之方向上在 NAND串之頂部上佈線,並與一或多個感測放大器連接。 124908.doc 1378456 該等字線(WL3、7 列。 WL1及WL0)包含記憶體陣列之 各記憶體單元可儲存 數位資料時(稱為二進制記憶體單二數位)°當儲存-位元 "〇,,… 成二個範圍,其係指派邏輯資料及 0。在-NAND型快閃 避科貝料1及 c愿體範例中,在技 ro- 後電麼臨限係負數並定義為邏輯” 體早- 向控制閉極施加〇伏特r 係負數並藉由 、a _ 衧來嘗试一讀取時,記憶體單元會接 通以4日不正在儲存通敍 制a _ —。當臨限電屢係正數並藉由向控 制閘極施加〇伏特來當分 ㈣嘗4 一讀取操作時,記憶體單元不會 接通,其指示儲存邏輯零。 > -己,體單兀還可儲存多個位準的資訊(稱為一多狀態 5己憶體單元)。在儲存多位康音祖夕降、口士 夕位早貢枓之情況中,可能臨限電 壓之範圍係劃分成資料位準的數目。例如,若儲存四位準 資訊,則指派四個臨限電壓範圍給資料值"η"、"ι〇·,、 ”01"及"00" »在-NAND型記憶體之範例中,在一抹除操 作後臨限電壓係負數並定義為"u,、正臨限電壓係用於 "10"、"01"及"00"之狀態。 NAND型快閃記憶體及其操作的相關範例係提供於以下 美國專利/專利申請案,其均以引用方式併入本文:美國 專利第5,570,3 15號、美國專利第5,774,397號、美國專利第 6,046,935號、美國專利第6,456,528號及美國專利公告案第 US 2003/0002348號》本文論述還可應用於除NAND外的其 124908.doc 1378456 他類型的快閃記憶體以及其他類型的非揮發性記憶體。 在程式化一快閃記憶體單元時,施加一程式化電壓至控 制閘極並將位元線接地。由於快閃記憶體單元之通道與浮 動閘極之間的電壓差’來自浮動閘極下面之通道區域之電 子被注入浮動閘極内。當電子在浮動閘極中累積時,浮動 閘極變成帶負電並且該記憶體單元之臨限電壓上升。為了 施加程式化電壓至程式化中單元之控制閘極,將程式化電 壓施加在適當字線上。該字線還連接至利用相同字線之其 他NAND串之各串内的一記憶體單元。當期望程式化在一 子線上的一單元而不程式化連接至相同字線的其他單元時 會出現一問題。因為將程式化電壓施加於連接至一字線連 接之全部記憶體單元,故可能無意間程式化相同字線上之 未選定記憶體單元(不需程式化之記憶體單元)。在選定字 線上未選定記憶體單元之意外程式化係稱為"程式干擾,,。 可運用若干技術來防止程式干擾^在一稱為"自行增壓" 之方法中’該等未選定NAND串係與該等對應位元線電性 隔離並在程式化期間將一傳遞電壓(例如7至1〇伏特,但不 限於此範圍)施加至該等未選定字線。該等未選定字線耦 合至該等未選定NAND串之通道區域,引起一電壓(例如, 6至10伏特)存在於該等未選定NAND串之通道内,藉此減 少程式干擾。自行增壓造成增壓電壓存在於通道内,其降 低橫跨穿隧氡化物之電壓差,因而減少程式干擾。應注意 增壓通道電壓可在很大程度上變化,因為增壓通道電壓取 決於傳遞電壓之值,也取決於記憶體單元之狀態,當 124908.doc 1378456 NAND串内所有記憶體單元處於抹除狀態時,增壓最具效 率(最面通道電壓)。 圖3及4描述使用自行增壓方法程式化並禁止的nAND 串。圖3描述一程式化中的NAND串。圖3之NAND串包括 八個記憶體單元 304、306、308、310、312、314、316 及IX. Description of the Invention: [Technical Field to Which the Invention Is Ascribed] The present invention relates to a technique for non-volatile storage. [Prior Art] Semiconductor memory devices have become more and more popular for use in various electronic devices. For example, 'non-volatile semiconductor memory is used in cellular phones, digital cameras, personal digital assistants, mobile computing devices, inactive computing devices, and other devices. It can erase the most popular non-volatile semiconductor memory of programmable read-only memory (EEPR0M) and flash memory system. Many types of EEPROM and flash memory utilize a floating gate that is over and insulated from a channel region within a semiconductor substrate. The floating gate is located between the source and drain regions. A control gate is provided over and insulated from the floating gate. The threshold voltage of the transistor is controlled by the amount of charge held on the floating gate. That is, the minimum amount of voltage that must be applied to the control gate to allow conduction between its source and drain before the transistor is turned on is controlled by the charge level on the floating gate. An example of a flash memory system uses a NAND structure that includes a plurality of transistors arranged in series and sandwiched between two select gates. The series transistor and the select gate are called NAND strings. Figure 1 shows a top view of a NAND string. Figure 2 is an equivalent circuit thereof. The NAND strings shown in Figures i and 2 include four transistors 1 in series and sandwiched between a first (or drain) select gate 12A and a second (or source) select gate 122. 1, 2, 1, 4, and 1. Gate 120 is selected to connect the NAND string to a bit line via bit line contact 126. Select gate 122 connects the NAND string to source line 128. Select gate 12 I I24908.doc 1378456 is controlled by applying an appropriate voltage to select line SGD. The selection gate 122 is controlled by applying an appropriate voltage to the selection line SGS. Each of the transistors 100, 102, 104 and 106 has a control gate and a floating gate. For example, the transistor 100 has a control gate 100CG and a floating gate 100FG. The transistor 102 includes a control gate 102CG and a floating gate 102FG. The transistor 104 includes a control gate 104CG and a floating gate 104FG. The transistor 106 includes a control gate 106CG and a floating gate 106FG. The control gate 100CG is connected to the word line WL3, the control gate 102CG is connected to the word line WL2, the control gate 104CG is connected to the word line WL1, and the control gate 106CG is connected to the word line WL0. It should be noted that although Figures 1 and 2 show four memory cells within a NAND string, the use of four transistors is only an example. A NAND string can have less than four memory cells or more than four memory cells. For example, some NAND strings include 8 memory cells, 16 memory cells, 32 memory cells, and the like. The discussion herein is not limited to any particular number of memory cells within a NAND string. A typical architecture for a flash memory system using a NAND structure includes several NAND strings. Each NAND string is connected to the source line by its source select gate controlled by a select line SGS and is connected to its associated bit line by its drain select gate controlled by a select line SGD. . Each of the bit lines and the individual NAND strings connected to the bit line via a bit line contact include the rows of the memory cell array. A plurality of NAND strings share a bit line. In general, the bit lines are routed on top of the NAND strings in a direction perpendicular to the word lines and are coupled to one or more sense amplifiers. 124908.doc 1378456 These word lines (WL3, 7 columns. WL1 and WL0) contain memory cells of the memory array that can store digital data (called binary memory single binary digits) ° when storage - bit " 〇,,... In two ranges, the system assigns logical data and 0. In the case of the -NAND type flash dodge 1 and c, in the case of the technique ro-, the negative is a negative number and is defined as a logic "body early" - applying a negative volt to the control closed-pole and using a _ 衧 When trying to read, the memory unit will be turned on for 4 days without storing the general a _ —. (4) Tasting 4 When the read operation is performed, the memory unit will not be turned on, and the indication stores the logic zero. > - The body unit can also store multiple levels of information (referred to as a multi-state 5 memory) Unit). In the case of storing a number of Kangyin Zuyue and Houshixiu early Gongga, the range of possible threshold voltages is divided into the number of data levels. For example, if four levels of information are stored, the assignment is made. The four threshold voltage ranges are given to the data values "η","ι〇·,, 01", and "00" »in the -NAND type memory example, the threshold voltage is negative after an erase operation And defined as "u, the positive threshold voltage is used for "10", "01" and "00"NAND-type flash memory and related examples of its operation are provided in the following U.S. patents/patent applications, which are incorporated herein by reference: U.S. Patent No. 5,570,315, U.S. Patent No. 5,774,397, U.S. Patent No. No. 6,046,935, U.S. Patent No. 6,456,528, and U.S. Patent Publication No. US 2003/0002348, the disclosure of which is hereby incorporated by reference in its entirety in its entirety, in its entirety, in its entirety, its its its body. When staging a flash memory cell, apply a stylized voltage to the control gate and ground the bit line. Since the voltage difference between the channel of the flash memory cell and the floating gate 'from the channel region under the floating gate is injected into the floating gate. When electrons accumulate in the floating gate, the floating gate becomes negatively charged and the threshold voltage of the memory cell rises. To apply a stylized voltage to the control gate of the stylized unit, a stylized voltage is applied to the appropriate word line. The word line is also connected to a memory cell within each string of other NAND strings that utilize the same word line. A problem arises when it is desired to program a unit on a sub-line without programmatically connecting to other units of the same word line. Because the stylized voltage is applied to all of the memory cells connected to a word line connection, unselected memory cells (without stylized memory cells) on the same word line may be inadvertently programmed. The unexpected stylization of unselected memory cells on the selected word line is called "program disturb,. A number of techniques can be employed to prevent program interference. In a method called "self-boosting", the unselected NAND strings are electrically isolated from the corresponding bit lines and pass a voltage during stylization. (eg, 7 to 1 volt, but not limited to this range) is applied to the unselected word lines. The unselected word lines are coupled to the channel regions of the unselected NAND strings, causing a voltage (e.g., 6 to 10 volts) to be present in the channels of the unselected NAND strings, thereby reducing program disturb. Self-pressurization causes a boost voltage to be present in the channel, which reduces the voltage difference across the tunneling compound, thereby reducing program disturb. It should be noted that the boost channel voltage can vary to a large extent because the boost channel voltage depends on the value of the transfer voltage and also on the state of the memory cell when all memory cells in the 124908.doc 1378456 NAND string are erased. In the state, the boost is most efficient (the most channel voltage). Figures 3 and 4 depict nAND strings that are programmed and disabled using a self-boosting method. Figure 3 depicts a stylized NAND string. The NAND string of Figure 3 includes eight memory cells 304, 306, 308, 310, 312, 314, 316 and
318。該等八個記憶體單元之各記憶體單元包括一浮動閘 極(FG)與一控制閘極(CG)。在該等浮動閘極之各浮動閘極 之間的係源極/沒極區330。在一些實施方案中,存在一 p 型基板(例如,矽)、在該基板内的—N井及在該\井内的一 p井(未全部描述以增加該等圖式可讀性)。應注意p井可包 3 —所謂通道植入,通常係一 p型植入,該植入決定或幫 助決定該等記憶體#元之臨限電壓及其他特性。該等源極 /汲極區330係形成於p井内的n+擴散區。318. Each of the memory cells of the eight memory cells includes a floating gate (FG) and a control gate (CG). A source/nomogram region 330 between the floating gates of the floating gates. In some embodiments, there is a p-type substrate (e.g., germanium), a -N well within the substrate, and a p-well within the well (not all described to increase the readability of the patterns). It should be noted that p-well can be used as a so-called channel implant, usually a p-type implant that determines or helps determine the threshold voltage and other characteristics of the memory #元. The source/drain regions 330 are formed in the n+ diffusion region in the p-well.
在該NAND串之一端的係一汲極側選擇閘極324。汲極選 擇閘極324經由位元線接點334而連接該nand串至對應位 '線在該NAND串之另一端的係一源極選擇閉極322。源 極選擇閘極322將該NAND串連接至共同源極線332。在程 式化期間,選擇用於程式化之記憶體單元在其相關聯字線 上接收一程式化電壓Vpgm。程式化電avpgm 一般可在以 至=伏特之間變化。在―具體實施例中,程式化電壓信號 係-組振幅隨各新脈衝而遞增的脈衝。將大約8伏特的一 ㈣電壓為-傳遞電壓)Vpass施加於未被選擇用於程 式化之記憶體單元的控制閘極。源極選擇閘極322係處於 一隔離狀㈣’在錢極(G)接收G伏特…低電 加 124908.doc 1378456 至共同源極線332 〇此低電壓可能係零伏特。然而,源極 電壓亦可能稍尚於零伏特,以提供源極側選擇閘極之更佳 隔離特性。將一電壓Vsgd施加至汲極側選擇閘極324,該 電壓一般在電源電壓Vdd(例如,2·5伏特)之範圍内。經由 對應位元線將零伏特施加至位元接點334以致能程式化選 定記憶體單元312。通道340在零伏特或靠近其。由於通道 與s己憶體單元312之浮動閘極之間的電壓差,電子藉由 Fowler-Nordheim穿隧法透過閘極氧化物(一般也稱為穿隧 氧化物)而穿隧至浮動閘極内。 圖4之NAND串描述禁止進行程式化之一NAND串。該 NAND串包括八個記憶體單元35〇、352、354、356、358、 360、362及364。該NAND串還包括經由位元線接點374將 NAND串連接至對應位元線之汲極選擇閘極366以及用於將 該NAND串連接至共同源極線332之源極選擇閘極刊^。在 該等浮動閘極堆疊之各浮動閘極堆疊之間的係源極/汲極 區370。圖4之NAND串具有施加至汲極選擇閘極366之閘極 的Vsgd、施加至源極侧選擇閘極368之閘極的〇伏特、及在 共同源極線332處的零伏特(或一略高電壓p位元線接點 374經由對應位元線接收電源電壓vdd,以便禁止程式化記 憶體單元358。 當施加vdd時,汲極選擇電晶體366最初將處於一傳導狀 態’因此’在NAND串下面的通道區域將被部分充電至一 較问電位(间於〇伏特’且一般等於或幾乎等於Vdd)。此充 € &係' $為預充電”。預充電將會在通道電位已到達 124908.doc -11· 1378456A gate 324 is selected on the one-drain side of one of the NAND strings. The drain select gate 324 connects the nand string to the corresponding bit by the bit line contact 334. The line selects the closed end 322 at the other end of the NAND string. A source select gate 322 connects the NAND string to a common source line 332. During programming, the memory cells selected for stylization receive a stylized voltage Vpgm on their associated word lines. The stylized electric avpgm can generally vary between up to volts. In a specific embodiment, the stylized voltage signal is a set of pulses whose amplitude is incremented with each new pulse. A (four) voltage of approximately 8 volts is a -transfer voltage) Vpass is applied to the control gate of the memory cell that is not selected for programming. The source select gate 322 is in an isolated state (four)' at the money pole (G) receiving G volts... low power plus 124908.doc 1378456 to the common source line 332 〇 This low voltage may be zero volts. However, the source voltage may also be slightly less than zero volts to provide better isolation characteristics of the source side select gate. A voltage Vsgd is applied to the drain side selection gate 324, which is typically in the range of the supply voltage Vdd (e.g., 2.5 volts). Zero volts is applied to bit contact 334 via the corresponding bit line to enable programmatic selection of memory cell 312. Channel 340 is at or near zero volts. Due to the voltage difference between the channel and the floating gate of the sigma cell 312, electrons tunnel through the gate oxide to the floating gate through a gate oxide (also commonly referred to as a tunneling oxide) by Fowler-Nordheim tunneling. Inside. The NAND string of Figure 4 describes the prohibition of stylizing one of the NAND strings. The NAND string includes eight memory cells 35, 352, 354, 356, 358, 360, 362, and 364. The NAND string also includes a drain select gate 366 that connects the NAND string to the corresponding bit line via the bit line contact 374 and a source select gate for connecting the NAND string to the common source line 332. . A source/drain region 370 between each of the floating gate stacks of the floating gate stacks. The NAND string of FIG. 4 has Vsgd applied to the gate of the drain select gate 366, volts applied to the gate of the source side select gate 368, and zero volts at the common source line 332 (or one) The slightly higher voltage p-bit line contact 374 receives the supply voltage vdd via the corresponding bit line to disable the stylized memory unit 358. When vdd is applied, the drain select transistor 366 will initially be in a conducting state 'so' The channel area underneath the NAND string will be partially charged to a potential (between volts and is generally equal to or nearly equal to Vdd). This charge is precharged. Precharge will be at the channel potential. Has arrived at 124908.doc -11· 1378456
Vdd或一較低電位(由Vsgd-Vt給出)時自動停止,其中%等 於汲極選擇閘極366之臨限電壓。一般而言,在預充電期 間,Vsgd係以一方式選擇使得Vsgd-Vt>Vdd,從而在 NAND串下面的通道區域可預充電至Vdd。在該通道已到 達該電位之後,選擇閘極電晶體係藉由將VSgd降低至一類 似於Vdd(例如2.5伏特)之值而不傳導或使得不傳導。隨 後’將該等電壓Vpass及Vpgm從零伏特升高至其個別最終 值(不必同時),且由於汲極側選擇閘極電晶體366處於一非 傳導狀態’故該通道電位將因為該等字線與通道區域之間 的電容性耦合而開始上升。此現象係稱為自行增壓。在圖 4之NAND串下面的通道區域係或多或少均勻地增壓至一增 壓電壓。由於記憶體單元358之浮動閘極與通道之間的電 壓差已減小’故禁止程式化。關於程式化NAND快閃記憶 體之更多資訊,包括自行增壓技術,可見於Lutze等人的 美國專利6,859,397 ’ "用於非揮發性記憶體之源極側自行 增壓技術",其全部内容以引用形式併入本文。應注意, 圖4顯示區域3 80’其包括在基板表面之一通道區域與在增 壓通道區域下面的一空乏層(由於通道增壓至一高電壓而 具有增加電場之一區域)。該通道區域存在於該等浮動閘 極/控制閘極堆叠之各堆疊下面且在該等源極/汲極區37〇之 間。 另一解決程式干擾之嘗試係抹除區域自行增壓 ("EASB")。EASB嘗試隔離先前程式化單元之通道與被禁 止記憶體單元之通道。在EASB方法中,選定NAND串之通 124908.doc 1378456 道區域分成兩個區域:可包含若干程式化(或抹除單元)記 憶體單元在選定字線之源極側的一區域以及其中各單元仍 處於抹除狀態或至少尚未處於最终程式化狀態在選定字線 之汲極側的一區域。該等兩個區域係藉由偏壓至一較低隔 離電壓(通常零伏特)之一字線來分離。由於此分離,可將 該等兩個區域增壓至不同電位。幾乎在所有情形中,將選 疋子線之汲極側區域將增壓至一高於源極側區域的電位。 由於最高增壓區域係具有抹除單元之區域,此增壓方法係 稱為抹除區域自行增壓》 另一增壓方案(稱為修訂抹除區域自行增壓(ReaSB))類 似於EASB ,除了在接收隔離電壓之字線與選定字線之間 的係一接收一中間電壓(在Vpass與隔離電壓之間)的字線。 儘管上述增壓方法已減小程式干擾,但其仍未排除該問 題。隨著記憶體裝置之比例縮放變得越來越迅速,程式干 擾之效應也變得越來越大。此外,需要緊密臨限電壓分佈 之多狀態快閃記憶體裝置可能經歷該等分佈之加寬。特定 吕之,靠近該等選擇閘極(尤其靠近源極選擇閘極)之字線 可能受到程式干擾影響。 可能發生在靠近源極選擇閘極之記憶體單元(例如,記 憶體單TC 350靠近圖4之源極選擇閘極368)的一效應係閘極 引發及極洩漏(GIDL),其還稱為帶間穿隧。GIDL引起在 禁止程式化NAND串下面的通道(增壓至一高電壓)時在源 極選擇閘極處產生電子^隨後,在較強橫向電場内將所產 生電子向靠近源極選擇閘極 < 記憶體單元之浮動間極加 124908.doc 1378456 速。該等電子之某些電子可獲得足夠能量以注入浮動閘極 下面的穿隧氧化物内或浮動閘極本身内,因而修改對應記 憶體單元之臨限電壓。 圖5顯示圖4之NAND串之一部分,具有在源極選擇閘極 之汲極與記憶體單元350之一部分通道的一放大。由於在 一程式化禁止操作期間增壓NAND串(例如正在程式化其他 NAND串時)’ 一高電壓係存在於該增壓nand串之通道區 域内。此高電壓還存在於源極選擇閘極368(一般偏壓在 0 V)與靠近源極選擇閘極368之記憶體單元35〇之間的接面 區域處。此偏壓條件可能引起產生電子電洞對,也稱為 GIDL。該荨電洞將到達p井區域3 84。該等電子將會移動 至增壓通道區域。一般而言,在該源極選擇閘極與靠近該 源極侧選擇閘極之記憶體單元之間的接面區域内有一橫向 電%,因為該接面(汲極/源極)之部分係空乏,由於在該等 記憶體單元下面的通道區域與在該選擇閘極下面的通道區 域之間存在較大電壓差異。該等電子可在電場中得到加速 並γ能獲得足夠能量以注入靠近該源極側選擇閘極之記憶 體單元之穿随氧化物内或可能甚至到達該記憶體單元之浮 動閘極。在兩情況中,該對應記憶體單元之臨限電壓將會 由於存在注入電子而變化’藉此在讀取靠近源極選擇間極 之圮憶體單元時遭受一錯誤風險。 :了減小該等GIDL之效應,可降低該等增壓電壓νρ 以減小在該禁止操作期間增壓的通道數量。然而,此可能 造成程式干擾’由於不充分的增磨(如上述)。因而選擇適 124908.doc 1378456 當電壓用於Vpass極為重要。 【發明内容】 本文說明用於減小程式干擾之技術,其包括施加不同增 麼電壓至未選定記憶體單元。 一具體實施例包括增壓一未選定非揮發性儲存元件群組 (或該未選定非揮發性儲存元件之至少一部分)並在增麼該 * 未選定非揮發性儲存元件群、組時施加一程式信號至一特定 #揮發性儲存元件。增壓未選定非揮發性儲存S件群組包 括該特定非揮發性儲存元件、由於用於該群組的一最後抹 除程序而尚未完成完整程式化的一組非揮發性儲存元件、 及其他非揮發性儲存元件。該增屢包括施加一或多個較高 增壓信號至該組非揮發性健存元件並施加一或多個不同增 屋信號至該等其他非揮發性儲存元件,其中該一或多個較 问增壓彳s號大於該一或多個不同增壓信號。 另具體實施例包括在一程式化操作期間施加增壓信號 • 1一未選定非揮發性儲存元件群組並在該程式化操作期間 施加-程式信號至一目標未選定非揮發性健存元件,使得 S *未選疋非揮發性健存元件不會由於該等增壓信號而 程式化。該未選定非揮發性錯存元件群組係在該目標未選 定非揮發性错存元件的一共同側。該群組之一相鄰非揮發 f生儲存7C件係靠近該目標未選定非揮發性儲存元件。該目 標未選疋非揮發性儲存元件與該未選定非揮發性儲存元件 独全部相互串聯。施加增壓信號至該群組包括施加-特 疋增I信號至該相鄰非揮發性儲存元件並施加一不同增壓 124908.doc -15· 1378456 信號至該群組之其他非揮發性儲存元件。該特定増壓信號 係高於該不同增虔信號。該群組之其他非揮發性储存元件 不會由於該群組之一最後抹除而受到程式化。 另一具體實施例包括使連接至—第—字線之非揮發㈣ 存元件受到局部程式化並使連接至一第二字線之非揮發性 儲存=件受到局部程式化。該第一字線係相對於一字線群 組而罪近該第二字線,該字線群組與一非揮發性儲存元件 =組相關聯’該非揮發性儲存元件群組包括連接至該第一 字線之該等非揮發性儲存元件與連接該第二字線之該等非 揮發性儲存元件。針對連接至該第一字線之該等非揮發性 儲存元件完成程式化,包括施加一程式化信號至該第一字 線、施加-第-傳遞信號至該第二字線並施加一或多個其 =傳遞信號至該字線群組之其他字線。針對連接至該第— 字線之該等非揮發性儲存元件完成程式化係在使連接至該 第二字線之該等非揮發性储存元件受到局部程式化之後執 行0 另一具體實施例包括施加一程式信號至一用於一非揮發 性儲存元件群組之選定字線、施加一第一傳遞信號至用於 該非揮發性儲存元件群組之一組未選定字線及施加_高 第傳遞心號之傳遞信號至一相鄰字線(相對於該選 疋字線)該相鄰字線係用於完成該非揮發性儲存元件群 組之其連接非揮發性儲存元件程式化之下一者。 某二範例性實施方案包含複數個非揮發性儲存元件與 該等非揮發性儲存①件通信的—組控㈣(例如,字線、 124908.doc •16· 1378456 位70線或其他控制線)、及一 麻± S理電路,包括一或多個電 壓提供電路,其與該等护制绐,圣户^ 电 寻控制線通偽以提供信號至該組非揮 發性儲存元件,詩執行本文所述之該等程序。 【實施方式】 一適用於實施本發明之記憶體系統之範例使用NAND快 =記憶體結構。“ ’還可制其他類型的非揮發性儲存 裝置。例如,還可與本發明一起使用一所謂tan〇s結構Vdd or a lower potential (given by Vsgd-Vt) automatically stops, with % equal to the threshold voltage of the drain select gate 366. In general, during pre-charge, Vsgd is selected in such a way that Vsgd-Vt > Vdd, so that the channel region under the NAND string can be precharged to Vdd. After the channel has reached this potential, the gate electro-optic system is selected to be non-conducting or not conducting by lowering VSgd to a value similar to Vdd (e.g., 2.5 volts). Then 'the voltages Vpass and Vpgm are raised from zero volts to their individual final values (not necessarily simultaneously), and since the drain side select gate transistor 366 is in a non-conducting state', the channel potential will be due to the words The capacitive coupling between the line and the channel region begins to rise. This phenomenon is called self-pressurization. The channel region below the NAND string of Figure 4 is more or less uniformly boosted to a boost voltage. Since the voltage difference between the floating gate and the channel of the memory cell 358 has been reduced, stylization is prohibited. More information on stylized NAND flash memory, including self-pressurization techniques, can be found in U.S. Patent 6,859,397 ' "Source-side self-boosting technology for non-volatile memory" The entire content is incorporated herein by reference. It should be noted that Figure 4 shows a region 3 80' which includes a channel region on the surface of the substrate and a depletion layer below the region of the voltage-increasing channel (having an area of increased electric field due to channel boosting to a high voltage). The channel region is present below each of the floating gate/control gate stacks and between the source/drain regions 37A. Another attempt to resolve program interference is to erase the area self-boosting ("EASB"). The EASB attempts to isolate the channel of the previously stylized unit from the channel of the disabled memory unit. In the EASB method, the selected NAND string pass 124908.doc 1378456 track area is divided into two areas: a region that can contain a number of stylized (or erase unit) memory cells on the source side of the selected word line and the cells therein An area that is still erased or at least not in the final stylized state on the drain side of the selected word line. The two regions are separated by biasing to a word line of a lower isolation voltage (typically zero volts). Due to this separation, the two regions can be boosted to different potentials. In almost all cases, the drain side region of the selected sub-wire will be boosted to a higher potential than the source side region. Since the highest boost zone has the area of the eraser unit, this boosting method is called the erase zone self-boosting. Another boosting scheme (called the revision erase zone self-boosting (ReaSB)) is similar to EASB. A word line that receives an intermediate voltage (between Vpass and the isolation voltage) is received by a line between the word line receiving the isolation voltage and the selected word line. Although the above boosting method has reduced program disturb, it has not ruled out the problem. As the scaling of memory devices becomes faster and faster, the effects of program interference become larger. In addition, multi-state flash memory devices that require tight threshold voltage distribution may experience widening of such distributions. Certainly, the word line close to the select gates (especially close to the source select gate) may be affected by program disturb. An effect system gate initiation and pole leakage (GIDL) that may occur in a memory cell near the source select gate (eg, the memory cell TC 350 is close to the source select gate 368 of FIG. 4), which is also referred to as Tunneling between belts. GIDL causes electrons to be generated at the source select gate when the channel under the stylized NAND string is disabled (boost to a high voltage). Subsequently, the generated electrons are brought closer to the source gate in a stronger transverse electric field. ; The floating position of the memory unit is extremely fast. 124908.doc 1378456 speed. Some of the electrons of the electrons are energized to implant into the tunneling oxide under the floating gate or within the floating gate itself, thereby modifying the threshold voltage of the corresponding memory cell. Figure 5 shows a portion of the NAND string of Figure 4 with an amplification of the gate of the source select gate and a portion of the channel of the memory cell 350. Since a NAND string is boosted during a program inhibit operation (e.g., while other NAND strings are being programmed), a high voltage system is present in the channel region of the boosted nand string. This high voltage is also present at the junction region between the source select gate 368 (typically at 0 V) and the memory cell 35A near the source select gate 368. This bias condition may cause an electron hole pair, also known as GIDL. The borehole will reach the p-well zone 3 84. These electrons will move to the boost channel area. Generally, there is a lateral power % in the junction region between the source selection gate and the memory cell adjacent to the source side selection gate, because part of the junction (drain/source) is Depletion, due to the large voltage difference between the channel region below the memory cells and the channel region below the select gate. The electrons are accelerated in the electric field and gamma is capable of obtaining sufficient energy to implant a floating gate within the oxide or possibly even to the memory cell of the memory cell adjacent the source side select gate. In both cases, the threshold voltage of the corresponding memory cell will vary due to the presence of injected electrons' thereby thereby experiencing a risk of error in reading the memory cell near the source select terminal. : To reduce the effects of these GIDLs, the boost voltages νρ can be reduced to reduce the number of channels that are boosted during the inhibit operation. However, this may cause program disturbances due to insufficient grinding (as described above). Therefore, the choice of 124908.doc 1378456 when the voltage is used for Vpass is extremely important. SUMMARY OF THE INVENTION Techniques for reducing program disturb are described herein that include applying different boost voltages to unselected memory cells. A specific embodiment includes boosting a group of unselected non-volatile storage elements (or at least a portion of the unselected non-volatile storage elements) and applying one when adding the non-selected non-volatile storage element groups, groups Program signals to a specific # volatile storage component. The boosted unselected non-volatile storage S-piece group includes the particular non-volatile storage element, a set of non-volatile storage elements that have not been fully programmed due to a final erase procedure for the group, and others Non-volatile storage element. The increment includes applying one or more higher boost signals to the set of non-volatile storage elements and applying one or more different increaser signals to the other non-volatile storage elements, wherein the one or more The boost 彳s number is greater than the one or more different boost signals. Another embodiment includes applying a boost signal during a stylizing operation. • an unselected group of non-volatile storage elements and applying a program signal to the target unselected non-volatile storage element during the stylizing operation, The S* unselected non-volatile storage elements are not programmed due to the boost signal. The unselected non-volatile memory element group is on a common side of the target non-volatile memory element. One of the cohort adjacent non-volatile storage 7C pieces is adjacent to the target unselected non-volatile storage element. The target unselected non-volatile storage element and the unselected non-volatile storage element are all in series with each other. Applying a boost signal to the group includes applying a special boost I signal to the adjacent non-volatile storage element and applying a different boost 124908.doc -15· 1378456 signal to other non-volatile storage elements of the group . The particular rolling signal is higher than the different boosting signal. Other non-volatile storage elements of the group will not be stylized due to the last erasure of one of the groups. Another embodiment includes locally staging the non-volatile (four) memory elements connected to the -first word lines and localizing the non-volatile storage = elements connected to a second word line. The first word line is adjacent to the second word line with respect to a word line group, the word line group being associated with a non-volatile storage element=group. The non-volatile storage element group includes a connection to the The non-volatile storage elements of the first word line and the non-volatile storage elements connected to the second word line. Stylizing the non-volatile storage elements connected to the first word line, including applying a stylized signal to the first word line, applying a -first pass signal to the second word line, and applying one or more It = pass signals to other word lines of the word line group. The staging of the non-volatile storage elements connected to the first word line is performed after the non-volatile storage elements connected to the second word line are partially programmed. Another embodiment includes Applying a program signal to a selected word line for a non-volatile storage element group, applying a first transfer signal to a group of unselected word lines for the non-volatile storage element group, and applying a _ high pass The signal of the heart is transmitted to an adjacent word line (relative to the selected word line). The adjacent word line is used to complete the non-volatile storage element group of the non-volatile storage element group. . A second exemplary embodiment includes a plurality of non-volatile storage elements in communication with the non-volatile storage ones (four) (eg, word lines, 124908.doc • 16·1378456 70 lines or other control lines) And a circuit, including one or more voltage supply circuits, which are connected to the protection device, the sacred control line to provide a signal to the group of non-volatile storage elements, The programs described. [Embodiment] An example of a memory system suitable for implementing the present invention uses a NAND fast = memory structure. Other types of non-volatile storage devices can also be made. For example, a so-called tan〇s structure can also be used with the present invention.
(由在m上的-TaN.Al2(VSiN siQ2堆疊層所組 成)朴’其基本上係使用-氮化物層(而非一浮動閉極)内捕獲 電荷之一記憶體單元。用於快閃EEPR〇M系統的另_類型 記憶體單元利用一非傳導介電材料來代替一傳導浮動閘 極’以採用-非揮發性方式來儲存電荷。此類單元係說明 於Chan等人的文早巾,標題為”真正的單電晶體氧化物氮 化物·氧化物EEPR0M裝置",IEEE電子裝置學報第胤_ 8卷’第3號’ 1987年3月’第93至95頁。由氧化石夕、氮化 矽及氧化矽("ONO")形成之一三層介電質係夹置與一傳導 控制閘極與在記憶體單元通道上的一半傳導基板表面之 門藉由將電子從單元通道注入氮化物内來程式化單元, 在氮化物中於一限制區内捕獲並儲存該等電子。此儲存電 荷接著採用一可偵測方式來改變一部分單元通道的臨限電 壓。藉由將熱電洞注入氮化物中來抹除單元。還參閱 Nozaki等人的"用於半導體碟片應用之具有記憶體 單元之1-Mb EEPR〇M”(IEEE固態電路期刊,第%卷,第4 號,1991年4月,第497至5〇1頁),其說明採用一分離閘極 124908.doc 1378456 組態之類似單元,其中一摻雜多晶梦閘極延伸於一部分記 憶體單疋通道上以形成一分離選擇電晶體。前述兩文章全 部内容係以引用方式併入本文。在WilHam D B卿η幻的 Ε.如醫所編輯之..非揮發性I導體記憶體技術"仰邱出 版社,1998年)的第12節中所提及的程式化技術還在該節 中說明可應用於介電電荷捕獲裝置。還可使用其他類型的 記憶體單元。 圖6說明依據一具體實施例之記憶體裝置396,其具有用 於平行讀取及程式化—頁記憶體單元之讀取/寫入電路。 記憶體裝置396可包括一或多個記憶體晶粒398。記憶體晶 粒398包括二維記憶體單元陣列4〇〇、控制電路41〇及讀取^ 寫入電路465。在一些具體實施例争,該記憶體單元陣列 可能係二維的。可藉由經由列解碼器43〇之字線以及經由 行解碼器460之位元線來定址記憶體陣列4〇〇。讀取/寫入 電路465包括多個感測區塊5〇〇並允許平行讀取或程式化一 記憶體單元頁。一控制器45〇可包括於與一或多個記憶體 晶粒398相同的記憶體裝置396(例如,一可移除儲存卡) 内。命令及資料係經由線420在主機與控制器450之間以及 經由線4 1 8在該控制器與一或多個記憶體晶粒398之間傳 送。 控制電路410與讀取/寫入電路465協作以在記憶體陣列 400上執行記憶體操作。控制電路4丨〇包括一狀態機412、 一晶片上位址解碼器414及一功率控制模組416。狀態機 412¼供記憶體操作之晶片級控制。晶片上位址解碼器 124908.doc .〇 1378456 提供由主機或一記憶體控制器使用的位址與由解碼器43〇 及460使用的硬體位址之間的一位址介面。功率控制模組 416控制在記憶體操作期間向該等字線及位元線供應的功 率及電壓。在一具體實施例中,功率控制模組包括一或多 個電壓提供電路,其可接收一基底電壓(例如,Vdd電源或 其他電壓)並產生本文所述之任一電壓。一電壓提供電路 之範例係一電荷幫浦。 在一些實施方案中’可組合圖6之該等組件之一些組 件。在各種設計中,除記憶體單元陣列4〇〇外,圖6之該等 組件之一或多個組件(單獨或組合)可視為一管理電路。例 如,一管理電路可包括控制電路41〇、狀態機412、解碼器 414/460、功率控制416、感測區塊5〇〇、讀取/寫入電路 465、控制器450等之任一者或一組合。 圖7說明圖6所示記憶體裝置396之另一配置。由該等各 種周邊電路存取記憶體陣列4〇〇係在該陣列之相對側上以 一對稱方式來實施’使得可將在各側上的存取線及電路之 密度減半。因此,將列解碼器分割成列解碼器43〇a及 43〇B並且將行解碼器分割成行解碼器460A及460B。同樣 地’將該等讀取/寫入電路分割成從底部連接至位元線的 讀取/寫入電路465A及從陣列400之頂部連接至位元線的讀 取/寫入電路465B。依此方式,該等讀取/寫入模組之密度 基本上減半。圖7之裝置還可包括一控制器,如上面關於 圖6之裝置所述。 圖8係一個別感測區塊5〇〇之一方塊圖,該區塊係劃分成 124908.doc 1378456 核心部分(稱為一感測模組480)與一共同部分490。在一 具體實施例中,存在用於各位元線的一分離感測模組48〇 與用於一組多個感測模組480之一共同部分49〇。在一範例 中,一感測區塊將包括一共同部分49〇與八個感測模組 480。在一群組内的各感測模組將經由一資料匯流排472與 相關聯共同部分進行通信。關於進一步細節,請參閱2〇〇4 年12月29日申請的美國專利申請案第11/〇26 536號標題 '•非揮發性記憶體及用於感測放大器集之共用處理方法", 其全部内容以引用方式併入本文。 感測模組480包含感測電路47〇,其決定在一連接位元線 中的一傳導電流是否高於或低於一預定臨限位準。感測模 組480還包括一位元線鎖存器482,其係用於在連接位元線 上設定一電壓條件。例如,鎖存於位元線鎖存器482内的 一預疋狀態將會造成將連接位元線拉至一指定程式化禁止 之狀態(例如,Vdd)。 共同部分490包括一處理器492、一組資料鎖存器494與 耦合於該板資料鎖存器494與資料匯流排420之間的I/C) 介面496。處理器492執行計算。例如,其功能之一係決定 儲存於感測記憶體單元内的資料並將所決定資料儲存於該 組資料鎖存器内。該組資料鎖存器494係用於儲存在一讀 取操作期間處理器492所決定之資料位元。其還用於儲存 在一程式化操作期間從該資料匯流排42〇匯入的資料位 疋。所匯入的資料位元表示試圖程式化於記憶體内的寫入 資料。I/O介面496在資料鎖存器494與資料匯流排42〇之間 124908.doc -20· 1378456 提供一介面。 在5賣取或感測期間,該系統之操作受狀態機412控制, 該狀I、機控制向定址單元供應不同的控制閘極電壓。隨著 其,步遍及對應於記憶體所支援之各種記憶體狀態的各種 預疋義控制閘極電壓,感測模組48〇將在該些電壓之一電 壓下跳脫並經由匯流排472從感測模組48〇向處理器492提 供一輸出。此時,處理器492藉由考量該感測模組之該(等) 跳脫事件以及關於經由輸入線493來自該狀態機之施加控 制閘極電壓的資訊來決定所產生的記憶體狀態。接著其針 對記憶體狀態來計算二進制編碼並將該等產生資料位元健 存於資料鎖存器494内。在該核心部分之另一具體實施例 中,位7C線鎖存器482服務於雙重任務,同時作為一用於 鎖存感測模組480之輸出之鎖存器並還作為上述之位元線 鎖存器。 預期一些實施方案將會包括多個處理器492。在一具體 實施例中,各處理器492將會包括一輸出線(在圖8中未描 述),使得該等輸出線之各輸出線係已線或(wired 〇R)連接 在一起。在一些具體實施例中,該等輸出線係在連接至線 或線之前反向。此組態實現在程式化驗証程序期間快速決 定程式化程序已完成之時間,因為接收線或的狀態機可決 定所有程式化中位元已到達所需位準之時間。例如,當各 位元已到達其所需位準時,會將用於該位元之一邏輯零發 送至該線或線(或反向一資料一當所有位元輪出一資料 〇(或一反向資料一)時,該狀態機便知道要終止該程式化程 124908.doc 1378456 序。在各處理器與八個感測模組通信之具體實施例中,因 此該狀態機需要對該線或線進行八次讀取,或可將邏輯添 加至處理器492以累積相關聯位元線之結果,使得該狀態 機僅需要讀取該線或線一次。(from -TaN.Al2 on m) (composed of a stacked layer of VSiN siQ2), which basically uses a memory cell within the nitride layer (rather than a floating closed-pole) for flashing. Another type of memory cell of the EEPR〇M system uses a non-conductive dielectric material instead of a conductive floating gate to store charge in a non-volatile manner. Such units are described in Chan et al. , titled "Real Single-Crystal Oxide Nitride Oxide EEPR0M Device", IEEE Transactions on Electronic Devices, 胤 _ 8 Volume 'No. 3 ' March 1987, pp. 93-95. , 矽 and 矽 (ONO") form a three-layer dielectric sandwich and a conductive control gate and a half-conducting substrate surface on the memory cell channel by means of electrons from the cell The channel is implanted into the nitride internal programming unit to capture and store the electrons in a restricted region in the nitride. The stored charge then uses a detectable manner to change the threshold voltage of a portion of the cell channel. Hole is injected into the nitride to erase the unit See also Nozaki et al., "1-Mb EEPR〇M with Memory Units for Semiconductor Disc Applications" (IEEE Solid State Circuits, Vol. I, No. 4, April 1991, 497-5) 〇1 page), which illustrates a similar unit configured with a split gate 124908.doc 1378456, in which a doped polysilicon dream gate extends over a portion of the memory monopole channel to form a separate select transistor. The entire content of the article is incorporated herein by reference. In Section 12 of WilHam DB η 幻 Ε 如. Edited by the Institute: Nonvolatile I Conductor Memory Technology " Yangqiu Press, 1998) The stylized techniques mentioned are also described in this section for application to dielectric charge trapping devices. Other types of memory cells can also be used. Figure 6 illustrates a memory device 396 having read/write circuits for parallel read and stylized-page memory cells in accordance with an embodiment. Memory device 396 can include one or more memory dies 398. The memory crystal 398 includes a two-dimensional memory cell array 4A, a control circuit 41A, and a read/write circuit 465. In some embodiments, the memory cell array may be two dimensional. The memory array 4 can be addressed by a word line via column decoder 43 and a bit line via row decoder 460. The read/write circuit 465 includes a plurality of sensing blocks 5 〇〇 and allows parallel reading or programming of a memory unit page. A controller 45A can be included in the same memory device 396 (e.g., a removable memory card) as the one or more memory dies 398. Commands and data are transferred between the host and controller 450 via line 420 and between the controller and one or more memory dies 398 via line 4 1 8 . Control circuit 410 cooperates with read/write circuit 465 to perform memory operations on memory array 400. The control circuit 4 includes a state machine 412, an on-chip address decoder 414, and a power control module 416. The state machine 4121⁄4 is used for wafer level control of memory operations. The on-chip address decoder 124908.doc .〇 1378456 provides an address interface between the address used by the host or a memory controller and the hardware address used by decoders 43 and 460. Power control module 416 controls the power and voltage supplied to the word lines and bit lines during memory operation. In one embodiment, the power control module includes one or more voltage supply circuits that can receive a substrate voltage (e.g., a Vdd power supply or other voltage) and generate any of the voltages described herein. An example of a voltage supply circuit is a charge pump. Some components of the components of Figure 6 may be combined in some embodiments. In various designs, one or more of the components (alone or in combination) of the components of Figure 6 can be considered a management circuit in addition to the memory cell array. For example, a management circuit can include any of control circuit 41, state machine 412, decoder 414/460, power control 416, sensing block 5, read/write circuit 465, controller 450, and the like. Or a combination. FIG. 7 illustrates another configuration of the memory device 396 of FIG. The peripheral memory access memory arrays 4 are implemented in a symmetrical manner on opposite sides of the array such that the density of access lines and circuitry on each side can be halved. Therefore, the column decoder is divided into column decoders 43a and 43B and the row decoder is divided into row decoders 460A and 460B. Similarly, the read/write circuits are divided into a read/write circuit 465A connected from the bottom to the bit line and a read/write circuit 465B connected from the top of the array 400 to the bit line. In this way, the density of the read/write modules is substantially halved. The apparatus of Figure 7 can also include a controller as described above with respect to the apparatus of Figure 6. Figure 8 is a block diagram of a different sensing block 5, which is divided into a core portion of 124908.doc 1378456 (referred to as a sensing module 480) and a common portion 490. In one embodiment, there is a separate sensing module 48 for each bit line and a common portion 49 for one of the plurality of sensing modules 480. In one example, a sensing block will include a common portion 49A and eight sensing modules 480. Each of the sensing modules within a group will communicate with an associated common portion via a data bus 472. For further details, please refer to the U.S. Patent Application Serial No. 11/26,536, filed on Dec. 29, 2008, entitled "Non-Volatile Memory and Common Processing Methods for Sensing Amplifier Sets", The entire content is incorporated herein by reference. Sensing module 480 includes sensing circuitry 47 that determines whether a conduction current in a connected bit line is above or below a predetermined threshold level. Sensing module 480 also includes a bit line latch 482 for setting a voltage condition on the connected bit line. For example, a pre-asserted state latched in bit line latch 482 will cause the connected bit line to be pulled to a specified stabilizing state (e.g., Vdd). The common portion 490 includes a processor 492, a set of data latches 494, and an I/C) interface 496 coupled between the board data latch 494 and the data bus 420. Processor 492 performs the calculations. For example, one of its functions is to determine the data stored in the sensing memory unit and store the determined data in the data latch of the group. The set of data latches 494 is used to store the data bits determined by the processor 492 during a read operation. It is also used to store data bits that are imported from the data bus 42 during a stylized operation. The data bits that are imported represent the write data that is attempting to be programmed into the memory. The I/O interface 496 provides an interface between the data latch 494 and the data bus 42 124 124908.doc -20· 1378456. During 5 sell or sense, the operation of the system is controlled by state machine 412, which supplies different control gate voltages to the address unit. Along with it, the various sensing threshold voltages corresponding to the various memory states supported by the memory, the sensing module 48A will trip at one of the voltages and pass through the bus 472. Sensing module 48 provides an output to processor 492. At this point, processor 492 determines the resulting memory state by taking into account the (equivalent) trip event of the sense module and information regarding the applied control gate voltage from the state machine via input line 493. It then calculates the binary code for the memory state and stores the generated data bits in the data latch 494. In another embodiment of the core portion, bit 7C line latch 482 serves a dual task while acting as a latch for latching the output of sensing module 480 and also as the bit line described above. Latches. It is contemplated that some embodiments will include multiple processors 492. In one embodiment, each processor 492 will include an output line (not depicted in Figure 8) such that the output lines of the output lines are wired or wired (R). In some embodiments, the output lines are reversed prior to being connected to the line or line. This configuration makes it quick to determine when the stylized program has completed during the stylized verification process because the receive line or state machine can determine when all stylized bits have reached the desired level. For example, when each element has reached its required level, a logical zero for one of the bits will be sent to the line or line (or reverse one data one when all bits turn out a data 〇 (or one inverse) In the case of data 1), the state machine knows to terminate the program sequence 124908.doc 1378456. In a specific embodiment where each processor communicates with eight sensing modules, the state machine needs to be on the line or The line is read eight times, or logic can be added to processor 492 to accumulate the result of the associated bit line such that the state machine only needs to read the line or line once.
在程式化或驗證期間,欲程式化資料係從資料匯流排 420而儲存於該組資料鎖存器494内的。受該狀態機控制的 程式化操作包含一系列程式化電壓脈衝(具有遞增振幅), 其係施加至該等定址記憶體單元之控制㈣。各程式化脈 衝之後跟隨一驗證程序以決定是$已將記憶體單元程式化 至所需狀。處理器492監控與所需記憶體狀態相關之驗 -a·»己隐體狀態。當該等二狀態—致時,處理器492設定位 元線鎖存器482,使得引起該位元線被拉至一減程式化 禁止之狀態。此舉禁止麵合至該位元線之單元受到進一步 程式化’即使在其控制閘極上出現程式化脈衝。在其他具 體實施例中,該處理器最初載人位元線鎖存器M2,而該 感測電路在該驗證程序期間將其設定為一禁止值。 資料鎖存器堆疊494包含對應於感測模組的-資料鎖存 器堆疊°在—具體實施例中,每個感龍組480存在三個 資料鎖存器。在—些實施方案(但不要求)中,該等資料鎖 存器係實施為-移位暫存器’使得將儲存於其内的平行資 料轉換成串列資料用於資料匯流排420,且反之亦然。在 八體實施例中,對應個記憶體單元之讀取/寫入區 =所有資料鎖存器可以係鏈結在—起以形成—區塊移位 暫存器’使得可藉由串列傳送來輸入或輸出一資料區塊。 124908.doc -22· 1378456 特定5之’ r個讀取/寫入模組庫係調適成使得#資料鎖存 器組之各資料鎖存器將資料依序移人/移出該資料匯流 排’如同其係—用於整個讀取/寫人區塊之移位暫存器之 部分。 關於非揮發性儲存裝置之各種具體實施例之結構及/或 操作之額外資訊可見於⑴2〇〇4年3月25日公告的美國專利 申請公告案第2004/0057287號,"非揮發性記憶體及減小 源極線偏壓錯誤之方法";(2) 2〇〇4年6月1〇日公告的美國 專利申請公告案第2004/0109357號,"非揮發性記憶體及 改良感測之方法|,;(3) 2004年12月16日申請的美國專利申 請案第11/015,199號’㈣為”改良記憶體感測電路及用於 低電壓操作之方法,,,發明人Raul_Adrian Cemea ; (4) 2()()5 年4月5日申請的美國專利申請案11/〇99,133,標題為"在非 揮發性記憶體之讀取操作期間的耦合補償",發明人】丨扣 Chen ;及(5) 2005年12月28曰申請的美國專利申請案第 11/321,953號,標題為”用於非揮發性記憶體之參考感測放 大器•,發明人Siu Lung Chan與Raul-Adrian Cernea。上面 直接所列專利文件之所有五個專利文件全部内容係以引用 形式併入本文。 圖9描述記憶體單元陣列4〇〇之一範例性結構。在一具體 實施例中,該記憶體單元陣列係劃分成大量記憶體單元區 塊。如同快閃EEPROM系統,區塊係抹除單位。即,各區 塊包含一起抹除的最小數目記憶體單元。各區塊一般劃分 成若干頁。一頁係一程式化單位。在一具體實施例中,個 124908.doc •23 · 別頁可劃分為多個片段且該等片 操作時一次寫入的最小數目單元 存於一記憶體單元之列内。一 -區段包括使用者資料與管理H管理資料—般包括從 區段之使用者資料已經計算的—錯誤校正碼(ECC)。控制 " 卩刀(如下述)在將資料程式化在陣列内時計算 默’並還在從該陣列讀取f料時檢查其。或者該等During stylization or verification, the data to be stylized is stored in the data latch 494 from the data bus 420. The stylized operation controlled by the state machine includes a series of stylized voltage pulses (with increasing amplitude) that are applied to the control of the addressed memory cells (4). Each stylized pulse is followed by a verification procedure to determine that the memory unit has been programmed to the desired state. The processor 492 monitors the status of the desired memory state - a · » has a hidden state. When the two states are met, the processor 492 sets the bit line latch 482 causing the bit line to be pulled to a reduced program disabled state. This prohibits the unit that is integrated into the bit line from being further stylized' even with stylized pulses on its control gate. In other embodiments, the processor initially carries a bit line latch M2, and the sensing circuit sets it to a disable value during the verification process. The data latch stack 494 includes a data latch stack corresponding to the sense module. In the particular embodiment, there are three data latches per sense group 480. In some embodiments (but not required), the data latches are implemented as a -shift register to cause parallel data stored therein to be converted into serial data for data bus 420, and vice versa. In the eight-body embodiment, the read/write area of the corresponding memory unit = all data latches can be linked to form a block shift register so that it can be transmitted by serial To input or output a data block. 124908.doc -22· 1378456 The specific r's read/write module library is adapted such that each data latch of the #data latch group sequentially shifts data out/out of the data bus. As with its system - part of the shift register for the entire read/write block. Additional information regarding the structure and/or operation of various embodiments of the non-volatile storage device can be found in (1) U.S. Patent Application Publication No. 2004/0057287, issued March 25, 2005, "non-volatile memory Method for reducing the bias of the source line bias"; (2) US Patent Application Bulletin No. 2004/0109357, published on June 1st, 2, 4, "Non-volatile memory and improvement Method of sensing|,; (3) U.S. Patent Application Serial No. 11/015,199, filed on Dec. 16, 2004, the disclosure of Raul_Adrian Cemea; (4) 2()() U.S. Patent Application Serial No. 11/99,133, filed on Apr. 5, 2005, entitled "Coupling Compensation During Non-Volatile Memory Read Operations" , inventor, 丨Chen Chen; and (5) U.S. Patent Application Serial No. 11/321,953, filed on December 28, 2005, entitled "Reference Sense Amplifier for Non-Volatile Memory", Invention People Siu Lung Chan and Raul-Adrian Cernea. All five patent documents of the patent documents listed above are hereby incorporated by reference in their entirety. Figure 9 depicts an exemplary structure of a memory cell array 4A. In a specific embodiment, the array of memory cells is divided into a plurality of memory cell blocks. Like the flash EEPROM system, the block erases the unit. That is, each block contains a minimum number of memory cells that are erased together. Each block is generally divided into several pages. One page is a stylized unit. In one embodiment, a 124908.doc • 23 page can be divided into a plurality of segments and the minimum number of cells written at a time during the slice operation is stored in a column of memory cells. The one-segment includes user data and management H management data—typically including the error correction code (ECC) that has been calculated from the user data of the segment. Control " Sickles (as described below) are calculated when the data is programmed into the array and are also checked when reading f material from the array. Or these
ECC及/或其他管理f料係不同於其所屬之使用者資料的儲 存於頁或甚至不同區塊内。一使用者資料之區段通常係 512個位元組,對應於磁碟機中的一區段之大小。管理資 料-般係-額外的16至2〇位元組…較大數目的頁形成一 區塊,(例如)從8頁直至32、64、128或更多頁,不論何處 皆如此。ECC and/or other management materials are stored on pages or even different blocks from the user data to which they belong. A section of user data is typically 512 bytes, corresponding to the size of a section in the drive. Management Data - General - Extra 16 to 2 Bytes... A larger number of pages form a block, for example from 8 pages up to 32, 64, 128 or more pages, wherever they are.
段可包含在一基本程式化 °通常將一或多頁資料儲 可以儲存一或多個區段。 作為一範例,一 NAND快閃EEPR〇M係描述於圖9中,其 係劃分成1,024個區塊。在此範例中,各區塊中有8,512 行’對應於位元線BL0、BL1、…BL8511。在一具體實施 例中,一區瑰之所有位元線可在讀取及程式化操作期間同 時選擇。沿一共同字線且連接至任一位元線之記憶體單元 可同時程式化。 在另一具體實施例中,將該等位元線劃分成偶數位元線 與奇數位元線《在一奇數/偶數位元線架構中,沿著一共 同字線並連接至該等奇數位元線的記憶體單元係在一時間 處程式化,而沿著一共同字線並連接至偶數位元線的記憶 體單元係在另一時間處程式化。 124908.doc -24- 1378456 圖9顯示串聯連接以形成一NAND串之四個記憶體單元。 雖然顯不各NAND串包括四個單元,但也可使用多於或少 於四個(例如,16.、32、64或另一數目或記憶體單元可在 一 NAND争上)。該NAND串之一端子係經由一 >及極選擇閉 極(連接至選擇閘極汲極線SGD)而連接至一對應位元線, 而另一端子係經由一源極選擇閘極(連接至選擇閘極源極 線SGS)而連接至c源極。 在一成功程式化程序(具有驗證)結束時,該等記憶體單 ,之臨限電塵適當時應處於一或多個用於已程式.化記憶體 單元之臨限電壓之分佈内或一用於已抹除記憶體單元的臨 限電壓之分佈内。圖10說明當各記憶體單元儲存二位元資 料時用於記憶體單元陣列之範例性臨限電壓分佈。然而, j他具體實施例可使用每記憶體單元多於或少於二位元的 資料。圖1G顯示用於已抹除記憶體單元之__第__臨限電壓 分佈Ε。還描述用於已程式化記憶體單元之三個臨限電壓 分佈A、Β及C。在—具體實施例中,Ε分佈中的臨限電壓 係負數而A、Β及C分佈中的臨限電壓係正數。 圖之各不同臨限電壓範圍對應於用於該組資料位元之 預定值。程式化於記憶體單元内的資料與該單元之臨限電 壓位準之間的較關係取決於該等單元所採用的資料編碼 方案。例如,美國專利第6,222,762號及在2〇〇3年6月13曰 申請的美國專利申請公告案第2〇〇4/〇255〇9〇號"用於記憶 體系統之循轨單元”說明用於多狀態快閃記憶體單元之各 種資料編碼方案,二者全部内容均以引用形式併入本文。 124908.doc -25- U/8456 在一具體實施例中,使用一格雷碼指派將資料值指派至臨 限電壓範圍,使得若一浮動閘極之臨限電壓錯誤地偏移至 其相鄰物理狀態,則只會影響一位元。一範例指派"u"至 臨限電壓範圍E(狀態E),"1〇"至臨限電壓範圍a(狀態A), 〇〇至臨限電壓範圍B(狀態B)而"〇1"至臨限電壓範圍c(狀 1C)。不過,在其他具體實施例中未使用格雷碼。儘管 圖11顯示四種狀態,但本發明還可與其他多狀態結構一起 使用,包括該等包括多於或少於四種狀態之結構。 圖ίο顯示三個讀取參考電壓Vra、Vrb及Vrc,用於從記 憶體單疋讀取資料。藉由測試一給定記憶體單元之臨限電 壓疋否超過或低於Vra、Vrb及Vrc,該系統可決定該記憶 體單元所處之狀態。 圖1〇還顯示三個驗證參考電壓Vva、¥讣及乂冗。當將記 隐體單元程式化至狀態A時,該系統將測試該等記憶體單 兀是否具有一大於或等於Vva之臨限電壓。當將記憶體單 元程式化至狀態B時,該系統將測試該等記憶體單元是否 具有一大於或等於Vvb之臨限電壓。當將記憶體單元程式 化至狀態C時,該系統將決定該等記憶體單元是否具有大 於或·#於Vvc之臨限電壓。 在一具體實施例(全序列程式化)中,將記憶體單元從抹 除狀態E直接程式化至該等程式化狀態a、8或^之任一 者。例如,可先抹除一群欲程式化記憶體單元,使得所有 。己隐體單元均在抹除狀態£下。儘管正在將某些記憶體單 兀從狀態E程式化至狀態A ,但正在將其他記憶體單元從 124908.doc •26- 1378456 狀態E程式化至狀態B及/或從狀態E至狀態c。全序列程式 化由圖10之三個曲線箭頭來圖形描述。 圖11說明一程式化多狀態記憶體單元之兩遍式技術之一 範例’該等多狀態記憶體單元針對二不同頁:一下部頁與 一上部頁來儲存資料。描述四種狀態:狀態Ε(11)、狀態 Α(10)、狀態Β(00)及狀態C(01)e對於狀態E,兩頁均儲^ 一 "1"。對於狀態A,下部頁儲存一 "〇"而上部頁儲存一 "1"。對於狀態B,兩頁皆儲存"0"。對於狀態c,下部頁儲 存’’ 1"而上部頁儲存”〇"。應注意,儘管已將特定位元圖案 指派給該等狀態之各狀態,但還可指派不同的位元圖案 在第一遍程式化中,單元之臨限電壓位準根據欲程式化 於下邏輯頁内的位元而設定。若該位元係一邏輯"1 ",則 臨限電壓不會變化,因為其由於更早時候已抹除而處於適 當狀態中。然而,若欲程式化的位元係一邏輯"〇||,則該 單元之臨限位準係增加至狀態A,如箭頭53〇所示。 在-第二遍程式化中,單元之臨限電壓位準係根據欲程 式化於上邏輯頁内的位元而設定。若上邏輯頁位元將儲存 一邏輯1,則不發生任何程式化,由於取決於下部頁位元 之程式化,該單元處於玆耸壯能p A A々_ .从Λ* .....Segments can be included in a basic stylized ° usually one or more pages of data can be stored in one or more sections. As an example, a NAND flash EEPR 〇M is described in Figure 9, which is divided into 1,024 blocks. In this example, 8,512 rows in each block correspond to bit lines BL0, BL1, ... BL8511. In one embodiment, all of the bit lines of a region can be selected simultaneously during reading and stylizing operations. Memory cells along a common word line and connected to any bit line can be programmed simultaneously. In another embodiment, the bit lines are divided into even bit lines and odd bit lines "in an odd/even bit line structure, along a common word line and connected to the odd bits The memory cells of the meta-line are stylized at one time, while the memory cells along a common word line and connected to the even bit lines are stylized at another time. 124908.doc -24- 1378456 Figure 9 shows four memory cells connected in series to form a NAND string. Although it is shown that each NAND string includes four cells, more or less than four (e.g., 16, 32, 64, or another number or memory cells may compete on a NAND) may be used. One terminal of the NAND string is connected to a corresponding bit line via a > and a pole selective closed pole (connected to the selected gate drain line SGD), and the other terminal is connected via a source select gate (connected Connect to the c source to select the gate source line SGS). At the end of a successful stylized program (with verification), the memory sheets, if appropriate, should be in one or more distributions of threshold voltages for programmed memory cells or Used within the distribution of the threshold voltage of the erased memory cell. Figure 10 illustrates an exemplary threshold voltage distribution for a memory cell array when each memory cell stores a binary material. However, his specific embodiment may use more or less than two bits per memory unit. Figure 1G shows the ____ threshold voltage distribution 用于 for the erased memory cell. Three threshold voltage distributions A, Β, and C for the programmed memory cells are also described. In a particular embodiment, the threshold voltage in the Ε distribution is negative and the threshold voltage in the A, Β, and C distributions is positive. The different threshold voltage ranges of the map correspond to predetermined values for the set of data bits. The relationship between the data programmed into the memory unit and the threshold voltage level of the unit depends on the data encoding scheme employed by the units. For example, U.S. Patent No. 6,222,762, and U.S. Patent Application Publication No. 2, 4, 255, 255, "Tracking Unit for Memory Systems," filed on June 13, 2003. Various data encoding schemes for multi-state flash memory cells, both of which are incorporated herein by reference. 124908.doc -25- U/8456 In a specific embodiment, a Gray code is used to assign data The value is assigned to the threshold voltage range such that if the threshold voltage of a floating gate is erroneously shifted to its neighboring physical state, only one bit is affected. An example assignment "u" to the threshold voltage range E (State E), "1〇" to the threshold voltage range a (state A), 临 to the threshold voltage range B (state B) and "〇1" to the threshold voltage range c (like 1C) However, Gray code is not used in other embodiments. Although Figure 11 shows four states, the invention can also be used with other multi-state structures, including those having more or less than four states. Figure ίο shows three read reference voltages Vra, Vrb and Vrc for the slave record The system reads the data. By testing whether the threshold voltage of a given memory unit exceeds or falls below Vra, Vrb, and Vrc, the system can determine the state of the memory unit. Three verification reference voltages Vva, ¥讣, and 乂 are displayed. When the hidden unit is programmed to state A, the system will test whether the memory unit has a threshold voltage greater than or equal to Vva. When the memory cells are programmed to state B, the system will test whether the memory cells have a threshold voltage greater than or equal to Vvb. When the memory cells are programmed to state C, the system will determine such Whether the memory cell has a threshold voltage greater than or equal to Vvc. In a specific embodiment (full sequence programming), the memory cells are directly programmed from the erase state E to the stylized states a, 8 Or ^. For example, you can erase a group of stylized memory cells so that all the hidden cells are in the erase state. Although some memory files are being stylized from state E. To state A, but other The memory unit is programmed from 124908.doc • 26-1378456 state E to state B and/or from state E to state c. The full sequence stylization is graphically depicted by the three curved arrows of Figure 10. Figure 11 illustrates a stylization An example of a two-pass technique for a multi-state memory cell. The multi-state memory cells store data for two different pages: a lower page and an upper page. Four states are described: state Ε (11), state Α (10), state Β (00) and state C (01) e for state E, both pages are stored ^1 "1" For state A, the lower page stores a "〇" and the upper page stores a ";1". For state B, both pages are stored "0". For state c, the lower page stores ''1" and the upper page stores '〇". It should be noted that although a particular bit pattern has been assigned to each state of the states, different bit patterns can also be assigned at In a stylization, the threshold voltage level of the unit is set according to the bit to be programmed in the lower logical page. If the bit is a logic "1 ", the threshold voltage will not change because It is in an appropriate state because it has been erased earlier. However, if the stylized bit is a logical "〇||, then the threshold of the unit is increased to state A, as indicated by arrow 53〇 In the second pass stylization, the threshold voltage level of the unit is set according to the bit to be programmed in the upper logical page. If the upper logical page bit will store a logic 1, then it does not occur. Any stylization, due to the stylization of the lower page bit, the unit is in a strong position p AA々_. From Λ* .....
一遍程式化已程式化至狀態A, 武化該單7〇 ’使得將臨限 丨所示。若該單元由於該第 則該記憶體單元係在該第 124908.doc -27- 1378456 二遍中進一步程式化,使得將臨限電壓増加至狀態B内, 如箭頭532所示。該第二遍之結果係將該單元程式化至指 定儲存一邏輯"0"用於上部頁之狀態,而不改變用於下部 頁之資料。 ° 在一具體實施例中,若寫入足夠資料填滿一字線,則可 設定一系統來執行全序列寫入。若寫入不足夠的資料則 該程式化程序可使用接收的資料來程式化下部頁。當接收 到後續資料時,系統會接著程式化上部頁。在另一^體實 施例中,系統可採用程式化下部頁之模式開始寫入且若後 續接收到足夠資料以填滿一整個(或大多數)字線之記憶體 單元,則轉換成全序列程式化模式。此類具體實施例之更 多細節係揭示於發明者Sergy Anat〇lievich 〇的〇^^與¥奶 Li於2004年12月14日申請的美國專利申請公告案第 2006/0126390號,標題為"使用先前資料之非揮發性記憶 體之管線程式化"中,其全部内容係以引用方式併入本 文。 圖12A至C揭示用於程式化非揮發性記憶體之另一程 序,其藉由對於任一特定記憶體單元,在寫入用於前頁之 相鄰記憶體單元之後相對於一特定頁寫入該特定記憶體單 元減小浮動閘極至浮動閘極耦合之效應。在圖12八至(:所 教導之程序之一實施方案之範例中,該等非揮發性記憶體 單元使用四種資料狀態,每記憶體單元儲存二位元資料。 例如,假定狀態E係抹除狀態而狀態A、B及^係程式化狀 態。狀態E儲存資料u ^狀態A儲存資料〇1。狀態B儲存資 124908.doc •28- 1378456 料ίο狀_ C儲存資料⑽。此係—非格雷編碼之範例因 :-在相鄰狀態八及Β之間變化。還可使用資料至物 理資料狀態之其他編摄。久印陪牌„ _ ^ ^各記憶體早疋儲存二頁資料。出 ;參考目#肖該些資料頁稱為上部頁與下部頁;然而, 可,予其其他標記。參考用於圖以至C之程序之狀態A, 上部頁儲存位元0而下部頁儲存位元1。參考狀態B,上部 頁儲存位元1而下部頁儲存位元〇。參考狀態c,二頁健存 位元資料〇。 圖12A至c之程式化程序係二步驟程序。在第一步驟 中程式化下部頁。若下部頁欲保留資料1,則該記憶體 单元狀態保持在狀態E。若該資料欲程式化至〇,則該記憶 體單元之電壓之臨限值係升高使得將該記憶體單元程式化 至狀態B’。圖12A因此顯示從狀態E至狀態B,來程式化記憶 體單元。圖12 A中所示之狀態b,係一中間狀態B ;因此,該 驗證點係描述為Vvb',其小於Vvb。 在一具體實施例中’在將一記憶體單元從狀態E程式化 至狀態B'之後’則接著相對於其下部頁程式化其在NAND 串内相鄰記憶體單元(WLn+1)。例如,在程式化將用於一 連接至WL0之記憶體單元之下部頁之後,將程式化用於一 在相同NAND串上但連接至WL1之記憶體單元(相鄰記憶體 單兀Ο之下部頁。在程式化該相鄰記憶體單元之後,若該 更早記憶體單元具有一從狀態E升高至狀態B,之臨限電 壓’則該浮動閘極至浮動閘極耦合效應將會升高更早欲程 式化之記憶體單元之表觀臨限電壓。此將會具有加寬用於 124908.doc -29· 1378456 狀態B’之臨限電壓分佈之效應,如圖〗2b所示。當程式化 上部頁時將會修復臨限電壓分佈之此明顯加寬。 圖12C描述該程式化上部頁之程序。若該記憶體單元係 在抹除狀態E中且上部頁欲保持在丨,則該記憶體單元將保 持在狀態E中》若該記憶體單元係在狀態£中且其上部頁資 料欲程式化至0,則該記憶體單元之臨限電壓係升高使得 該記憶體單元在狀態A中。若該記憶體單元係在一中間臨 限電壓分佈550内且上部頁資料欲保持在1,則該記憶體單 元將程式化至最終狀態B。若該記憶體單元係在一中間臨 限電壓分佈550内且上部頁資料欲變成資料〇,則該記憶體 單元之臨限電壓係升高使得該記憶體單元在狀態C中。圖 12A至C所示之程序降低浮動閘極之間的耦合效應,因為 僅相鄰記憶體單元之上部頁程式化將會影響一給定記憶體 單元之表觀臨限電壓。 儘管圖12A至C相對於四種資料狀態及二資料頁提供一 範例,但圖12A至C所教導之該等概念可施加至多於或小 於四種狀態、不同於二頁之其他實施方案及/或其他資料 編碼。 圖13係說明利用圖12A至C之程式化方法來程式化記憶 體單元之次序之一具體實施例之一表格。對於連接至字線 WL0之記憶體單元’下部頁形成頁〇而上部頁形成頁2。對 於連接至字線WL1之記憶體單元,下部頁形成頁i而上部 頁形成頁4。對於連接至字線WL2之記憶體單元,下部頁 形成頁3而上部頁形成頁6»對於連接至字線WL3之記憶體 124908.doc •30· 單元,下部頁形成頁5而上部頁形成頁7。記憶體單元係從 頁0至頁7依據頁碼而程式化。在其他具體實施例中,還可 使用其他程式化次序。 在些具體實施例中,資料係沿一共同字線而程式化至 記隐體單元。因而,在施加該等程式化脈#f之前,選擇該 等字線之一用於程式化。該字線將稱為選定字線。一區塊 之剩餘字線係稱為未選定字線。該選定字線可具有一或兩 個相鄰字線。若該選定字線具有兩個相鄰字線,則在汲極 側的相鄰予線係稱為沒極側相鄰字線而在源極側的相鄰字 線係稱為源極側相鄰字線。例如,若WL2係選定字線則 WL1係源極側相鄰字線而WL3係汲極側相鄰字線。在一些 具體實施例中,一區塊的記憶體單元係從源極至汲極側來 程式化。例如,先程式化連接至WL〇之記憶體單元,隨後 程式化在WL1上的記憶體單元,隨後程式化在WL2上的記 憶體單7G等等》如上述,圖13描述在此次序上的一微小變 化,其仍一般從源極侧程式化至汲極側。 圖14係說明一用於程式化連接至一選定字線之記憶體單 元之程式化程序之一流程圖。因而,圖14之程序係用於實 施圖10之全序列程式化、圖u之兩遍程式化技術之一遍 (第一遍或第二遍)或圖12A至C及13之二步驟程式化技術之 一遍(第一遍或第二遍)。因為一程式化程序可包括程式化 多個頁,故該程式化程序可包括多次執行圖14之程序。 在圖14之程序之一實施方案中,在程式化之前抹除(按 區塊或其他單位)記憶體單元(步驟64〇)。在一具體實施例 124908.doc •31 · 中記憶體單兀係藉由升高卩井至一抹除電壓(例如,20伏特) 持續足夠時期並在該等源極及位元線正在浮動時將一選定 區塊之字線接地來加以抹除。由於電容性耦合,未選定字 線、位7L線、選擇線及C源極也上升至抹除電壓之一明顯 部分。因而將一強電場施加至選定記憶體單元之穿隧氧化 物層並隨著一般藉由Fowler-Nordheim穿随機制將該等 汗動閘極之電子發射至基板側來抹除該等選定記憶體單元 之資料。隨者將電子係從浮動閘極傳送至p井區域,降低 選定單兀之臨限電壓《抹除可在整個記憶體陣列、分離 區塊或其他單元單位上執行β在抹除記憶體單元之區塊之 後可按本文所述來程式化或局部程式化各種記憶體單 元。應注意,在步驟040執行的抹除不一定在程式化一區 塊之各字線之前執行。確實,可抹除該區塊,接著可程式 化各字線而不在程式化該等字線之間進行抹除。 在步驟642,執行軟程式化以使用於該等抹除記憶體單 几之抹除臨限電壓之分佈變窄。某些記憶體單元由於該抹 除程序而處於一深於所必需之抹除狀態下。軟程式化可施 加較小程式化脈衝來移動該等抹除記憶體單元之臨限電 壓,使之更靠近該抹除驗證位準。在步驟650中,一"資料 載入"命令係由控制器450發出並輸入至狀態機412。在步 驟652,提供指定頁位址之位址資料至解碼器。在步驟 654,輸入用於該定址頁之一頁程式資料用於程式化。例 如,在一具體實施例中,可輸入528位元組的資料。將該 資料鎖存於用於該等選定位元線之適當暫存器/鎖存器 124908.doc -32- 1378456 中°在一些具體實施例中,還將資料鎖存於用於該等選定 位元線之一第二暫存器中以供驗證操作使用。在步驟 656,—"程式"指令係由控制器450接收並提供至狀態機 412 〇 由該"程式"指令觸發後,在步驟654中鎖存的資料將使 用施加至適當字線之脈衝而程式化至受狀態機412控制的 該等選定記憶體單元中》在步驟658,將程式化電壓信號 Vpgm(例如,一系列脈衝)初始化至開始量值(例如,〜i2 v 或另一適當位準)並將狀態機412所維持之一程式計數器pc 初始為0。在步驟660,施加該程式信$Vpgm之一脈衝至 選疋予線。备·將邏輯"0"儲存在一特定資料鎖存器中,指 示應程式化對應記憶體單元,則將對應位元線接地。另一 方面,若將邏輯"丨"儲存在特定鎖存器中,指示對應記憶 體單TL應該保持在其當前資料狀態,則將對應位元線連接 至Vdd以禁止程式化。 在步驟662,使用適當目標位準組來驗證該等選定記憶 體單凡之狀態,如上所述。若偵測到一選擇單元之臨限電 壓已達到適#目標位準,則將儲存於對應資料鎖存器中的 貝料變成-邏輯"1"。若偵測到臨限電壓尚未達到適當目 標位準,則不改變儲存在對應資料鎖存器中的資料。以此 =式’不必程式化具有__邏輯"丨"儲存於其對應資料鎖存 器中之一位元線。當所右咨%‘ 田m有貢枓鎖存器均在儲存邏輯"Γ, 時’該狀態機瞭解已程式化所 π— W有選定早兀^在步驟664 中,檢查所有資料鎖存9§熹 仔器疋否均儲存邏輯"1"。若是,則 124908.doc •33- 該程式化程序已完成並成功,因為所有選定記憶體單元均 已程式化並驗5a·至其目標狀態。在步驟666中報告__ II通過" 狀悲。應注意’在-些實施方案+,在步驟664,檢查至 少一預定數目的資料鎖存器是否正在儲存一邏輯"丨"\此 預定數目可小於所有資料鎖存器數目,藉此允許該程式化 程序以在所有記憶體單元已到達其適當驗證位準之前停 止。在讀取程序期間,可則吏用錯誤校正來校正未成功程 式化的記憶體單元。 右在步驟664決定並非所有資料鎖存器均正在儲存邏輯 1 ,則該程式化程序會繼續。在步驟668,針對一程式化 限值來檢查該程式計數器pc ^然而,一程式化限值之範例 係20 ;然而可在各種實施方案中使用其他值。若該程式計 數器pc不小於該程式化限值,則在步驟669決定尚未成功 程式化之記憶體單元數目是否等於或小於一預定數目。若 未成功程式化記憶體單元數目等於或小於該預定數目,則 在步驟671標記該程式化程序為通過並報告一通過狀態。 在許多情形令’可在讀取程序期間使用錯誤校正來校正未 成功程式化之記憶體單元。然而’若未成功程式化記憶體 單疋數目大於該預定數目’則在步驟670標記該程式化程 序為失敗並報告一失敗狀態。若該程式計數器PC小於該程 式化限值(例如,20) ’則在步驟672按步階大小(例如,〇.2 至0.4伏特步階大小)增加Vpgm脈衝之量值並遞增該程式計 數器PC。在步驟672後’該程序迴路返回至步驟660以施加 下一 Vpgm脈衝》 124908.doc •34- 1378456 般而。,在該等驗證操作(例如在圖14之步驟662期間 執行之驗證操作)及讀取操作期間,該選定字線係連接至 一電壓,其位準係指定用於各讀取及驗證操作以便決定 相關記憶體單元之一臨限電壓;^已㈣此位準。在施加 字線電壓之後,測量該記憶體單元之傳導電流,以決定是 否回應施加至該字線之電壓接通該記憶體單元。若測量該 傳導電流大於一特定值,則假定該記憶體單元接通且施加 至該子線之電壓大於該記憶體單元之臨限電遷。若測量該 傳導電流不大於該特定值,則假定該記憶體單元不接通且 施加至該字線之電壓不大於該記憶體單元之臨限電壓。 存在許多方法以在一讀取或驗證操作期間測量一記憶體 單元之傳導電流。在一範例中,一記憶體單元之傳導電流 係根據感測放大器中的一專用電容器的放電或充電速率來 測量。在另一範例中’選定記憶體單元之傳導電流允許 (或無法允許)包括該記憶體單元之NAnd串釋放對應位元 線電荷。在一段時間之後測量位元線上的電壓以查看其是 否已放電。 已觀察到’特別用於靠近源極側選擇閘極之WL〇(相對 於其他字線在該NAND串之末端處並最先程式化的字線)在 選疋字線上的程式干擾邊界取決於相鄰記憶體單元之狀 態。若該相鄰記憶體單元在抹除狀態中,則在增壓程序期 間(在程式化禁止期間)在該相鄰記憶體單元下面的通道區 域應在一傳導狀態中。然而,若局部程式化該相鄰記憶體 單元(例如在中間狀態550,如圖12A至c中B,所示,或者另 J24908.doc -35- 1378456 一並非期望最终程式化狀態),則在該相鄰記憶體單元下 面的通道區域可變成在截止狀態或比其局部程式化情況下 傳導更少。在此後者情況t,將不同於期望來增壓在連接 至選定字線之禁止記憶體單元下面的通道區域並可能不充 分地增壓以避免程式干擾。由此,用以獲得適當減小程式 干擾之Vpass之最佳值取決於相鄰記憶體單元之狀態。為 了移除此資料相依性,提出比施加至其他未選定記憶體單 元之增壓電壓’施加一較高增壓電壓至相鄰記憶體單元。 由此,在相鄰記憶體單元下面的通道區域將會獨立於相鄰 記憶體單元内所儲存之資料而處於其適當傳導狀態中。 圖15描述一 NAND串,其如所提出將較高增壓電壓施加 至相鄰a己憶體單元來偏壓,欲在圖μ之步驟660之一反覆 處理期間禁止程式化。由於如圖15所述施加該等電壓,增 壓該NAND串之至少一部分(若非全部),使得禁止程式 化。圖15之NAND串包括八個記憶體單元75〇、752、754、 756、75 8、760、762及764。記憶體單元750及764係相對 於其他記憶體單元在該NAND串之末端處。該等八個記憶 體單元之各記憶體單元包括一浮動閘極(FG)以及一控制閘 極(CG)。浮動閘極之各浮動閘極間係源極/汲極區77〇。在 些實施方案中,存在一 P型基板(例如,矽)、在該基板 内的N井及在]^井内的一p井(未繪示全部以增加圖式可讀 性)。應注意,P井可包含一所謂通道植入,其通常係一 P ^植入’該p型植入決定或幫助決定該等記憶體單元之臨 限電壓及其他特性。該等源極/及極區77G係形成於P井内 124908.doc -36- 1378456 的N+擴散區。 在該NAND串之一端係一汲極側選擇閘極766。汲極側選 擇閘極766經由位元線接點774將該nAnd串連接至對應位 元線。在該NAND串之另一端係一源極選擇閘極768。源極 選擇閘極768將該NAND串連接至共同源極線772。在程式 化期間,連接至選定字線(例如,記憶體單充記憶體 單元經由選定字線在其控制閘極接收程式化.電壓Vpgm。 將大約8至9伏特之增壓電壓vpass施加至不選擇用於程式 化之記憶體單元(例如’記憶體單元754、756、758、 760、762、及764)之控制閘極,除相鄰記憶體單元外。一 較高增壓電壓VpassH係經由字線WL1而提供至相鄰記憶體 單元752之控制閘極。VpassH係一高於Vpass之電壓。在一 具體實施例中,VpassH係比VpassH高1至4伏特,或可針 對特定具體實施例在適當時運用另一差異。在一實施方案 中,VpassH比Vpass高一量,該量等於狀態£與狀態B,(參 見圖12B)之間的差異。應注意,使VpassH過高可能在接收 VpassH之記憶體單元上引起程式干擾。 取決於欲儲存的資料,可局部程式化或可不程式化相鄰 記憶體單兀752 ^例如,若欲儲存資料係一 ”丨",則記憶體 單元將仍處於抹除狀態E。若欲儲存資料係一 "〇",則可能 已將記憶體單元移動至中間狀態B,(圖12A至c之臨限電壓 550)。 源極選擇閘極768係在一隔離狀態下,在其閘極(G)接收 0伏特。將一低電壓施加至共同源極線772。此低電壓可以 124908.doc •37- 1378456 係〇伏特。然而,源極電壓亦可稍高於〇伏特,以提供源極 側選擇閘極之較佳隔離特性。向汲極側選擇閘極766施加 通常處於電源電壓Vdd(例如,2 5伏特)範圍内的電壓 Vsgd。經由對應位元線將〇伏特施加於位元線接點,以 致能選定記憶體單元750之程式化。由於該等增壓電壓, 增壓該NAND串之通道區域(如上所述)。由於記憶體單元 750之浮動閘極與通道之間的電壓差已減小,故禁止程式 化。圖15顯示區780,其包括位於基板表面之增壓通道區 域781 (在源極/汲極區77〇之間並在該等浮動閘極/控制閘極 堆疊下面)與在該增壓通道區域下面的一空乏層(一由於增 壓至一高電壓的通道而增加電場之區域)。 圖16係一時序圖’其說明針對圖14之步驟660之一反覆 處理用於施加圖15所示各種信號用於一未選定NAND串之 時序之一範例。圖16顯示從t2至t6位元線電壓VBL在Vdd(例 如’ 2_5伏特)’從而禁止與該特定位元線相關聯之nand 串。選擇閘極電壓VSGD(在選擇電晶體SGD之控制閘極處 的電壓)係在tl升高至5伏特,接著在t2降低至2.5伏特(例如 Vdd) ’此處其保持直至t6。在^與^之間Vsgd在5伏特的週 期係視需要地用於增加NAND串之預充電電壓位準。在未 選疋子線Vuwl上的電壓係在tl處升高至Vdd以允許預充 電’接著在t2升高至大約Vpass以增壓與該等未選定位元 線相關聯之NAND串。傳遞電壓vpass將會保持在未選定字 線,直至大約t5 »應注意,在圖丨5之具體實施例中,將 VUWL施加至除相鄰字線外的所有字線。在未選定相鄰字線 124908.doc -38· 1378456 (在圖15中係WL1(對應於記憶體單元752))上的電壓Vnuwl 係在11升高至vdd以允許預充電並接著在t2升高至大約 VpassH以幫助增壓與該等未選定位元線相關聯之nand 串。傳遞電壓VpassH將會保持在該等未選定相鄰字線上, 直至大約t5。在選定字線上的電壓vSWL(例如在圖15中係 WLO ’對應於記憶體單元750)係在ti升高至vdd以允許預 充電。在t3’施加程式化脈衝直至t5。在一範例中,該等 程式化脈衝可在12伏特與20伏特之間變化。應注意,該源 極側選擇閘極(vSGS)之控制閘極係始終〇伏特而源極電壓Vs 係在11之前升咼至並保持於此直至t6。應注意,上述各種 信號之確切時序可根據特定實施方案而變化。應注意,圖 1 5對應於在圖16之時間t4的該等電壓信號之狀態。在一些 實施方案中’在tl至t2期間,Vuwl、Vswl及Vnuwl係連接 至Vdd(或另一電壓>〇 v)。在其他實施方案中,在時間間 隔 tl 至 t2期間,Vuwl、Vswl及 Vnuwl為 〇 V。 圖1 5及16屬於當選擇其WL0時使用一較高增壓電壓用於 一未選定相鄒字線之情況。然而,本文所述之技術還屬於 選擇其他字線用於程式化之情形。例如,圖17顯示當選擇 連接至記憶體單元756之字線(例如,WL3)時禁止程式化 NAND串之情形β在此情況中,靠近選定字線之字線(例如 WL4)將會接收較高傳遞電壓VpassH。更明確而言,圖I? 顯示目標但未選定記憶體單元756正在接收Vpgm。相鄰記 憶體單元758接收VpassH。記憶體單元750、752、754、 760、762及764接收Vpass ^由於施加該等增壓電壓,在區 124908.doc •39· 1378456 780之表面處的通道區域781係增壓並完全禁止程式化記憶 體單元756。在圖17所述之信號時序類似於圖16之信號時 序。圖17僅顯示一範例,且在選擇其他字線用於程式化時 可使用施加VpassH至相鄰者。 應注意,在某些替代例中,在接收程式化電壓之記憶體 單元之源極側上的該等記憶體單元可接收一高於Vpass之 傳遞電壓《例如,記憶體單元750、752及/或754可接收 Vpass、VpassH或VpassO ’其中VpassO可能高於、低於或 類似於VpassH或Vpass。 圖15顯示記憶體單元754、756、758、760、762及764均 接收相同信號Vpass。同樣地,圖17顯示記憶體單元750、 752、758、760、762及764均接收相同信號Vpass。然而, 在一些具體實施例中,該些記憶體單元不一定接收確切相 同的電壓。例如,該等電壓可按字線變化,只要其(或一 子集)小於VpassH。 圖15及圖17二者描述用於所提出技術來修改上述自行增 壓方案。然而,所提出技術還可用於修改其他增壓方案。 圖18描述選擇WL0外的一字線用於程式化並使用所提出 技術來修改EASB增壓方案時的NAND串。目標但未選定記 憶體單元756接收Vpgm。相鄰記憶體單元758接收 VpassH。記憶體單元 750、752、760、762 及 764 接收Once stylized, it has been programmed to state A, and the order of the single 7〇 is made as shown by the threshold. If the unit is further programmed in the second pass of the 124908.doc -27-1378456 due to the second step, the threshold voltage is added to the state B as indicated by the arrow 532. The result of this second pass is to program the unit to specify the state of the storage "0" for the upper page without changing the data for the lower page. ° In one embodiment, if enough data is written to fill a word line, a system can be set to perform a full sequence of writes. If insufficient data is written, the stylized program can use the received data to program the lower page. When the follow-up data is received, the system will then program the upper page. In another embodiment, the system can start writing in the mode of the programmed lower page and convert to a full sequence if the subsequent received sufficient data to fill the memory unit of an entire (or most) word line. Mode. Further details of such specific embodiments are disclosed in U.S. Patent Application Publication No. 2006/0126390, filed on Dec. 14, 2004, by the inventor Sergy Anat 〇 vi vi vi , , , , , , , , , , , , , , , , , , , The use of non-volatile memory pipeline stylization of previous data is incorporated herein by reference. 12A-C disclose another procedure for programming non-volatile memory by writing to a particular page after writing to an adjacent memory cell for the previous page for any particular memory cell. Into this particular memory cell reduces the effect of floating gate to floating gate coupling. In the example of one of the embodiments of the teachings of Figures 12 to (10), the non-volatile memory cells use four data states, each of which stores two-bit data. For example, assume that state E is a wipe. In addition to the state, states A, B, and ^ are stylized states. State E stores data u ^ state A stores data 〇 1. State B storage resources 124908.doc • 28- 1378456 material ίο _ C storage data (10). Examples of non-Gray coding are: - change between adjacent states and Β. Other data can be used to record the physical data status. Jiuyin _ _ ^ ^ each memory stores two pages of data. The reference page is called the upper page and the lower page; however, it can be given other marks. Referring to the state A for the program of the figure and C, the upper page stores the bit 0 and the lower page stores the bit. Element 1. Reference state B, the upper page stores bit 1 and the lower page stores bit 〇. Reference state c, two pages of storage bit data 〇. The stylized program of Figures 12A to c is a two-step program. Stylize the lower page in the step. If the lower page wants to keep the data 1, The memory cell state remains in state E. If the data is to be programmed to 〇, the threshold of the voltage of the memory cell is raised to program the memory cell to state B'. The memory cell is programmed from state E to state B. The state b shown in Figure 12A is an intermediate state B; therefore, the verification point is described as Vvb', which is less than Vvb. In the example 'after staging a memory cell from state E to state B', then it is then programmed relative to its lower page to its adjacent memory cell (WLn+1) in the NAND string. For example, in stylized Will be used after a page connected to the lower part of the memory cell of WL0, will be programmed for a memory cell on the same NAND string but connected to WL1 (next page of the adjacent memory cell. After the adjacent memory unit, if the earlier memory unit has a rising state E to a state B, the threshold voltage 'the floating gate to floating gate coupling effect will rise earlier. Apparent threshold voltage of the memory unit. This will It has the effect of widening the threshold voltage distribution for state B' of 124908.doc -29· 1378456, as shown in Figure 2b. This apparent widening of the threshold voltage distribution will be fixed when the upper page is programmed. 12C describes the program of the stylized upper page. If the memory unit is in the erase state E and the upper page is to remain in the 丨, the memory unit will remain in the state E if the memory unit is in the state In the case where the upper page data is to be programmed to 0, the threshold voltage of the memory unit is raised such that the memory unit is in state A. If the memory unit is within an intermediate threshold voltage distribution 550 And if the upper page data is to remain at 1, the memory unit will be programmed to the final state B. If the memory cell is within an intermediate threshold voltage distribution 550 and the upper page data is to become data 则, then the threshold voltage of the memory cell is raised such that the memory cell is in state C. The procedure shown in Figures 12A through C reduces the coupling effect between the floating gates because only the top page stylization of adjacent memory cells will affect the apparent threshold voltage of a given memory cell. Although Figures 12A-C provide an example with respect to four data states and two data pages, the concepts taught by Figures 12A-C can be applied to more or less than four states, other embodiments than two pages and/or Or other material code. Figure 13 is a table showing one of the specific embodiments of the sequence of staging memory cells using the stylized methods of Figures 12A-C. For the memory cell unit connected to the word line WL0, the lower page forms a page and the upper page forms a page 2. For the memory cell connected to word line WL1, the lower page forms page i and the upper page forms page 4. For the memory cells connected to word line WL2, the lower page forms page 3 and the upper page forms page 6» for memory 124908.doc • 30· cells connected to word line WL3, the lower page forms page 5 and the upper page forms page 7. The memory unit is programmed from page 0 to page 7 according to the page number. In other embodiments, other stylized sequences may also be used. In some embodiments, the data is programmed to a hidden unit along a common word line. Thus, one of the word lines is selected for stylization prior to application of the stylized pulses #f. This word line will be referred to as the selected word line. The remaining word lines of a block are referred to as unselected word lines. The selected word line can have one or two adjacent word lines. If the selected word line has two adjacent word lines, the adjacent lines on the drain side are referred to as the non-polar side adjacent word lines and the adjacent word lines on the source side are referred to as the source side phases. Neighbor word line. For example, if WL2 is a selected word line, then WL1 is the source side adjacent word line and WL3 is the drain side adjacent word line. In some embodiments, a block of memory cells is programmed from the source to the drain side. For example, first programmatically connect to the memory cell of WL〇, then program the memory cell on WL1, then program the memory bank 7G on WL2, etc. As described above, Figure 13 depicts in this order. A small change is still typically programmed from the source side to the drain side. Figure 14 is a flow diagram showing one of the stylized procedures for staging a memory cell connected to a selected word line. Thus, the program of FIG. 14 is used to implement the full sequence of FIG. 10, one of the two-pass programming techniques of FIG. 10 (first pass or second pass) or the second step stylization techniques of FIGS. 12A to C and 13 One pass (first pass or second pass). Since a stylized program can include stylizing a plurality of pages, the stylized program can include executing the program of Figure 14 multiple times. In one embodiment of the process of Figure 14, the memory cells are erased (by block or other unit) prior to programming (step 64). In a specific embodiment 124908.doc •31 · the memory unit is maintained by raising the well to a wipe voltage (eg, 20 volts) for a sufficient period of time and when the source and bit lines are floating The word line of a selected block is grounded to be erased. Due to capacitive coupling, the unselected word line, bit 7L line, select line, and C source also rise to a significant portion of the erase voltage. Thus, a strong electric field is applied to the tunneling oxide layer of the selected memory cell and the selected memory is erased as the electrons of the sweat gates are generally emitted to the substrate side by Fowler-Nordheim randomization. Unit information. The electron system is transferred from the floating gate to the p-well region, and the threshold voltage of the selected single-turn is reduced. "Erasing can be performed on the entire memory array, the separation block or other unit units to erase the memory unit. The blocks can be followed by stylized or partially programmed various memory cells as described herein. It should be noted that the erasure performed at step 040 is not necessarily performed prior to stylizing the word lines of a block. Indeed, the block can be erased and the word lines can then be programmed without being erased between the stylized word lines. At step 642, a soft stylization is performed to narrow the distribution of the erase threshold voltages for the erased memory. Some memory cells are in a deeper state than necessary due to the erase procedure. The soft stylization can apply a smaller stylized pulse to move the threshold voltage of the erased memory cells closer to the erase verify level. In step 650, a "data load" command is issued by controller 450 and input to state machine 412. At step 652, address data specifying the page address is provided to the decoder. At step 654, one of the page program data for the addressed page is entered for stylization. For example, in one embodiment, 528 bytes of data can be entered. The data is latched into the appropriate register/latch 124908.doc -32-1378456 for the selected location line. In some embodiments, the data is also latched for use in the selection. One of the bit lines is in the second register for use in the verify operation. At step 656, the -"program" command is received by controller 450 and provided to state machine 412. After being triggered by the "program" command, the data latched in step 654 is applied to the appropriate word line. The pulse is programmed into the selected memory cells controlled by state machine 412. At step 658, the programmed voltage signal Vpgm (eg, a series of pulses) is initialized to a starting magnitude (eg, ~i2v or another A suitable level) and one of the program counters pc maintained by state machine 412 is initially zero. At step 660, one of the program signals $Vpgm is applied to the select line. Storing the logic "0" in a specific data latch, indicating that the corresponding memory cell should be stylized, then grounding the corresponding bit line. On the other hand, if the logic "丨" is stored in a specific latch indicating that the corresponding memory single TL should remain in its current data state, the corresponding bit line is connected to Vdd to disable stylization. At step 662, the appropriate target level set is used to verify the status of the selected memory, as described above. If it is detected that the threshold voltage of a selected unit has reached the appropriate target level, the bedding stored in the corresponding data latch becomes -logic"1". If it is detected that the threshold voltage has not reached the appropriate target level, the data stored in the corresponding data latch is not changed. It is not necessary to stylize a bit line stored in its corresponding data latch with __logic "丨". When the right side of the %' Tianm has a Gonggao latch in the storage logic "Γ, the state machine knows that the programmed π-W has been selected early ^ In step 664, check all data latches 9 § 熹 疋 疋 储存 储存 储存 储存 储存 储存 储存 储存 。 。 。 。 。. If so, then 124908.doc •33- The stylized program has completed and succeeded because all selected memory units have been programmed and tested 5a to their target state. In step 666, report __ II through " It should be noted that in some embodiments +, in step 664, it is checked whether at least a predetermined number of data latches are storing a logic "丨"\ this predetermined number may be less than the number of all data latches, thereby allowing The stylized program stops before all memory cells have reached their proper verify level. During the reading process, error correction can be used to correct unsuccessfully programmed memory cells. Right at step 664, it is determined that not all data latches are storing logic 1, and the stylized program continues. At step 668, the program counter pc is checked for a stylized limit. However, an example of a stylized limit is 20; however, other values may be used in various implementations. If the program counter pc is not less than the stylized limit, then in step 669 it is determined whether the number of memory cells that have not been successfully programmed is equal to or less than a predetermined number. If the number of unsynchronized memory cells is equal to or less than the predetermined number, then the stylized program is marked as passed and a pass status is reported in step 671. In many cases, error correction can be used during the reading process to correct unsuccessfully programmed memory cells. However, if the number of unsynchronized memory cells is greater than the predetermined number, then the stylized program is marked as failed in step 670 and a failure status is reported. If the program counter PC is less than the stylized limit (eg, 20)', then in step 672, the magnitude of the Vpgm pulse is incremented by step size (eg, 〇.2 to 0.4 volt step size) and the program counter PC is incremented. . After step 672, the program loop returns to step 660 to apply the next Vpgm pulse, 124908.doc • 34-1378456. During such verify operations (eg, verify operations performed during step 662 of FIG. 14) and during read operations, the selected word line is coupled to a voltage whose level is designated for each read and verify operation. Determine the threshold voltage of one of the relevant memory cells; ^ has (four) this level. After the word line voltage is applied, the conduction current of the memory cell is measured to determine whether the voltage applied to the word line is turned on to turn on the memory cell. If the measured conduction current is greater than a particular value, it is assumed that the memory cell is turned "on" and the voltage applied to the sub-line is greater than the threshold current of the memory cell. If the measured conduction current is not greater than the specific value, it is assumed that the memory cell is not turned on and the voltage applied to the word line is not greater than the threshold voltage of the memory cell. There are many ways to measure the conduction current of a memory cell during a read or verify operation. In one example, the conduction current of a memory cell is measured based on the rate of discharge or charge of a dedicated capacitor in the sense amplifier. In another example, the conduction current of the selected memory cell allows (or does not allow) the NAnd string including the memory cell to release the corresponding bit line charge. The voltage on the bit line is measured after a period of time to see if it has been discharged. It has been observed that the program disturbing boundary on the selected word line depends on the WL 特别 (closed to the word line at the end of the NAND string relative to other word lines). The state of adjacent memory cells. If the adjacent memory cell is in the erased state, then the channel region underneath the adjacent memory cell during the boosting process (during stylization prohibition) should be in a conducting state. However, if the adjacent memory unit is partially programmed (eg, in intermediate state 550, as shown in B of Figures 12A-c, or another J24908.doc-35-1378456 is not expected to be finalized), then The channel area underneath the adjacent memory cell can become less conductive in the off state or in the case of its partial stylization. In this latter case t, it will be different than would be expected to boost the channel area underneath the memory block connected to the selected word line and may not be sufficiently boosted to avoid program disturb. Thus, the optimum value of Vpass used to obtain a suitable reduction in program disturb depends on the state of the adjacent memory cells. In order to remove this data dependency, a higher boost voltage is applied to the adjacent memory cells than the boost voltage applied to other unselected memory cells. Thus, the channel area underneath adjacent memory cells will be in its proper conduction state independent of the data stored in adjacent memory cells. Figure 15 depicts a NAND string that is biased by applying a higher boost voltage to an adjacent a memory cell as proposed, and is not intended to be stabilized during one of the steps 660 of the process. Since the voltages are applied as described in Figure 15, at least a portion, if not all, of the NAND strings are boosted, stabilizing. The NAND string of Figure 15 includes eight memory cells 75A, 752, 754, 756, 758, 760, 762, and 764. Memory cells 750 and 764 are at the end of the NAND string relative to other memory cells. Each of the memory cells of the eight memory cells includes a floating gate (FG) and a control gate (CG). The floating gates of the floating gates are between the source/drain regions 77〇. In some embodiments, there is a P-type substrate (e.g., germanium), a N-well in the substrate, and a p-well in the well (not all shown to increase readability). It should be noted that the P well may comprise a so-called channel implant, which is typically a P^ implant' that determines or helps determine the threshold voltage and other characteristics of the memory cells. The source/pole regions 77G are formed in the N+ diffusion region of the P well 124908.doc -36-1378456. A gate 766 is selected on one of the drain sides of the NAND string. The drain side select gate 766 connects the nAnd string to the corresponding bit line via bit line contact 774. A source select gate 768 is coupled to the other end of the NAND string. Source select gate 768 connects the NAND string to a common source line 772. During stylization, it is connected to the selected word line (for example, the memory single-charge memory unit receives the stylized voltage at its control gate via the selected word line. Voltage Vpgm. Apply a boost voltage vpass of approximately 8 to 9 volts to no Selecting control gates for staging memory cells (eg, 'memory cells 754, 756, 758, 760, 762, and 764), except for adjacent memory cells. A higher boost voltage VpassH is via Word line WL1 is provided to the control gate of adjacent memory cell 752. VpassH is a voltage higher than Vpass. In a specific embodiment, VpassH is 1 to 4 volts higher than VpassH, or may be for a specific embodiment Another difference is applied where appropriate. In one embodiment, VpassH is one higher than Vpass, which is equal to the difference between state £ and state B, (see Figure 12B). It should be noted that making VpassH too high may be receiving Program noise is caused by the memory unit of VpassH. Depending on the data to be stored, the adjacent memory unit can be partially or unprogrammed. For example, if you want to store the data system, the memory unit will In erasing state E. If you want to store the data system "〇", the memory unit may have been moved to the intermediate state B (Fig. 12A to c threshold voltage 550). Source selection gate 768 system In an isolated state, it receives 0 volts at its gate (G). A low voltage is applied to the common source line 772. This low voltage can be 124908.doc • 37-1378456 volts. However, the source voltage is also It may be slightly higher than 〇V to provide better isolation characteristics of the source side selection gate. The gate 766 is applied to the drain side to apply a voltage Vsgd which is usually in the range of the power supply voltage Vdd (for example, 25 volts). The bit line applies volts to the bit line contacts to enable programming of the selected memory cell 750. Due to the boost voltage, the channel region of the NAND string is boosted (as described above) due to the memory cell. The voltage difference between the floating gate and the channel of 750 has been reduced, so stylization is prohibited. Figure 15 shows a region 780 that includes a boost channel region 781 on the surface of the substrate (between source/drain regions 77〇) And below the floating gate/control gate stack) a depletion layer below the region of the booster channel (a region that increases the electric field due to pressurization to a high voltage channel). Figure 16 is a timing diagram of the description of one of the steps 660 of Figure 14 for repeated processing. An example of the timing of applying the various signals shown in Figure 15 for an unselected NAND string. Figure 16 shows that the bit line voltage VBL from t2 to t6 is at Vdd (e.g., '2_5 volts'' thereby inhibiting correlation with the particular bit line. The nand string is selected. The gate voltage VSGD (the voltage at the control gate of the selected transistor SGD) is raised to volts at tl, then lowered to 2.5 volts at t2 (eg, Vdd) 'here it remains until T6. The Vsgd between 5 and volts is optionally used to increase the precharge voltage level of the NAND string. The voltage on the unselected dice line Vuwl is raised to Vdd at t1 to allow pre-charging' then raised to approximately Vpass at t2 to boost the NAND strings associated with the unselected positioning elements. The pass voltage vpass will remain at the unselected word line until approximately t5 » It should be noted that in the particular embodiment of Figure 5, VUWL is applied to all word lines except the adjacent word line. The voltage Vnuwl on the unselected adjacent word line 124908.doc -38· 1378456 (in Figure 15 WL1 (corresponding to memory unit 752)) is raised to 11 in vdd to allow pre-charging and then at t2 liters. Up to about VpassH to help boost the nand string associated with the unselected positioning elements. The pass voltage VpassH will remain on the unselected adjacent word lines until approximately t5. The voltage vSWL on the selected word line (e.g., WLO ' corresponds to memory cell 750 in Figure 15) is raised to vdd at ti to allow for pre-charging. A stylized pulse is applied at t3' until t5. In one example, the stylized pulses can vary between 12 volts and 20 volts. It should be noted that the control gate of the source side select gate (vSGS) is always volts and the source voltage Vs is ramped up to and maintained until t6 before 11. It should be noted that the exact timing of the various signals described above may vary depending on the particular implementation. It should be noted that Fig. 15 corresponds to the state of the voltage signals at time t4 of Fig. 16. In some embodiments 'Vuwl, Vswl and Vnuwl are connected to Vdd (or another voltage > 〇 v) during t1 to t2. In other embodiments, Vuwl, Vswl, and Vnuwl are 〇 V during the time interval tl to t2. Figures 1 and 16 pertain to the use of a higher boost voltage for an unselected phase of the word line when WL0 is selected. However, the techniques described in this article are also part of the case of choosing other word lines for stylization. For example, Figure 17 shows the situation in which the stylized NAND string is disabled when a word line (e.g., WL3) connected to the memory cell 756 is selected. In this case, the word line (e.g., WL4) near the selected word line will receive High transfer voltage VpassH. More specifically, Figure I? shows the target but the unselected memory unit 756 is receiving Vpgm. Adjacent memory unit 758 receives VpassH. The memory cells 750, 752, 754, 760, 762, and 764 receive Vpass. Due to the application of the boost voltage, the channel region 781 at the surface of the region 124908.doc • 39· 1378456 780 is boosted and completely disabled. Memory unit 756. The signal timing described in Figure 17 is similar to the signal timing of Figure 16. Figure 17 shows only an example, and applying VpassH to neighbors can be used when selecting other word lines for stylization. It should be noted that in some alternatives, the memory cells on the source side of the memory cells receiving the stylized voltage may receive a transfer voltage that is higher than Vpass "eg, memory cells 750, 752, and / Or 754 can receive Vpass, VpassH, or VpassO 'where VpassO may be higher, lower, or similar to VpassH or Vpass. Figure 15 shows that memory cells 754, 756, 758, 760, 762, and 764 each receive the same signal Vpass. Similarly, Figure 17 shows that memory cells 750, 752, 758, 760, 762, and 764 all receive the same signal Vpass. However, in some embodiments, the memory cells do not necessarily receive exactly the same voltage. For example, the voltages can vary by word line as long as they (or a subset) are less than VpassH. Both Figure 15 and Figure 17 describe the techniques used to modify the self-voltage boosting scheme described above. However, the proposed technique can also be used to modify other boosting schemes. Figure 18 depicts a NAND string when a word line other than WL0 is selected for programming and using the proposed technique to modify the EASB boosting scheme. The target but unselected memory unit 756 receives Vpgm. Adjacent memory unit 758 receives VpassH. Memory unit 750, 752, 760, 762 and 764 receive
Vpass»記憶體單元754接收隔離電壓(例如,〇伏特)《由於 施加該專增壓電壓,產生一較高增壓通道區域與一較低增 壓通道區域。例如’圖18描述區782,其包括位於基板表 124908.doc 40- 1378456 面之較高增壓通道區域783與在該較高增壓通道區域下面 的一空乏層。圖18還顯示區域784,包括位於基板表面之 較低增壓通道區域785與在該較低增壓通道區域下面的一 空乏層。該較高增壓通道區域引起完全禁止程式化記憶體 單元756。在圖18所述之信號時序類似於圖16之信號時 序。圖18僅顯示一範例’且在選擇其他字線用於程式化時 可使用施加VpassH至相鄰者。應注意,若選擇WL〇用於程 式化’則如圖15所示’施加所提出技術至EASB增壓方 案,由於在自行增壓與EASB之間的差異係基於源極側且 在選擇WLO用於程式化時不存在任何源極側。 圖19描述選擇WLO外的一字線用於程式化並使用所提出 技術來修改REASB增壓方案時的NAND串。目標但未選定 記憶體單元756接收Vpgm。相鄰記憶體單元758接收 VpassH。記憶體單元750、760、762及764接收Vpass。記 憶體單元752接收隔離電壓(例如,〇伏特)^記憶體單元754 經由其連接字線接收中間電壓Vgp(例如,2至5伏特)。由 於施加該等增壓電壓’產生一較高增壓通道區域與一較低 增壓通道區域。例如,圖19描述區788,其包括位於基板 表面之較高增壓通道區域789與在該較高增壓通道區域下 面的一空乏層。圖19還顯示區790,其包括位於基板表面 之較低增壓通道區域791與在該較低增壓通道區域下面的 一空乏層。該較高增壓通道區域引起完全禁止程式化記憶 體單元756。在圖19所述之信號時序類似於圖16之信號時 序,Vgp具有類似於Vpass之時序。圖19僅顯示一範例,且 124908.doc •41 · 1378456 在選擇其他字線用於程式化時可使用施加VpassH至相鄰 者。應注意,若選擇WLO用於程式化,則如圖15所示,施 加所提出技術至rEASB增壓方案,由於在自行增壓與 EASB之間的差異係基於源極側且在選擇wl〇用於程式化 時不存在任何源極側。 圖20描述選擇WL〇外的一字線用於程式化並使用所提出 技術來修改一替代性增壓方案時的NAND串。目標但未選 定記憶體單元758接收Vpgm。相鄰記憶體單元760接收The Vpass»memory unit 754 receives an isolated voltage (e.g., volts) "by applying the dedicated boost voltage, a higher boost channel region and a lower boost channel region are created. For example, Figure 18 depicts a region 782 that includes a higher pressurized channel region 783 on the surface of the substrate table 124908.doc 40-1378456 and a depletion layer below the region of the higher pressurized channel. Figure 18 also shows a region 784 comprising a lower pressurized channel region 785 on the surface of the substrate and a depletion layer below the lower pressurized channel region. This higher boost channel region causes a complete disable of the stylized memory unit 756. The signal timing described in Figure 18 is similar to the signal timing of Figure 16. Figure 18 shows only an example ' and can use VpassH to the neighbor when selecting other word lines for stylization. It should be noted that if WL〇 is selected for stylization' then the proposed technique is applied to the EASB boosting scheme as shown in Figure 15, since the difference between self-boosting and EASB is based on the source side and is used in selecting WLO. There is no source side when stylized. Figure 19 depicts a NAND string when a word line outside the WLO is selected for programming and using the proposed technique to modify the REASB boost scheme. The target but not selected memory unit 756 receives Vpgm. Adjacent memory unit 758 receives VpassH. Memory cells 750, 760, 762, and 764 receive Vpass. Memory unit 752 receives an isolation voltage (e.g., volts). Memory unit 754 receives an intermediate voltage Vgp (e.g., 2 to 5 volts) via its connected word line. A higher boost channel region and a lower boost channel region are created by the application of the boost voltages. For example, Figure 19 depicts a region 788 that includes a higher pressurized channel region 789 on the surface of the substrate and a depletion layer below the region of the higher pressurized channel. Figure 19 also shows a region 790 comprising a lower pressurized channel region 791 on the surface of the substrate and a depletion layer below the lower pressurized channel region. This higher boost channel region causes the program bus memory unit 756 to be completely disabled. The signal timing described in Fig. 19 is similar to the signal timing of Fig. 16, and Vgp has a timing similar to Vpass. Figure 19 shows only one example, and 124908.doc •41 · 1378456 can be used to apply VpassH to neighbors when selecting other word lines for stylization. It should be noted that if WLO is selected for stylization, as shown in Figure 15, the proposed technique is applied to the rEASB boosting scheme, since the difference between self-boosting and EASB is based on the source side and is used in the selection of wl There is no source side when stylized. Figure 20 depicts a NAND string when a word line outside of WL is selected for programming and using the proposed technique to modify an alternative boosting scheme. The target but unselected memory unit 758 receives Vpgm. Adjacent memory unit 760 receives
VpassH。記憶體單元 75〇、756、760、762、766、…接收VpassH. Memory unit 75〇, 756, 760, 762, 766, ... receive
Vpass °如上述,本文所述技術可使用長於八個記憶體單 元之NAND串。圖20顯示具有八個以上的記憶體單元之一 NAND串之一部分。記憶體單元752及764經由其連接字線 接收隔離電壓。記憶體單元754經由其連接字線接收中間 電壓Vpg。由於施加該等增壓電壓,產生一較高增壓通道 區域與一中間增壓通道區域及一較低增壓通道區域。例 如’圖20描述區794,其包括位於基板表面處的較高增壓 通道區域795與一在較高增壓通道區域下面的空乏層;區 798’其包括位於基板表面之中間增壓通道區域799與一位 於較低增壓通道區域下面的空乏層;及區域796,其包括 位於基板表面的較低增壓通道區域797與一在較低增壓通 道區域下面的空乏層^該較高增壓通道區域引起完全禁止 程式化記憶體單元758。在圖20所述之信號時序類似於圖 16之信號時序,Vgp具有類似於vpass之時序。圖20僅顯示 一範例’且在選擇其他字線用於程式化時可使用施加 124908.doc -42- 1378456Vpass ° As described above, the techniques described herein can use NAND strings that are longer than eight memory cells. Figure 20 shows a portion of a NAND string having more than eight memory cells. Memory cells 752 and 764 receive the isolation voltage via their connected word lines. The memory unit 754 receives the intermediate voltage Vpg via its connected word line. Due to the application of the boosted voltages, a higher boost channel region and an intermediate boost channel region and a lower boost channel region are created. For example, Figure 20 depicts a region 794 that includes a higher pressurized channel region 795 at the surface of the substrate and a depletion layer below the region of the higher pressurized channel; a region 798' that includes a region of the pressurized channel in the middle of the surface of the substrate. 799 and a depletion layer located below the lower pressurized passage region; and region 796 comprising a lower pressurized passage region 797 on the surface of the substrate and a depleted layer below the lower pressurized passage region The compressed channel region causes the program memory unit 758 to be completely disabled. The signal timing described in Figure 20 is similar to the signal timing of Figure 16, and Vgp has a timing similar to vpass. Figure 20 shows only an example ' and can be used when selecting other word lines for stylization. 124908.doc -42 - 1378456
VpassH至相鄰者。 在一些具體實施例中,該系統可在完成一當前選定記憶 體單元之程式化之前局部程式化一 NAND串之一個以上的 記憶體早元。例如’圖12A至C之程式化程序可修改以在 繼續完成第一子線上的程式化之前執行第一遍/步驟用於 三個字線。在一將三個資料頁儲存於一記憶體單元内之範 例中,以下列次序來寫入資料:(1)將下部頁資料寫入 WLn ’(2)將下部頁資料寫入WLn+1,(3)將中間頁資料寫 入WLn,(4)將下部頁資料寫入wLn+2,(5)將中間頁資料 寫入WLn+1,及(6)將下部頁資料寫入WLn以完成將所有3 頁寫入WLn。還可使用其他方法/方案。在該些範例中, 存在二個字線可能(取決於欲儲存之資料)已受到局部程式 化並可在程式化第一字線期間接收VpassH ^ 圖21顯不二字線接收VpassH之範例。更明確而言,連接 至選定字線(例如,WLO)之記憶體單元750經由該選定字線 在其控制閘極接收程式化電壓Vpgm。將增壓電壓Vpass施 加至°亥等未選疋5己憶體單元756、758、760、762及764之 控制閘極。铰向增壓電壓VpassH係經由字線WL1提供至記 隐體單元752之控制閘極並經由字線WL2提供至記憶體單 ,754。應注意,儘管圖21顯示兩個接收的記憶體 單元,因為該等二個記憶體單元由於該NAND串之最後抹 除程序而可能或可能未局部程式化,但其他具體實施例可 包括一個以上的接收VpassH之記憶體單元,因為該等記憶 體單元由於用於NAND串之最後抹除程序而可能或可能未 124908.doc -43. 1378456 局部程式化。 應注意在一個以上的字線接收較高增壓電壓之具體實施 例中,其不一定全部接收確切相同的VpassH。該等接收較 向增壓電壓之字線可接收不同的VpassH變化。在一具體實 施例中,各VpassH變化大於Vpass。 考量使用圖12A至C之程序來程式化一記憶體單元區塊 時的一範例。依據圖13之圖表,連接至WLO之該等記憶體 早元將會被局部程式化(參見圖12 A),使得其下部頁具有 資料。隨後’連接至WL1之該等記憶體單元將會被局部程 式化(參見圖12A),使得其下部頁具有資料。在此刻,由 於以前抹除該區塊(參見圖14之步驟640),連接至WLO及 WL1之該專6己憶體早元均未完成完全程式化。隨後,連接 至WL0之該等記憶體單元將會完成其程式化(參見圖 12C) ’使得其上部頁也具有資料。當完成用於WL〇之程式 化(例如’程式化上部頁)時’如圖15所示偏壓該等Nand 串之各種字線》該些概念可延伸至其他字線,如根據圖16 至20。此外’在程式化下部頁時可如圖15所示來偏壓該等 NAND串之各種字線。 出於例示及說明目的已呈現本發明之前述詳細說明。並 不希望徹底詳盡或將本發明限於所揭示的精確形式。在以 上教導的啟發下,可能有許多修改及變更.所述具體實施 例係選擇’以便清楚地說明本發明之原理及其實際應用, 從而使其他習知此項技術者能以各種具體實施例並採用適 合於所預期特定使用之各種修改來最佳地利用本發明。希 124908.doc -44- 1378456 望本發明之範疇由隨附申請專利範圍來定義。 【圖式簡單說明】 圖1係一 NAND串之一俯視圖。 圖2係NAND串之一等效電路圖。 圖3描述一 NAND串及在一程式化操作期間施加至該 NAND串之一組電壓。 圖4描述一 NAND串及A 一招n揭u 甲汉隹程式化刼作期間施加至該 NAND串之一組電壓。 圖5描述一 NAND串之一部分。 圖6係一非揮發性記憶體系統之一方塊圖。 圖7係一非揮發性記憶體系統之一方塊圖。 圖8係描述一感測區塊之一具體實施例之—方塊圖。 圖9係描述一記憶體陣列之一具體實施例之一方塊圖。 圖10描述一組範例性臨限電壓分佈並說明—用於程式化 非揮發性記憶體之程序。 圖11描述一組範例性臨限電壓分佈並說明—用於程式化 非揮發性記憶體之程序。 圖12A至C顯示各種臨限電壓分佈並說明一用於程式化 非揮發性記憶體之程序。 圖13係一描述在一具體實施例中程式化非揮發性記憶體 之次序之表格。 圖14係說明一用於程式化非揮發性記憶體之程序之一具 體實施例之一流程圖。 圖1 5描述一 NAND串及在一程式化操作期間施加至該 124908.doc •45- 1378456 NAND串之一組電墨。 圖1 6係解釋在一程式化操作期間特定信號之行為之一時 序圖。 圖1 7描述一 NAND串及在一程式化操作期間施加至該 NAND串之一組電壓。 圖1 8描述一 NAND串及在一程式化操作期間施加至該 NAND串之一組電壓。VpassH to the neighbor. In some embodiments, the system can partially program more than one memory early element of a NAND string prior to completing programming of a currently selected memory unit. For example, the stylized program of Figures 12A through C can be modified to perform a first pass/step for three word lines before continuing to complete the stylization on the first sub-line. In an example in which three data pages are stored in a memory unit, data is written in the following order: (1) writing the lower page data to WLn '(2) writing the lower page data to WLn+1, (3) Write intermediate page data to WLn, (4) Write lower page data to wLn+2, (5) Write intermediate page data to WLn+1, and (6) Write lower page data to WLn to complete Write all 3 pages to WLn. Other methods/schemes can also be used. In these examples, there are two word lines (depending on the data to be stored) that have been partially programmed and can receive VpassH during the staging of the first word line. Figure 21 shows an example of the VpassH. More specifically, memory unit 750 coupled to a selected word line (e.g., WLO) receives a stylized voltage Vpgm at its control gate via the selected word line. The boosting voltage Vpass is applied to the control gates of the unselected 己5 memory cells 756, 758, 760, 762 and 764. The hinge boost voltage VpassH is supplied to the control gate of the body unit 752 via the word line WL1 and to the memory bank 754 via the word line WL2. It should be noted that although FIG. 21 shows two received memory cells, because the two memory cells may or may not be partially programmed due to the final erase process of the NAND string, other embodiments may include more than one. The memory cells of the VpassH are received because the memory cells may or may not be partially programmed by 124908.doc -43. 1378456 due to the final erase program for the NAND strings. It should be noted that in a particular embodiment where more than one word line receives a higher boost voltage, it may not all receive exactly the same VpassH. The word lines that receive the relatively boosted voltage can receive different VpassH variations. In a specific embodiment, each VpassH change is greater than Vpass. Consider an example when using the program of Figures 12A through C to program a memory cell block. According to the graph of Fig. 13, the memories connected to the WLO will be partially programmed (see Fig. 12A) such that the lower page has data. These memory cells connected to WL1 will then be partially localized (see Figure 12A) such that the lower page has data. At this point, since the block was previously erased (see step 640 of Figure 14), the dedicated 6-message early connected to WLO and WL1 has not been fully programmed. Subsequently, the memory cells connected to WL0 will complete their stylization (see Figure 12C) so that their upper page also has data. When the stylization for WL〇 (eg, 'programming the upper page') is performed, the various word lines of the Nand strings are biased as shown in FIG. 15. These concepts can be extended to other word lines, as shown in FIG. 20. In addition, the various word lines of the NAND strings can be biased as shown in Figure 15 when the lower page is programmed. The foregoing detailed description of the invention has been presented for purposes of illustration It is not intended to be exhaustive or to limit the invention to the precise form disclosed. In the light of the above teachings, many modifications and variations are possible. The specific embodiments are chosen to clearly illustrate the principles of the present invention and its practical application. The invention is best utilized with various modifications that are suitable for the particular use contemplated.希 124908.doc -44- 1378456 The scope of the invention is defined by the scope of the accompanying patent application. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a top view of a NAND string. Figure 2 is an equivalent circuit diagram of one of the NAND strings. Figure 3 depicts a NAND string and a set of voltages applied to the NAND string during a stylized operation. Figure 4 depicts a NAND string and a set of voltages applied to the NAND string during a stylized operation. Figure 5 depicts a portion of a NAND string. Figure 6 is a block diagram of a non-volatile memory system. Figure 7 is a block diagram of a non-volatile memory system. Figure 8 is a block diagram depicting one embodiment of a sensing block. Figure 9 is a block diagram depicting one embodiment of a memory array. Figure 10 depicts an exemplary set of threshold voltage distributions and illustrates the procedure for programming non-volatile memory. Figure 11 depicts an exemplary set of threshold voltage distributions and illustrates the procedure for programming non-volatile memory. Figures 12A through C show various threshold voltage distributions and illustrate a procedure for programming non-volatile memory. Figure 13 is a table depicting the order in which non-volatile memory is programmed in a particular embodiment. Figure 14 is a flow chart showing a specific embodiment of a program for staging non-volatile memory. Figure 15 depicts a NAND string and a set of ink applied to the 124908.doc • 45-1378456 NAND string during a stylization operation. Figure 16 illustrates a timing diagram of the behavior of a particular signal during a stylized operation. Figure 17 depicts a NAND string and a set of voltages applied to the NAND string during a stylized operation. Figure 18 depicts a NAND string and a set of voltages applied to the NAND string during a stylized operation.
圖1 9描述一 NAND串及在一程式化操作期間施加至該 NAND串之一組電壓。 圖20描述一 NAND串及在一程式化操作期間施加至該 NAND串之一組電壓。 圖2 1描述一 NAND串及在一讀取操作期間施加至該 NAND串之一組電壓。Figure 19 depicts a NAND string and a set of voltages applied to the NAND string during a stylized operation. Figure 20 depicts a NAND string and a set of voltages applied to the NAND string during a stylized operation. Figure 21 depicts a NAND string and a set of voltages applied to the NAND string during a read operation.
【主要元件符號說明】 100 電晶體 100CG 控制閘極 100FG 浮動閘極 102 電晶體 102CG 控制閘極 102FG 浮動閘極 104 電晶體 104CG 控制閘極 104FG 浮動閘極 106 電晶體 124908.doc -46- 1378456[Main component symbol description] 100 transistor 100CG control gate 100FG floating gate 102 transistor 102CG control gate 102FG floating gate 104 transistor 104CG control gate 104FG floating gate 106 transistor 124908.doc -46- 1378456
106CG 控制閘極 106FG 浮動閘極 120 第一(或汲極)選擇閘極 122 第二(或源極)選擇閘極 126 位元線接點 128 源極線 304 記憶體單元 306 記憶體單元 308 記憶體單元 3 10 記憶體單元 312 記憶體單元 3 14 記憶體單元 316 記憶體單元 318 記憶體單元 322 源極選擇閘極 324 汲極側選擇閘極 330 源極/ >及極區 332 共同源極線 334 位元線接點 340 通道 350 記憶體單元 352 記憶體單元 354 記憶體單元 356 記憶體單元 124908.doc -47- 1378456 358 記憶體單元 360 記憶體單元 362 記憶體單元 364 記憶體單元 366 汲極選擇閘極/汲極選擇電晶體 368 源極側選擇閘極 370 源極/汲極區 374 位元線接點 380 區域/增壓通道 384 P井區域 396 記憶體裝置 398 記憶體晶粒 400 二維記憶體單元陣列 410 控制電路 412 狀態機 414 晶片上位址解碼器 416 功率控制模組 418 線 420 線 430 列解碼器 430A 列解碼器 430B 列解碼器 450 控制器 460 行解碼器 124908.doc -48- 1378456 行解碼器 行解碼器 讀取/寫入電路 讀取/寫入電路 讀取/寫入電路 感測電路 資料匯流排 感測模組 位元線鎖存器 共同部分 處理器106CG control gate 106FG floating gate 120 first (or drain) select gate 122 second (or source) select gate 126 bit line contact 128 source line 304 memory unit 306 memory unit 308 memory Body unit 3 10 memory unit 312 memory unit 3 14 memory unit 316 memory unit 318 memory unit 322 source select gate 324 drain side select gate 330 source / > and polar region 332 common source Line 334 bit line contact 340 channel 350 memory unit 352 memory unit 354 memory unit 356 memory unit 124908.doc -47- 1378456 358 memory unit 360 memory unit 362 memory unit 364 memory unit 366 汲Pole selection gate/drain selection transistor 368 source side selection gate 370 source/drain region 374 bit line contact 380 region/boost channel 384 P well region 396 memory device 398 memory die 400 Two-dimensional memory cell array 410 control circuit 412 state machine 414 on-chip address decoder 416 power control module 418 line 420 line 430 column decoder 430A Column Decoder 430B Column Decoder 450 Controller 460 Row Decoder 124908.doc -48- 1378456 Row Decoder Row Decoder Read/Write Circuit Read/Write Circuit Read/Write Circuit Sensing Circuit Data Busbar sensing module bit line latch common part processor
460A460A
460B 465460B 465
465A465A
465B 470 472 480 # 482 490 492 493 輸入線 494 496 500 530 532 534 550 750 752 754 756 758465B 470 472 480 # 482 490 492 493 Input line 494 496 500 530 532 534 550 750 752 754 756 758
資料鎖存器 I/O介面 感測區塊 箭頭 箭頭 箭頭 中間臨限電壓分佈/中間狀態 記憶體單元 記憶體單元 記憶體單元 記憶體單元 記憶體單元 124908.doc •49· 1378456Data Latch I/O Interface Sensing Block Arrow Arrow Arrow Intermediate Threshold Voltage Distribution / Intermediate Status Memory Unit Memory Unit Memory Unit Memory Unit Memory Unit 124908.doc •49· 1378456
760 762 764 766 768 770 772 774 _ 780 781 782 783 784 785 788 789 790 791 794 795 796 797 798 799 124908.doc 記憶體單元 記憶體單元 記憶體單元 汲極側選擇閘極 源極選擇閘極 源極/汲極區 共同源極線 位元線接點 增壓通道區域 區 較高增壓通道區域 區 較低增壓通道區域 區 較高增壓通道區域 區 較低增壓通道區域 區 較高增壓通道區域 區 較低增壓通道區域 區 中間增壓通道區域 -50-760 762 764 766 768 770 772 774 _ 780 781 782 783 784 785 788 789 790 791 794 795 796 797 798 799 129 124908.doc Memory unit memory unit memory unit drain side select gate source select gate source / bungee area common source line bit line contact point booster channel area higher boost channel area lower boost channel area higher boost channel area lower boost channel area higher boost Channel area area lower pressure channel area intermediate pressure channel area -50-
Claims (1)
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/535,634 US8184478B2 (en) | 2006-09-27 | 2006-09-27 | Apparatus with reduced program disturb in non-volatile storage |
US11/535,628 US8189378B2 (en) | 2006-09-27 | 2006-09-27 | Reducing program disturb in non-volatile storage |
Publications (2)
Publication Number | Publication Date |
---|---|
TW200822344A TW200822344A (en) | 2008-05-16 |
TWI378456B true TWI378456B (en) | 2012-12-01 |
Family
ID=39230870
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW96135141A TWI378456B (en) | 2006-09-27 | 2007-09-20 | Method and system for reducing program disturb in non-volatile storage |
Country Status (2)
Country | Link |
---|---|
TW (1) | TWI378456B (en) |
WO (1) | WO2008039667A2 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11495297B2 (en) | 2020-05-27 | 2022-11-08 | Windbond Electronics Corp. | Semiconductor device and reading method thereof |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7848144B2 (en) * | 2008-06-16 | 2010-12-07 | Sandisk Corporation | Reverse order page writing in flash memories |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR100272037B1 (en) * | 1997-02-27 | 2000-12-01 | 니시무로 타이죠 | Non volatile simiconductor memory |
KR100385230B1 (en) * | 2000-12-28 | 2003-05-27 | 삼성전자주식회사 | Method for programming a nonvolatile semiconductor memory device |
JP3957985B2 (en) * | 2001-03-06 | 2007-08-15 | 株式会社東芝 | Nonvolatile semiconductor memory device |
KR100502412B1 (en) * | 2002-10-23 | 2005-07-19 | 삼성전자주식회사 | Non-volatile semiconductor memory device and program method thereof |
US7466590B2 (en) * | 2004-02-06 | 2008-12-16 | Sandisk Corporation | Self-boosting method for flash memory cells |
US7161833B2 (en) * | 2004-02-06 | 2007-01-09 | Sandisk Corporation | Self-boosting system for flash memory cells |
US7170793B2 (en) * | 2004-04-13 | 2007-01-30 | Sandisk Corporation | Programming inhibit for non-volatile memory |
WO2007078793A1 (en) * | 2005-12-19 | 2007-07-12 | Sandisk Corporation | Method for programming non-volatile memory with reduced program disturb using modified pass voltages |
-
2007
- 2007-09-19 WO PCT/US2007/078842 patent/WO2008039667A2/en active Application Filing
- 2007-09-20 TW TW96135141A patent/TWI378456B/en not_active IP Right Cessation
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11495297B2 (en) | 2020-05-27 | 2022-11-08 | Windbond Electronics Corp. | Semiconductor device and reading method thereof |
Also Published As
Publication number | Publication date |
---|---|
WO2008039667A2 (en) | 2008-04-03 |
TW200822344A (en) | 2008-05-16 |
WO2008039667A3 (en) | 2008-06-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8270217B2 (en) | Apparatus for reducing the impact of program disturb | |
KR101519081B1 (en) | Enhanced bit-line pre-charge scheme for increasing channel boosting in non-volatile storage | |
TWI326487B (en) | Non-volatile storage system and method for programming non-volatile storage | |
US7436713B2 (en) | Reducing the impact of program disturb | |
US7515463B2 (en) | Reducing the impact of program disturb during read | |
TW201324513A (en) | Substrate bias during program of non-volatile storage | |
US8189378B2 (en) | Reducing program disturb in non-volatile storage | |
US7426137B2 (en) | Apparatus for reducing the impact of program disturb during read | |
JP4995264B2 (en) | Reduction of program disturbance during reading | |
TWI312155B (en) | Last-first mode, method and apparatus for programming of non-volatile memory with reduced program disturb | |
TW200841344A (en) | Methods and systems for controlled boosting in non-volatile memory soft programming | |
US8184478B2 (en) | Apparatus with reduced program disturb in non-volatile storage | |
TWI378456B (en) | Method and system for reducing program disturb in non-volatile storage | |
TWI334607B (en) | Method and apparatus for reducing the impact of program disturb | |
TWI352404B (en) | Providing local boosting control implant for non-v | |
KR20090089348A (en) | Reducing program disturb in non-volatile memory using multiple boosting modes |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
MM4A | Annulment or lapse of patent due to non-payment of fees |