TW200837760A - Systems utilizing variable program voltage increment values in non-volatile memory program operations - Google Patents

Systems utilizing variable program voltage increment values in non-volatile memory program operations Download PDF

Info

Publication number
TW200837760A
TW200837760A TW096137701A TW96137701A TW200837760A TW 200837760 A TW200837760 A TW 200837760A TW 096137701 A TW096137701 A TW 096137701A TW 96137701 A TW96137701 A TW 96137701A TW 200837760 A TW200837760 A TW 200837760A
Authority
TW
Taiwan
Prior art keywords
program voltage
state
programmed
voltage pulses
applying
Prior art date
Application number
TW096137701A
Other languages
Chinese (zh)
Other versions
TWI385665B (en
Inventor
Yan Li
Fang-Lin Zhang
Toru Miwa
Farookh Moogat
Original Assignee
Sandisk Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from US11/548,267 external-priority patent/US7450426B2/en
Priority claimed from US11/548,264 external-priority patent/US7474561B2/en
Application filed by Sandisk Corp filed Critical Sandisk Corp
Publication of TW200837760A publication Critical patent/TW200837760A/en
Application granted granted Critical
Publication of TWI385665B publication Critical patent/TWI385665B/en

Links

Classifications

    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/34Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/04Erasable programmable read-only memories electrically programmable using variable threshold transistors, e.g. FAMOS
    • G11C16/0483Erasable programmable read-only memories electrically programmable using variable threshold transistors, e.g. FAMOS comprising cells having several storage transistors connected in series
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C11/00Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C11/56Digital 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
    • G11C11/5621Digital 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
    • G11C11/5628Programming or writing circuits; Data input circuits
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/10Programming or data input circuits
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/10Programming or data input circuits
    • G11C16/12Programming voltage switching circuits
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/30Power supply circuits
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/34Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
    • G11C16/3436Arrangements for verifying correct programming or erasure
    • G11C16/3454Arrangements for verifying correct programming or for detecting overprogrammed cells
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/34Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
    • G11C16/3436Arrangements for verifying correct programming or erasure
    • G11C16/3454Arrangements for verifying correct programming or for detecting overprogrammed cells
    • G11C16/3459Circuits or methods to verify correct programming of nonvolatile memory cells
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/34Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
    • G11C16/3436Arrangements for verifying correct programming or erasure
    • G11C16/3468Prevention of overerasure or overprogramming, e.g. by verifying whilst erasing or writing
    • G11C16/3486Circuits or methods to prevent overprogramming of nonvolatile memory cells, e.g. by detecting onset or cessation of current flow in cells and using the detector output to terminate programming
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C16/00Erasable programmable read-only memories
    • G11C16/02Erasable programmable read-only memories electrically programmable
    • G11C16/06Auxiliary circuits, e.g. for writing into memory
    • G11C16/34Determination of programming status, e.g. threshold voltage, overprogramming or underprogramming, retention
    • G11C16/3418Disturbance prevention or evaluation; Refreshing of disturbed memory data
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C2211/00Indexing scheme relating to digital stores characterized by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C2211/56Indexing scheme relating to G11C11/56 and sub-groups for features not covered by these groups
    • G11C2211/562Multilevel memory programming aspects
    • G11C2211/5621Multilevel programming verification

Landscapes

  • Engineering & Computer Science (AREA)
  • Microelectronics & Electronic Packaging (AREA)
  • Computer Hardware Design (AREA)
  • Read Only Memory (AREA)

Abstract

The lowest programmed state in multi-state non-volatile flash memory devices can suffer from an increased level of bit line to bit line capacitive charge coupling when compared with other states. Program voltages applied to memory cells as increasing voltage pulses can be incremented using smaller values when programming memory cells to the lowest programmable state. Smaller increments in the applied voltage allow for greater precision and a narrower threshold voltage distribution to compensate for the disproportionate charge coupling experienced by cells programmed to this state. Smaller increment values can be used when switching from lower page to upper page programming in some implementations. In a pipelined programming architecture where cells forming a physical page store two logical pages of data and programming for one logical page begins before receiving data for the other logical page, the increment value can be increased when switching from programming the first logical page to programming both pages concurrently.

Description

200837760 九、發明說明: 【發明所屬之技術領域】 根據本發明之具體實施例係關於程式化非揮發性記憶 體。 【先前技術】 • 半導體記憶體裝置已變成愈來愈普遍運用在各種電子裝 • 置中。舉例而言,行動電話、數位攝影機、個人數位助 理、行動運算裝置、非行動運算裝置及其他裝置中皆使用 f 非揮發性半導體記憶體。電可擦除式可程式化唯讀記憶體 (Electrical Erasable Programmable Read Only Memory ; EEPROM)(包括快閃EEPROM)及電可程式化唯讀記憶體 (Electronically Programmable Read Only Memory ; EPROM)係最普遍的非揮發性半導體記憶體。200837760 IX. Description of the Invention: [Technical Field of the Invention] A specific embodiment according to the present invention relates to a stylized non-volatile memory. [Prior Art] • Semiconductor memory devices have become more and more widely used in various electronic devices. For example, non-volatile semiconductor memory is used in mobile phones, digital cameras, personal digital assistants, mobile computing devices, non-mobile computing devices, and other devices. Electrically Erasable Programmable Read Only Memory (EEPROM) (including flash EEPROM) and Electrically Programmable Read Only Memory (EPROM) are the most common Non-volatile semiconductor memory.

快閃記憶體系統之一項實例使用NAND結構,其包括夾 在兩個選擇閘極之間串聯排列的多個電晶體。串聯的該等 電晶體與該等選擇閘極被稱為一 NAND串。圖1繪示NANDOne example of a flash memory system uses a NAND structure that includes a plurality of transistors arranged in series between two select gates. The transistors in series and the select gates are referred to as a NAND string. Figure 1 shows NAND

C 串的俯視圖。圖2繪示其同等電路。圖1及2所示之該NAND 串包括夾在一第一選擇閘極120與一第二選擇閘極122之間 — 串聯的四個電晶體100、102、104與106。選擇閘極120連 接該NAND串至位元線126。選擇閘極122連接該NAND串 至源極線128。藉由將適當電壓經由選擇線SGD施加至控 制閘極120CG來控制選擇閘極120。藉由將適當電壓經由 選擇線SGS施加至控制閘極122CG來控制選擇閘極122。電 晶體100、102、104和106各包括一控制閘極及一浮動閘 125212.doc 200837760 極,形成一記憶體單元的閘極元件。舉例而言,電晶體 1〇〇具有控制閘極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包括控制閘極 106CG及浮動閘極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。 請注意,雖然圖1與圖2繪示出在該NAND串中有四個記 憶體單元,但是使用四個記憶體單元僅係作為一項實例予 以提供。一 NAND串可具有少於四個記憶體單元或多於四 個記憶體單元。舉例而言,一些NAND串將包括8個記憶體 單元、1 6個記憶體單元、32個記憶體單元等等。本文中之 論述未限定一 NAND串中的任何特定記憶體單元數量。 一種使用NAND結構之快閃記憶體系統的典型架構將包 括數個NAND串。舉例而言,圖3繪示一具有更多NAND串 之記憶體陣列的三個NAND串202、204和206。圖3所示之 該等NAND串中的每一 NAND串包括兩個選擇電晶體(或選 擇閘極)及四個記憶體單元。舉例而言,NAND串202包括 選擇電晶體220和230及記憶體單元222、224、226和228。 NAND串204包括選擇電晶體240和250及記憶體單元242、 244、246和248。每串係藉由一個選擇閘極(例如,選擇閘 極23 0與選擇閘極250)而連接至源極線。一選擇線SGS係用 於控制源極側選擇閘極。各種NAND串係藉由選擇閘極 125212.doc 200837760 220、240 (受控於選擇線SGD)等等而連接至各自位元線。 在其它具體實施例中,該等選擇線未必然成為共同線。字 線WL3被連接至記憶體單元222及記憶體單元242的控制閘 極。字線WL2被連接至記憶體單元224及記憶體單元244的 • ㈣閘極。字線WL1被連接至記憶體單元226及記憶體單Top view of the C string. Figure 2 shows its equivalent circuit. The NAND string shown in Figures 1 and 2 includes four transistors 100, 102, 104 and 106 sandwiched between a first select gate 120 and a second select gate 122. Select gate 120 connects the NAND string to bit line 126. Select gate 122 connects the NAND string to source line 128. The selection gate 120 is controlled by applying an appropriate voltage to the control gate 120CG via the select line SGD. The selection gate 122 is controlled by applying an appropriate voltage to the control gate 122CG via the selection line SGS. The transistors 100, 102, 104 and 106 each comprise a control gate and a floating gate 125212.doc 200837760 pole forming a gate element of a memory cell. For example, the transistor 1 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. Control gate 100CG is coupled to word line WL3, control gate 102CG is coupled to word line WL2, control gate 104CG is coupled to word line WL1, and control gate 106CG is coupled to word line WL0. Note that although Figures 1 and 2 illustrate that there are four memory cells in the NAND string, the use of four memory cells is provided as an example only. A NAND string can have fewer than four memory cells or more than four memory cells. For example, some NAND strings will include 8 memory cells, 16 memory cells, 32 memory cells, and the like. The discussion herein does not limit the number of any particular memory cells in a NAND string. A typical architecture for a flash memory system using a NAND structure would include several NAND strings. For example, Figure 3 illustrates three NAND strings 202, 204, and 206 having a memory array of more NAND strings. Each of the NAND strings shown in Figure 3 includes two select transistors (or select gates) and four memory cells. For example, NAND string 202 includes select transistors 220 and 230 and memory cells 222, 224, 226, and 228. NAND string 204 includes select transistors 240 and 250 and memory cells 242, 244, 246, and 248. Each string is connected to the source line by a select gate (e.g., select gate 23 0 and select gate 250). A select line SGS is used to control the source side select gate. The various NAND strings are connected to their respective bit lines by selecting gates 125212.doc 200837760 220, 240 (controlled by select line SGD) and the like. In other embodiments, the select lines are not necessarily to be common lines. Word line WL3 is connected to the control unit of memory unit 222 and memory unit 242. The word line WL2 is connected to the (four) gate of the memory unit 224 and the memory unit 244. Word line WL1 is connected to memory unit 226 and memory list

元246的控制閘極。字線WL〇被連接至記憶體軍元2以及記 ' 憶、體單元248的控制閘極。如所示,一位元線及各自NAND 广 串構成記憶體單元陣列之一行。該等字線(WL1、WL2、 WL3和WL4)構成陣列之列。每一字線連接該列中每一記 憶體單元的控制閘極。舉例而言,字線饥2被連接至記憶 體單元224、244及252的控制閘極。 母°己隱體單元可儲存資料(類比或數位)。當儲存一位 元之數位資料時,記憶體單元之可能的臨限電壓範圍被劃 分成經指派為邏輯資料"丨,,及"〇,,的兩段範圍。在nand型 >、]己L體之項實例中,記憶體單元被擦除之後的臨限 iy 電(為負且被疋義為邏輯"1"。程式化操作之後的臨限電 墨為正且被定義為邏輯,,『。當臨限電壓為負且嘗試施加〇 、彳工制閘極來進行讀取時,記憶體單元將開通以指示出 在儲存逯輯1 。當臨限電壓為正且嘗試施加〇伏至控制 .閘極來進行w貝取操作時,記憶體單元未開通,其指示出儲 存邏輯”0”。記憶體翠元亦可以儲存多位準資訊,舉例而 口夕位元數位資料。假使儲存多位準資料,則按資料位 準之數里來里j刀可能的臨限電壓範圍。舉例而言,如果儲 存四位準之貝訊’則有四段臨限電壓範圍被指派為資料值 125212.doc 200837760 ”11”、”10”、”01”及”00”。在NAND型記憶體之一項實例 中,擦除操作之後的臨限電壓為負且被定義為”丨丨,,。三個 不同正臨限電壓係用於狀態”10”、”01,,及,,〇〇,,。 以下美國專利案/專利申請案中提供NAND型快閃記情體 及其運作的相關實例,所有該等案均以引用方式併入本文 中·美國專利案第5,570,3 15號;美國專利案第5 774 397 號;美國專利案第6,046,935號;美國專利案第6,456,528 號;及美國專利申請案序號第〇9/893,277號(公告第us 2003/0002348號)。 當程式化EEPROM或快閃記憶體裝置時,典型地,施加 一程式電壓至控制閘極且使位元線接地。來自通道的電子 被注入至浮動閘極。當電子累積於浮動閘極中時,浮動閘 極變成荷載負電荷狀態,並且記憶體單元的臨限電壓上 升,使得記憶體單元係處於經程式化狀態。該記憶體單元 的浮動閘極電荷及臨限電壓可指示出一相對應於所儲存之 貧料的特定狀態。如需關於程式化之詳細資訊,請參閱 2003年3月5日申請之美國專利申請案第1〇/379,6〇8號標題 為"Self Boosting Technique";及 2003 年 7 月 29 日申請之美 國專利申請案第10/629,068號標題為”Detecting 〇ver Programmed Memory”,該等案整份内容以引用方式併入本 文中。 浮動閘極上儲存之表觀電荷(apparent charge)的偏移可 起因於基於相鄰浮動閘極中儲存之電荷的電場耦合而發 生。美國專利第5,867,429號中描述此浮動閘極至浮動閘極 125212.doc 200837760 箱合現象’該案整份内容以引 ’谷以%用方式併入本文中。目標浮 動閘極的相鄰浮動閘極可包括· 〜 j包括·位於相同位元線上的鄰近 浮動閘極;位於相同丰磕μ沾牛 — 一 U子線上的鄰近洋動閘極,·或位於鄰近 位元線與鄭近字線兩者上的、、聿 百的子動閘極,並且因此往對角線 方向彼此交叉。 ㈣閘極至浮動閘極輕合現象最顯著(然而非係專門)發 生於在不同時間已程式化之若干組相鄰記憶體單元之間。 Γ 相而言’第—記憶體單^被程式化以將-電荷位準加 至其净動閘極’其對應於一組資料。其後,一或多個相鄰 把憶體單元經程式化,以將一電荷位準加至其浮動閘極, ㈣應於-組資料。該等相鄰記憶體單元中之—或多者經 $式化之後’因為該等相鄰記憶體單元上的電荷耦合至該 第-記憶體單元之效應,所以讀取自該第一記憶體單元的 電^立準似乎不同於所程式化的電荷位準。來自相鄰記憶 體早几的耦合可使讀取中之表觀電荷位準偏移,其偏移量 足以導致錯誤讀取所儲存之資料。 ^因為在多狀態式裝置中的受允許之臨限電壓範圍與禁用 範圍(’I於表不相異記憶體狀態之相異臨限電壓範圍之間 :1巳圍)車乂乍於一兀式裝置(binary device),所以對於多狀 :以裝置較關切浮動閘極至浮動閘極耦合之效應。因此, /子動間極至浮動閘極麵合現象可導致記憶體單it自-受允 許臨限電壓範圍偏移至禁用範圍。 又 思著口己U體單元尺寸持續縮小,預期自然臨限電壓程 化一 t除刀佈歸因於短通道效應、較大之氧化物厚度 125212.doc 200837760 合比率變化及更大之通道摻雜物波動而增大,因而減小介 於相鄰狀態之間的可用分隔。與僅使用兩種狀態之二元式 記憶體相比,多狀態式記憶體之此效應更加顯著。另外, 介於子線之間的空間及介於位元線之間的空間之減小亦將 亦增大介於相鄰浮動閘極之間的耦合。 【發明内容】 在多狀態非揮發性快閃記憶體中,與其 低經程式狀態可遭遇到一增大位準之位元線至位元線電容 電荷耦合。當程式化記憶體單元至最低可程式位準或狀態 時,可隨著增大的電壓脈衝,使用較小值遞增經施加至記 憶體單元的程式電壓。在施加之電壓中的遞增量愈小,允 許精確度愈高且臨限電壓分佈愈窄,以補償經程式化至該 狀態之記憶體單元所遇到的不均衡電荷輕合。在一項具體 5施例中’當程式化一第一邏輯頁時,可使用較小之遞增 值,、以及當程式化其它頁時,可使用較大之遞增值。在管 線式私式化架構中(其中形成一實體頁之記憶體單元儲存 ㈣邏輯頁:資料,並且對-邏輯頁之程式化係在接收另 =頁之*料之前開始),當自程式化第—邏輯頁切換 至叫程式化該兩頁時,可增大遞增值。 在一項具體實施例中,袒也 # < 1 置之方法,其包括··接此= 非揮發性儲存裝 發性儲存元件之1 . &式化貝料至-組多狀態非揮 至該纽非揮加預先決㈣ 性儲存元件,·及:力 程式化該資料至該等非揮發 σ或多個額外程式電壓脈衝至該組非 125212.doc 200837760 揮發性儲存元件,以完成„料之程式化。施加預先決定 數量之程式電隸衝至該組多狀態非揮發性儲存元件可包 括以帛-遞增值來增大該等程式電麼脈衝之每一者之 一大小’直到抵達該預先決定數量。施加—或多個額外程 式電壓脈衝至該組多狀態非揮發㈣存元件可包括:以一 第一遞增值來增大該一或多個額外程式電壓脈衝之每一者 之一大小。Control gate of element 246. The word line WL 〇 is connected to the memory cell 2 and the control gate of the memory cell 248. As shown, a single bit line and a respective NAND wide string form one row of a memory cell array. The word lines (WL1, WL2, WL3, and WL4) form a matrix. Each word line connects to the control gate of each of the memory cells in the column. For example, word line hunger 2 is connected to the control gates of memory cells 224, 244, and 252. The parent-hidden unit can store data (analog or digital). When storing a one-digit digital data, the possible threshold voltage range of the memory unit is divided into two ranges that are assigned as logical data "丨,, and "〇,. In the nand type>, the case of the L-body, the threshold iy after the memory cell is erased (negative and derogated as logic "1". The threshold ink after the stylization operation Positive and defined as logic, "When the threshold voltage is negative and an attempt is made to apply a 〇, 彳 制 gate for reading, the memory unit will be turned on to indicate that the memory is being stored. When the voltage is positive and an attempt is made to apply the sag to the control gate, the memory cell is not turned on, indicating that the logic "0" is stored. The memory cell can also store multiple levels of information, for example, Evening digit data. If multiple levels of data are stored, the possible threshold voltage range can be obtained from the number of data. For example, if you store four quasi-before, there are four paragraphs. The voltage limit range is assigned as the data value 125212.doc 200837760 "11", "10", "01" and "00". In one example of the NAND type memory, the threshold voltage after the erase operation is negative and Is defined as "丨丨,,. Three different positive threshold voltages are used for "10", "01,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, U.S. Patent No. 5,570,315; U.S. Patent No. 5,774,397; U.S. Patent No. 6,046,935; U.S. Patent No. 6,456,528; and U.S. Patent Application Serial No. 9/893,277, US 2003/0002348). When programming an EEPROM or flash memory device, typically a program voltage is applied to the control gate and the bit line is grounded. The electrons from the channel are injected into the floating gate. When accumulating in the floating gate, the floating gate becomes a negative load state, and the threshold voltage of the memory cell rises, so that the memory cell is in a stylized state. The floating gate charge and threshold of the memory cell The voltage may indicate a particular state corresponding to the stored lean material. For more information on stylization, please refer to U.S. Patent Application Serial No. 1/379,6,8, filed on March 5, 2003. U.S. Patent Application Serial No. 10/629,068, filed on Jul. 29, 2003, entitled "Detecting 〇ver Programmed Memory", the entire contents of which are incorporated herein by reference. The offset of the apparent charge stored on the floating gate can be caused by the electric field coupling based on the charge stored in the adjacent floating gate. This floating gate to floating gate is described in U.S. Patent No. 5,867,429. 125212.doc 200837760 Boxing phenomenon 'The whole content of this case is incorporated herein by reference. The adjacent floating gates of the target floating gate may include: ~ j including: adjacent floating gates on the same bit line; adjacent oceanic gates located on the same Fengqi μ dim - a U sub-line, or located Near the bit line and the Zheng near word line, the hundreds of sub-gates, and thus cross each other in the diagonal direction. (d) The most significant (but not exclusive) phenomenon of gate-to-floating gates occurs between several sets of adjacent memory cells that have been programmed at different times. In contrast, the 'memory memory cell' is programmed to add the -charge level to its net-drive gate' which corresponds to a set of data. Thereafter, one or more adjacent memory cells are programmed to add a charge level to their floating gate, and (4) to the group data. After - or more of the adjacent memory cells are processed, the first memory is read because the charge on the adjacent memory cells is coupled to the first memory cell. The unit's electrical standard seems to be different from the programmed charge level. Coupling from an adjacent memory allows the apparent charge level in the read to be offset with an offset sufficient to cause erroneous reading of the stored data. ^Because the allowable threshold voltage range and the disable range in the multi-state device ('I differs between the different threshold voltage ranges of the memory state: 1巳) Binary device, so for multi-shape: the device is more concerned with the effect of floating gate to floating gate coupling. Therefore, the /-sub-pole-to-floating gate face-to-face phenomenon can cause the memory single-it-permitted threshold voltage range to be shifted to the disabled range. Also thinking about the size of the U-body unit continues to shrink, it is expected that the natural threshold voltage is a knives due to the short channel effect, the larger oxide thickness 125212.doc 200837760 ratio change and larger channel doping The debris increases and fluctuates, thus reducing the available separation between adjacent states. This effect of multi-state memory is more pronounced than binary memory using only two states. In addition, the space between the sub-lines and the reduction in space between the bit lines will also increase the coupling between adjacent floating gates. SUMMARY OF THE INVENTION In a multi-state non-volatile flash memory, a bit line-to-bit line capacitance charge coupling with an increased level can be encountered with its low programmed state. When the memory cell is programmed to the lowest programmable level or state, the program voltage applied to the memory cell can be incremented with a smaller value as the voltage pulse is increased. The smaller the amount of increase in the applied voltage, the higher the accuracy and the narrower the threshold voltage distribution to compensate for the unbalanced charge encounters encountered by the memory cells programmed into this state. In a specific example 5, when a first logical page is programmed, a smaller increment value can be used, and when other pages are programmed, a larger increment value can be used. In a pipelined privateization architecture (where the memory cells forming a physical page store (4) logical pages: data, and the stylization of the logical pages begins before receiving another = page), when self-programming When the first-logical page is switched to the two pages that are stylized, the increment value can be increased. In a specific embodiment, the method of 袒# < 1 includes: 接接 = non-volatile storage of the storage element 1 & formulaized batting to - group multi-state non-swing To the New Zealand to add pre-determined (four) storage elements, and: to programmatically program the data to the non-volatile σ or a plurality of additional program voltage pulses to the group of non-125212.doc 200837760 volatile storage elements to complete „ Styling the material. Applying a predetermined number of programs to the set of multi-state non-volatile storage elements may include increasing the size of each of the program's pulses by a 帛-increment value until arrival The predetermined number. Applying - or a plurality of additional program voltage pulses to the set of multi-state non-volatile (four) storage elements can include: increasing each of the one or more additional program voltage pulses by a first increment value One size.

在-項具體實施例中,提供_種程式化非揮發性儲存裝 置之方法。接收-第一組資料,f亥第一組資料經指定用於 儲存於非揮發性儲存裝置之一實體頁中。該第一組資料可 包括少於該實體頁可儲存之最大量資料。在—項具體實施 例中,該第一組資料形成於一下部邏輯頁資料。該第一組 貝料被程式化至該實體頁,其方式為使用一程式電壓訊號 來程式化該資料,該程式電壓訊號具有一峰值,當程式化 该資料至該實體頁時,以一第一遞增值來遞增該峰值。接 收一第二組資料,該第二組資料亦經指定用於儲存於該實 體頁中。可在開始程式化該第一組資料至該實體頁之後並 且在完成程式化該第一組資料至該實體頁之前,接收該第 二組資料。回應於在完成程式化該第一組資料之前接收該 第一組資料’停止或中斷程式化該第一組資料。停止程式 化該第一組資料之後,並行地程式化該第一組資料與該第 二組資料至該實體頁。並行地程式化可包括:使用一程式 電壓訊號來程式化該第一資料與該第二資料,該程式電壓 訊號具有一峰值,當程式化該第一資料與該第二資料至該 125212.doc -12- 200837760 只體頁卞以一第一遞增值來遞增該峰值。 且2項具體實施例中,提供一種非揮發性記憶體系統, 4括.硬數個儲存元件;複數個資料緩衝器,盆In a specific embodiment, a method of staging a non-volatile storage device is provided. Receiving - The first set of data, the first set of data is designated for storage in a physical page of a non-volatile storage device. The first set of data may include less than the maximum amount of data that can be stored on the physical page. In the specific embodiment, the first set of data is formed on the lower logical page data. The first set of bead material is programmed into the physical page by using a program voltage signal to program the data. The program voltage signal has a peak value. When the data is stylized to the physical page, the first An increment value is used to increment the peak. A second set of data is received, which is also designated for storage in the physical page. The second set of data may be received after beginning to program the first set of data to the physical page and before completing the stylization of the first set of data to the physical page. Responding to receiving the first set of data prior to completing the stylization of the first set of data 'stops or interrupts the stylized first set of data. After the programming of the first set of data is stopped, the first set of data and the second set of data are programmed in parallel to the physical page. The parallel programming may include: using a program voltage signal to program the first data and the second data, the program voltage signal having a peak value, when the first data and the second data are programmed to the 125212.doc -12- 200837760 Body page 递增 increments the peak with a first increment value. And in two specific embodiments, a non-volatile memory system is provided, 4 is a hard number of storage elements; a plurality of data buffers, basins

2存7°件通訊;及管理電路,其與該等緩衝器㈣存元件 官理電路接收待儲存於該等儲存㈣中的第一資 料,並且作出回應,提供該第一資料至該等緩衝器中之 一組緩衝器。該管理電路使用以—第—遞增值增大之複數 個程式電壓脈衝來程式化該第一資料至該等储存元件。該 官理電路在程式化該第―資料時接收待儲存於該等儲存元 =Μ二資料’並且作出喊,提供該第:資料至該等 &衝Ο之第二組緩衝ϋ。該管理電路停止程式化該第一 資料至該等儲存元件’並且開始使用以一第二遞增值增大 之複數個程式電壓脈衝來並行地程式化該第一資料與該第 二資料至該等儲存元件。 攸口兄月田、附圖及中請專利範圍將可明白所揭示技術之 具體實施例的其他特徵、態樣及目的。 【實施方式】 圖4繪示可用於實施本發明之一或多項具體實施例的快 閃:己憶體系統之一具體實施例的方塊圖。可使用其他系統 及實施方案。記憶體單元陣列3〇2受控於行控制電路3〇4、 列控制電路306、共同源極線控制電路310及ρ井控制電路 308。仃控制電路3〇4被連接至記憶體單元陣列的位元 線,用於:讀取記憶體單元中料的資料;在程式化操作 期間敎記憶體單元之狀態;以及控制位元線之電位位 125212.doc •13- 200837760 準,以促進或禁止程式化及擦除。列控制電路306被連接 至子線’用以:選擇其中一個字線;施加讀取電壓;結合 受控於行控制電路304的位元線電位位準來施加程式電 Μ ’以及施加擦除電壓。共同源極線控制電路3丨〇控制一 連接至纪憶體單元的共同源極線(圖5中標示為”共同源極 線)。ρ井控制電路308控制ρ井電壓。 口己i思體單元中儲存的資料係藉由行控制電路3 予以讀 出且係經由資料輸入/輸出緩衝器312而輸出至外部"〇線。 待儲存於記憶體單元中的資料係經由該等外部1/〇線而輸 入至該資料輸入/輸出緩衝器312,並且傳送至該行控制電 路3 04。該等1/〇線被連接至控制器3 j 8。 用於控制快閃記憶體裝置的命令資料被輸入至控制器 318。該命令資料將所要求的操作告知快閃記憶體。輸入 的命令被傳送至屬於控制電路315之部件的狀態機316。狀 悲機316控制行控制電路3〇4、列控制電路3〇6、共同源極 線控制電路310、P井控制電路308及資料輸入/輸出緩衝器 3 12。狀悲機3 16亦可輸出快閃記憶體的狀態資料,諸如,, 就緒/忙碌”(rEADY/busy)或”通過/失敗"(pass/fail)。 控制器318被連接至或可連接於—主㈣統,諸如個人 ~ t腦、數位攝影機或個人數位助理等等。控制器與起始命 7的主機通訊,諸如儲存資料至記憶體陣列或從記憶 體哮列3〇2讀取資料,以及提供或接收此等資料。控制器 3 18將此等命令轉換成可藉由命令電路314 (其屬於控制電 路315之部件)解譯及執行的命令訊號。命令電路314係與 125212.doc -14- 200837760 狀態機316通訊。控制器318典型包括緩衝器記憶體,用於 寫入至記憶體陣列或自記憶體陣列讀取的使用者資料。 一項示範性記憶體系統包括一個積體電路(其包括控制 器3 18)及一或多個積體電路晶片(每一積體電路晶片包含一 記憶體陣列及相關聯之控制、輸入/輸出及狀態機電路)。 一項趨勢係在一或多個積體電路晶片上將一系統的記憶體 陣列及控制器電路整合在一起。記憶體系統可被嵌入為主 機系統的部件,或可被包括於一以可卸除方式插入至主機 系統中的§己憶卡(或其他封裝)中。此一記憶卡可包括整個 記憶體系統(例如,包括控制器),或僅包括記憶體陣列與 相關聯之周邊電路(連同嵌入於主機中的控制器或控制功 旎)。因此,控制器可被嵌入為主機中或被包括於可卸除 式s己憶體系統内。 °月參閱圖5 ’描述記憶體單元陣列302之示範性結構。作 為項貝例’描述一種被分割成1,〇24個區塊的NAND快閃 EEPROM。可以同時擦除每一區塊中儲存的資料。在一項 具體實她例中,區塊係被同時擦除之記憶體單元的最小單 位。在此實例中,每一區塊有8,5 12行。每一區塊典型被 ^分成若干頁(可能係一程式化單位)。其他程式化之資料 °°亦係可實行且列入考量。在一項具體實施例中,個別 頁可被刀成若干節段(segment),並且節段可包含作為一 土本私式化操作而一次寫入的最少數量之記憶體單元。一 或夕頁貪料典型被儲存於一列記憶體單元中。 |^| 5 0|Tr — 不只例之每一區塊中有8,5 12行,其被劃分成偶 125212.doc -15· 200837760 數行及奇數行。位元線被劃分成偶數位元線(Β^)及奇數 位元線(BLo)。在一種奇數/偶數位元線架構中,對沿一共 同字線且連接至奇數位元線的記憶體翠元進行一次程式 化,並且對沿一共同字線且連接至偶數位元線的記憶體單 元進行另-次程式化。圖5繪示串聯連接以形成一 ΝΑ·串 的四個記憶體單元。雖然圖中繪示每_ Ναν〇串中包括四 個記憶體單元’但是可以使用四個以上或以下記憶體單元 (例如H或其他數量)。NAND串的—終端係經由一 第一選擇電晶體或閘極(其連接至選擇間極沒極線SGD)而 連接至-相對應之位元線’並且另一終端係經由—第二選 擇電晶體(其連接至選擇閘極源極線SGS)而連接至一共同 源極線。 在其它具體實施財,該等位元線未被劃分成偶數及奇 數位元線。此類架構通常稱為"全位元線架構"。在一種全 位元線架構中’於讀取及程式化操作期間,同時選擇一2 塊的所有位元線。沿-共同字線且連接至任何位元線的記 憶體單元被同時程式化。 於-具體實施例之讀取及程式化操作期間,同時選擇 4’256個記憶體單元。該等所選記憶體單元具有相同的字 線(例如’WL2-0及同一種位元線(例如,偶數位元線)。因 此,可同時讀取或程式化532個位元組資料。該 取或程式化的532個位元組資料形成一 、咳 ^ ^ 邏輯頁。因此,在 此實例中’-個區塊可儲存至少八頁。當每一記憶體單元 儲存兩個位元的資料時(例如,—種多位準記憶體單元), 125212.doc * 16 - 200837760 一個區塊儲存16頁。具體實施例也可使用其他大小之區塊 及頁另外亦可使用除圖4及5以外的架構來實施具體實 施例。 擦除記憶體單元之方式為··使p井上升至一擦除電壓(例 , 如,20伏),並且使一所選區塊的字線接地。源極線及位 元線係處於浮動狀態。可對整個記憶體陣列、分開的區塊 或其他記憶體單元單位來執行擦除。電子從浮動閘極轉移 (Λ 至Ρ井區,並且臨限電壓變成負(在一具體實施例中)。 在讀取與驗證操作中,使得一所選區塊的選擇閘極上升 至一或多個選擇電壓,並且使得該所選區塊的非所選字線 (例如,WL0、WL1和WL3)上升至一讀取傳送電壓(例如, 4.5伏),以使電晶體運作為傳送閘極。該所選區塊的所選 字線(例如,WL2)被連接至一參考電壓,對於每一讀取與 驗證操作來指定該參考電壓的位準,以判定所涉及的記憶 體單元的臨限電壓是否高於或低於該參考電壓的位準。舉 例而言,在一位元記憶體單元的讀取操作中,該所選字線 WL2被接地,使得以偵測其臨限電壓是否高於〇伏。在一 位元记憶體單元的驗證操作中,該所選字線WL2被連接至 (例如)0.8伏,使得以在程式化進行過程中偵測其臨限電 - 壓是否已到達0.8伏。於讀取與驗證期間,源極與p井為零 伏。所選位元線(BLe)被預充電至(例如)〇7伏之位準。如 果臨限電壓高於讀取或驗證位準,則因為相關聯之非傳導 狀態記憶體單元,所以涉及之位元線(BLe)的電位位準維 持高位準。另一方面,如果臨限電壓低於讀取或驗證位 125212.doc -17- 200837760 準,則因為傳導狀態之記憶體單元,所以涉及之位元線 (BLe)的電位位準減小至低位準,例如,小於0.5伏。藉由 連接至位元線的感測放大器來偵測記憶體單元之狀態,並 且感測所得之位元線電壓。記憶體單元是否被程式化或擦 除之間的差異取決於淨負電荷是否被儲存於浮動閘極中。 舉例而言,如果負電荷被儲存於浮動閘極中,則臨限電壓 變成較高且電晶體可能處於增強操作模式。 在一項實例中,當程式化一記憶體單元時,汲極及p井 接收0伏,而控制閘極接收量值遞增的一連串程式化脈 衝。在一項具體實施例中,該連串脈衝的脈衝量值在Μ伏 至24伏範圍β。在其它具體實施針,f亥連串脈衝的脈衝 範圍可能不同’舉例而言,具有一高於12伏的開始位準。 於記憶體單元之程式化期間’在介於程式化脈衝之間的週 期中實行驗證操作。即,在每一程式化脈衝之間讀取正被 並行程式化之一記憶體單元群組中之每一記憶體單元的程 式化位準,以判定記憶體單元是否已到達或超過其正被程 式化時施加至其的驗證位準。驗證程式化的一項手段係測 試一特定比較點的傳導。經驗證已被充分程式化之記憶體 單元被鎖定,舉例而言,在NAND記憶體單元中,鎖定方 式係對於所有後續程式化脈衝,使位元線從〇上升至Vdd (例如,2·5伏)’以終止彼等記憶體單元的程式化處理程 序。在一些情況下,將限制脈衝數量(例如,2〇個脈衝), 並且如果最後一個脈衝未充分程式化一既定記憶體單元, 則假設-錯誤。在-些實施方案中,在程式化之前先擦除 125212.doc 18 200837760 記憶體單元(以區塊為單位或其他單位)。 圖6繪示根據一具體實族 餸貝施例之程式化電壓訊號。此訊號 具有含增大量值的一組脈衝。該等脈衝之量值係按一預先 決定步進大小而隨每-脈㈣以遞增。在—項包含館存多 位元資料的記憶體單⑽具體實施例中…示範性步進大 小為〇·2伏(或〇·4伏)。介於每一程式化脈衝之間者係驗證 脈衝。圖6之訊號假設—種四狀態式記憶體單元,因此, 該訊號包括三個驗證脈衝。舉例而言,介於程式化脈衝 330與332之間係三個連續的驗證脈衝。所繪示之一第一驗 證脈衝334係處於零伏驗證電壓位準。接在第—驗證脈衝 後的第二驗證脈衝336係處於第二驗證電壓位準。接在第 二驗證脈衝後的第三驗證脈衝338係處於第三驗證電壓位 準 種肖b夠以八種狀悲來儲存資料的多狀態記憶體單元 可能必須在七個比較點處實行驗證操作。因此,在兩個連 貝程式化脈衝之間依序施加七個驗證脈衝,以在七個驗證 C, 位準實行驗證操作。系統可依據七個驗證操作來決定記憶 體單元的狀態。-項減小驗證時間貞荷的手段係使用更高 效率的驗證處理程序,舉例而言,如下列專利申請案所揭 不· 2002年12月5日提出之專利申請案第1〇/314,〇55號標題 為"Smart Verify for Multi_State Memories" ; 2005年 1〇月 27 曰提出之專利申請案第11/259,799號標題為”Apparatus f〇r Programming of Multi-State Non-Volatile Memory Using Smart Verify” ;以及2005年10月27曰提出之專利申請案第 11/260,658 號標題為’’Method for Programming 〇f Multi- 125212.doc -19· 2008377602 storing 7° pieces of communication; and managing circuitry, and receiving, by the buffer (4) component management circuit, the first data to be stored in the storage (4), and responding, providing the first data to the buffers One of the group buffers in the device. The management circuit programs the first data to the storage elements using a plurality of program voltage pulses that are incremented by a -th increment value. The official circuit receives the data to be stored in the storage element when the first data is programmed and makes a call to provide the second data to the second set of buffers of the & The management circuit stops programming the first data to the storage elements and begins to program the first data and the second data in parallel using a plurality of program voltage pulses that are incremented by a second increment value Store components. Other features, aspects, and objectives of the specific embodiments of the disclosed technology will be apparent from the scope of the appended claims. [Embodiment] FIG. 4 is a block diagram showing one embodiment of a flash:recognition system that can be used to implement one or more embodiments of the present invention. Other systems and implementations can be used. The memory cell array 3〇2 is controlled by a row control circuit 〇4, a column control circuit 306, a common source line control circuit 310, and a ρ well control circuit 308. The 仃 control circuit 〇4 is connected to the bit line of the memory cell array for: reading data of the material in the memory unit; 敎 the state of the memory cell during the stylizing operation; and controlling the potential of the bit line Bit 125212.doc •13- 200837760 is intended to facilitate or prohibit stylization and erasure. Column control circuit 306 is coupled to the sub-wire ' to: select one of the word lines; apply a read voltage; apply a program voltage 结合 in conjunction with a bit line potential level controlled by row control circuit 304 and apply an erase voltage . The common source line control circuit 3 丨〇 controls a common source line (labeled as "common source line" in Fig. 5) connected to the memory unit. The ρ well control circuit 308 controls the ρ well voltage. The data stored in the unit is read by the row control circuit 3 and output to the external "〇 line via the data input/output buffer 312. The data to be stored in the memory unit is via the external 1/ The line is input to the data input/output buffer 312, and is transmitted to the line control circuit 304. The 1/〇 lines are connected to the controller 3 j 8. Command data for controlling the flash memory device It is input to the controller 318. The command data informs the flash memory of the required operation. The input command is transmitted to the state machine 316 belonging to the components of the control circuit 315. The sorcerer 316 controls the row control circuit 3〇4, Column control circuit 〇6, common source line control circuit 310, P-well control circuit 308, and data input/output buffer 3 12. The sorcerer 3 16 can also output status data of the flash memory, such as, ready /busy" (rEADY/busy) or "pass/fail". The controller 318 is connected to or connectable to the main (four) system, such as a personal ~ t brain, a digital camera or a personal digital assistant, etc. The controller and the starting life 7 Host communication, such as storing data to or reading data from the memory array, and providing or receiving such data. The controller 3 18 converts the commands into commands via the command circuit 314 (which belongs to The command signal is interpreted and executed by the control circuit 315. The command circuit 314 is in communication with the state machine 316 of 125212.doc -14-200837760. The controller 318 typically includes a buffer memory for writing to the memory array or User data read from the memory array. An exemplary memory system includes an integrated circuit (which includes controller 3 18) and one or more integrated circuit chips (each integrated circuit chip contains a memory) Body arrays and associated control, input/output, and state machine circuits. One trend is to integrate a system of memory arrays and controller circuits on one or more integrated circuit chips. The system can be embedded as a component of the host system or can be included in a removable memory card (or other package) that can be removably inserted into the host system. This memory card can include the entire memory system (eg, , including the controller), or only the memory array and associated peripheral circuits (along with controllers or control functions embedded in the host). Therefore, the controller can be embedded in the host or included in the removable Within the system of the memory system, see FIG. 5' to describe an exemplary structure of the memory cell array 302. As an item, a NAND flash EEPROM divided into 1, 24 blocks is described. The data stored in each block can be erased at the same time. In a specific example, the block is the smallest unit of memory cells that are simultaneously erased. In this example, each block has 8, 5 12 rows. Each block is typically divided into several pages (possibly a stylized unit). Other stylized information °° is also practicable and considered. In a specific embodiment, individual pages can be knifed into segments, and the segments can include a minimum number of memory cells that are written at a time as a localized private operation. One or the eve page is typically stored in a list of memory cells. |^| 5 0|Tr — There are not only 8 and 5 12 lines in each block, but it is divided into even 125212.doc -15· 200837760 lines and odd lines. The bit lines are divided into even bit lines (Β^) and odd bit lines (BLo). In an odd/even bit line architecture, a memory of a memory element along a common word line connected to an odd bit line is performed, and a memory along a common word line and connected to an even bit line is performed. The body unit is further stylized. Figure 5 illustrates four memory cells connected in series to form a ΝΑ string. Although four memory cells are included in each _ Ναν〇 string, four or more memory cells (e.g., H or other numbers) may be used. The terminal of the NAND string is connected to the corresponding bit line ' via a first selection transistor or gate (which is connected to the selected interpole line SGD) and the other terminal is via the second selection A crystal (which is connected to the select gate source line SGS) is connected to a common source line. In other implementations, the bit lines are not divided into even and odd bit lines. This type of architecture is often referred to as the "full bit line architecture". In a full bit line architecture, all bit lines of a block are selected simultaneously during read and program operations. Memory cells along the - common word line and connected to any bit line are simultaneously programmed. During the read and stylization operations of the particular embodiment, 4'256 memory cells are simultaneously selected. The selected memory cells have the same word line (eg, 'WL2-0 and the same bit line (eg, even bit line). Therefore, 532 bytes of data can be read or programmed simultaneously. The 532 bytes of data obtained or stylized form a cough ^ ^ logical page. Therefore, in this example, '-blocks can store at least eight pages. When each memory unit stores two bits of data Time (for example, a multi-level memory unit), 125212.doc * 16 - 200837760 One block stores 16 pages. The specific embodiment can also use other sizes of blocks and pages can also be used in addition to Figures 4 and 5 The embodiment is implemented by a different architecture. The way to erase the memory cell is to raise the p-well to an erase voltage (eg, 20 volts) and ground the word line of a selected block. The line and bit line are in a floating state. Erasing can be performed on the entire memory array, separate blocks, or other memory unit units. Electrons are transferred from the floating gate (Λ to the Sakai area, and the threshold voltage becomes Negative (in a specific embodiment). In the verifying operation, the select gate of a selected block is raised to one or more select voltages, and the unselected word lines (eg, WL0, WL1, and WL3) of the selected block are raised to a read transfer voltage. (eg, 4.5 volts) to operate the transistor as a transfer gate. The selected word line (eg, WL2) of the selected block is connected to a reference voltage, which is specified for each read and verify operation The level of the voltage to determine whether the threshold voltage of the memory cell involved is higher or lower than the level of the reference voltage. For example, in the read operation of the one-bit memory cell, the selected The word line WL2 is grounded to detect whether its threshold voltage is higher than the stagnation. In the verify operation of the one-bit memory cell, the selected word line WL2 is connected to, for example, 0.8 volts, so that During the stylization process, it is detected whether the threshold voltage has reached 0.8 volts. During reading and verification, the source and p wells are zero volts. The selected bit line (BLe) is precharged (eg ) 〇 7 volts. If the threshold voltage is higher than the read or verify level Because of the associated non-conducting state memory cell, the potential level of the bit line (BLe) involved is maintained at a high level. On the other hand, if the threshold voltage is lower than the read or verify bit 125212.doc -17 - 200837760 准, because of the conduction state of the memory cell, the potential level of the bit line (BLe) involved is reduced to a low level, for example, less than 0.5 volts. By means of a sense amplifier connected to the bit line Detecting the state of the memory cell and sensing the resulting bit line voltage. The difference between whether the memory cell is programmed or erased depends on whether the net negative charge is stored in the floating gate. If a negative charge is stored in the floating gate, the threshold voltage becomes higher and the transistor may be in an enhanced mode of operation. In one example, when staging a memory cell, the drain and p wells receive 0 volts, while the gate receives a series of stylized pulses of increasing magnitude. In a specific embodiment, the pulse magnitude of the series of pulses is in the range of Μ to 24 volts. In other embodiments, the pulse range of the pulse may be different ', for example, having a starting level above 12 volts. During the stylization of the memory unit, the verify operation is performed during the period between the stylized pulses. That is, the programmed level of each memory cell in a group of memory cells being serialized in parallel is read between each stylized pulse to determine if the memory cell has reached or exceeded The level of verification applied to it when stylized. One means of verifying stylization is to test the conduction of a particular comparison point. Memory cells that have been verified to be fully programmed are locked. For example, in a NAND memory cell, the lock mode is such that for all subsequent stylized pulses, the bit line rises from 〇 to Vdd (eg, 2·5)伏)' to terminate the stylized processing of their memory cells. In some cases, the number of pulses (e.g., 2 pulses) will be limited, and if the last pulse does not adequately program a given memory cell, then an error is assumed. In some embodiments, the 125212.doc 18 200837760 memory unit (in blocks or other units) is erased prior to stylization. Figure 6 illustrates a stylized voltage signal according to a specific real family of mussels. This signal has a set of pulses with increasing magnitudes. The magnitude of the pulses is incremented by a predetermined step size with each pulse (four). In the embodiment of the memory list (10) containing the multi-bit data of the museum, the exemplary step size is 〇·2 volts (or 〇·4 volts). The verification pulse is between each stylized pulse. The signal of Figure 6 assumes a four-state memory unit, so the signal includes three verify pulses. For example, there are three consecutive verify pulses between the stylized pulses 330 and 332. One of the first verification pulses 334 is shown to be at a zero volt verify voltage level. The second verify pulse 336 following the first-verify pulse is at the second verify voltage level. The third verification pulse 338 connected to the second verification pulse is at the third verification voltage level. The multi-state memory unit capable of storing data in eight kinds of sorrows may have to perform verification operations at seven comparison points. . Therefore, seven verify pulses are sequentially applied between the two consecutive stylized pulses to perform the verify operation at the seven verification C levels. The system can determine the state of the memory unit based on seven verification operations. - The means of reducing the verification time load is to use a more efficient verification process, for example, as disclosed in the following patent application, Patent Application No. 1/314, filed on December 5, 2002, 〇55 titled "Smart Verify for Multi_State Memories"; Patent Application No. 11/259,799, filed on January 27, 2005, entitled "Apparatus f〇r Programming of Multi-State Non-Volatile Memory Using Smart Verify ; and Patent Application No. 11/260,658, filed October 27, 2005, entitled ''Method for Programming 〇f Multi- 125212.doc -19· 200837760

State Non_v〇iatile Mem〇ry using Smart Verify”,該等案整 份内容均以引用方式併入本文。 上文所述之擦除、磧取與驗證操作係根據此項技術中熟 知的技術予以實行。因此,熟悉此項技術者可改變所解說 的許多細節。 右適用,在成功程式化處理程序結束時,記憶體單元的 阳電C應在經私式化之記憶體單元的一或多項臨限電壓State Non_v〇iatile Mem〇ry using Smart Verify, the entire contents of which are incorporated herein by reference. The erasing, capture and verification operations described above are carried out according to techniques well known in the art. Therefore, those skilled in the art can change many of the details explained. Right, at the end of the successful stylization process, the memory unit's solar C should be in one or more of the private memory cells. Voltage limit

刀佈内或在經擦除之記憶體單元的一臨限電壓分佈内。圖 7繪示當每一記憶體單元儲存兩個位元之資料時記憶體單 凡陣列的fe限電壓分佈。圖7繪示經擦除之記憶體單元的 第一臨限電壓分佈E。亦描繪出經程式化之記憶體單元的 三種臨限電壓分佈A、BW。在一項具體實施例中,£分 佈中的臨限電壓係負值,A、B*c分佈中的臨限電壓係正 值。 圖7之母一相異臨限電壓範圍對應於一用於各組資料位 元的預S決定值。介於程式化於記㈣單元巾之資料與記 憶體單it之臨限電壓位準之間的特定關係取決於記憶體單 元所採用的資料編碼方案。在—項具體實施例中,使用一 ㈣雷碼(Gray code)指派,將資料值指派給該等臨限電壓 耗圍,使付如果一浮動閘極的臨限電壓錯誤地偏移至其鄰 近實體狀態,則僅一個位元將受到影響。然而,在其他具 體實施例中,不使用格雪紙 粍田編碼法。一項實例指派”11”給臨 限電壓範圍E (狀離E) ·和,, 1 (,礼派10,,給臨限電壓範圍A (狀離 A);指派”00"給臨限電壓範圍B (狀態B);及指派"〇i "㈣ 125212.doc -20- 200837760 限電壓範圍c (狀態c)。雖然圖7繪示四種狀態,但是亦可 配合其他多狀態結構(包括具有四種以上或以上狀態之多 狀態結構)運用本發明。 圖7亦繪示用於從記憶體單元讀取資料的三個讀取參考 電壓VrA、VrB與VrC。藉由測試一既定記憶體單元的臨限電 . 壓是否高於或低於Vm、VrB與VrC,系統可判定該記憶體單 - 元所處之狀態。圖7亦繪示三個驗證參考電壓VvA、V B與 當將記憶體單元程式化至狀態A時,系統將測試記憶 體單70是否具有大於或等於VvA之臨限電壓。當將記憶體 單兀程式化至狀態B時,系統將測試記憶體單元是否具有 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,系統將判定記憶體單元是否具有大於或等於νπ之 臨限電壓。 圖7繪示利用全序列程式化之一具體實施例。在全序列 程式化中,可將記憶體單元從經擦除狀態E直接程式化至 該等經程式化狀態A、Β或C中之任一狀態。舉例而言,待 程式化的一群體記憶體單元可先予以擦除,使得該群體中 的所有記憶體單元皆處於經擦除狀態E。接著,使用下文 • 參閱圖7描述之處理程序(使用一連串程式電壓脈衝施加至 „ 所選記憶體單元之控制閘極),將記憶體單元直接程式化 至狀態A、B或C。當一些記憶體單元正被從狀態E程式化 至狀態A時,其他記憶體單元正被從狀態E程式化至狀態β 及/或從狀態E程式化至狀態C。 圖8繪示一種程式化多狀態式記憶體單元之兩次進程 125212.doc -21 - 200837760 (two-pass)技術之實例,其儲存兩個不同頁(一下部頁與一 上部頁)的資料。圖中繪示四種狀態:狀態E (11)、狀態A (10)、狀態B (00)及狀態C (01)。對於狀態£,彼兩頁儲存 ’· 1π。對於狀態A,下部頁儲存”〇”且上部頁儲存,,丨,,。對於 狀態B,彼兩頁儲存”〇”。對於狀態c,下部頁儲存,,1,,且上 部頁儲存”〇”。請注意,雖然特定位元型樣(bit pattern)已 被指派給每一狀態,但是可指派不同的位元型樣。在第一 次程式化進程中,按照待程式化至下部邏輯頁中的位元來 設定記憶體單元的臨限電壓位準。如果該位元係一邏輯 Π1",則由於已在早先予以擦除而處於適當狀態,所以未 使臨限電壓變化。但是,如果待程式化之位元係一邏輯 〇 ,則圮憶體單元之臨限電壓位準被增大至狀態A,如箭 頭302所示。這使第一程式化進程終止。 在第二次程式化進程中,按照正被程式化至上部邏輯頁 中的位元來設定記憶體單元的臨限電壓位準。如果該上部 邏輯頁位元係儲存一邏輯”1”,則由於該記憶體單元係處 於狀態E或A (取決於該下部頁位元之程式化),彼兩種狀 態皆載有上部頁位元”1”,所以未發生程式化。如果該上 部頁位元係邏輯”0”,則使臨限電壓偏移。如果第一進程 導致該記憶體單元維持在經擦除狀態E,則在第二階段 中,該記憶體單7L被程式化,使得臨限電壓被增大至狀熊 C範圍内,如箭頭306所示。如果第一程式化進程導致該記 體單元已被私式化為狀態a,則在第二進程中進一步程 式化該記憶體單元,使得臨限電壓被增大至狀態B範圍 125212.doc -22- 200837760 内,如箭頭304所示。第二進程的結果係將記憶體單元程 式化為經指定用以使上部頁儲存邏輯之狀態,而且未 變更下部頁之資料。Within a knife cloth or within a threshold voltage distribution of the erased memory cell. Figure 7 shows the fe limit voltage distribution of the memory array when each memory cell stores two bits of data. Figure 7 illustrates a first threshold voltage distribution E of the erased memory cell. Three threshold voltage distributions A, BW of the stylized memory unit are also depicted. In a specific embodiment, the threshold voltage in the distribution is negative, and the threshold voltage in the A, B*c distribution is positive. The parent-individual threshold voltage range of Figure 7 corresponds to a pre-S decision value for each set of data bits. The specific relationship between the data of the stylized (4) unit towel and the threshold voltage level of the memory unit depends on the data encoding scheme used by the memory unit. In a specific embodiment, a (four) Gray code assignment is used to assign data values to the threshold voltage limits such that if a threshold voltage of a floating gate is erroneously offset to its vicinity Entity status, only one bit will be affected. However, in other specific embodiments, the snow paper coding method is not used. An example assigns "11" to the threshold voltage range E (distance from E) · and, 1 (, ritual 10, to the threshold voltage range A (displaced from A); assign "00" to the threshold voltage Range B (state B); and assignment "〇i " (4) 125212.doc -20- 200837760 Limit voltage range c (state c). Although Figure 7 shows four states, it can also be combined with other multi-state structures ( The invention is applied to a multi-state structure having four or more states. Figure 7 also shows three read reference voltages VrA, VrB and VrC for reading data from a memory cell. By testing a given memory The threshold of the body unit. Whether the voltage is higher or lower than Vm, VrB and VrC, the system can determine the state of the memory unit. Figure 7 also shows three verification reference voltages VvA, VB and When the memory unit is programmed to state A, the system will test whether the memory unit 70 has a threshold voltage greater than or equal to VvA. When the memory unit is programmed to state B, the system will test whether the memory unit has greater than Or equal to the threshold voltage of VvB. When the memory unit is programmed to state c, The system will determine whether the memory cell has a threshold voltage greater than or equal to νπ. Figure 7 illustrates one embodiment using full sequence stylization. In full sequence programming, the memory cell can be erased from state E. Directly stylized to any of the stylized states A, Β, or C. For example, a population of memory cells to be programmed may be erased first, such that all memory cells in the population are In the erased state E. Next, use the processing procedure described below with reference to Figure 7 (using a series of program voltage pulses applied to the control gate of the selected memory cell) to program the memory cell directly to state A, B or C. When some memory cells are being programmed from state E to state A, other memory cells are being programmed from state E to state β and/or from state E to state C. Figure 8 An example of a two-step process of a stylized multi-state memory unit 125212.doc -21 - 200837760 (two-pass), which stores data for two different pages (the lower page and the upper page). Paint Four states: state E (11), state A (10), state B (00), and state C (01). For state £, the two pages store '·1π. For state A, the lower page stores 〇' And the upper page stores, 丨,,. For state B, the two pages store "〇". For state c, the lower page stores, 1, and the upper page stores "〇". Note that although the specific bit type A bit pattern has been assigned to each state, but different bit patterns can be assigned. In the first stylization process, the memory cells are set according to the bits to be programmed into the lower logical page. Threshold voltage level. If the bit is a logical Π1", the threshold voltage is not changed because it has been properly erased earlier. However, if the bit to be programmed is a logical one, the threshold voltage level of the memory cell is increased to state A, as indicated by arrow 302. This causes the first stylization process to terminate. In the second stylization process, the threshold voltage level of the memory unit is set according to the bit being programmed into the upper logical page. If the upper logical page bit stores a logic "1", since the memory cell is in state E or A (depending on the stylization of the lower page bit), both states carry the upper page bit. Yuan "1", so no stylization occurred. If the upper page bit is logic "0", the threshold voltage is shifted. If the first process causes the memory cell to remain in the erased state E, then in the second phase, the memory bank 7L is programmed such that the threshold voltage is increased to the range of the bear C, as indicated by arrow 306. Shown. If the first stylization process causes the record unit to be privateized to state a, the memory unit is further programmed in the second process such that the threshold voltage is increased to the state B range 125212.doc -22 - 200837760, as indicated by arrow 304. The result of the second process is to program the memory cells into data that is designated to cause the upper page to store logic and without changing the lower page.

1 在一項具體實施例中,可設定一系統用以如果寫入足以 填滿一整頁的資料,則實行全序列寫入。如果資料不足以 寫入一全頁,則程式化處理程序可用所接收之資料來程式 化下部頁。當接收後續資料時,系統將接著程式化上部 頁。在另一項具體實施例中,系統可開始使用兩次進程技 術寫入資料,並且接著如果後續接收到足夠的資料,則轉 換至全序列程式化模式,以填滿一整個(或大多數)字線的 記憶體單元。如需此具體實施例之詳細資訊,請參閱發明 人 Sergy Anatolievich Gorobets及 Yan Li於 2004 年 12 月 04 日 申請之美國專利申請案第11/〇13,125號標題為”pipelined gramming 〇f Non-Volatile Memories Using Early Data",該案整份内容以引用方式併入本文中。 除了如圖8所示之全序列與上部頁_下部頁程式化以外, ,據各項具體實施例’亦可使用其它程式化技術。舉例而 :,在一項具體實施例中,可使用一種兩次進程技術,其 2 下歹】方式減小浮動閘極至浮動閘極耦合:對於任何特 疋口己U體單兀’繼寫入至相鄰記憶體單元的先前頁之後, ^ 忒特疋圮憶體單元的一特定頁。在此項技術之一實 碼,並°己L、體單兀使用四種資料狀態(例如,非格雷編 別二中、、二擦除狀储存11、經程式化狀態A、B與c分 別儲存01、10盥— /、) 儲存母記憶體單元兩個位元資料。 125212.d〇i -23- 200837760 母一^體单元健存兩頁資料,基於參考目的 部頁與下部頁。在第一進程中,下部頁被程式化 部頁中儲存資料!的記憶體單元維持在狀態e, 料0的記憶體單元上升至-中間狀態B·。在第二進程:貝 f 上!7被程式化。待在上部頁中健存資料1的處於經捧除 狀悲E之記憶體單元維持在狀態E,而待在上部頁中儲存: 料〇的處於狀態E之記憶體單元被程式化至狀態A。待在二 部 '中儲存資料1的處於中間狀態B,之記憶體單元被程式 化至取終狀態B,而待在上部頁中儲存資❹的處於中間狀 態B,之記憶體單元被程式化至狀態c。1 In one embodiment, a system can be set up to perform a full sequence of writes if sufficient data is written to fill a full page. If the data is not sufficient to write a full page, 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 begin to write data using two process techniques, and then if sufficient data is subsequently received, transition to full sequence stylized mode to fill an entire (or majority) The memory unit of the word line. For more information on this specific example, please refer to the inventor Sergy Anatolievich Gorobets and Yan Li on December 4, 2004, U.S. Patent Application Serial No. 11/13, No. 125 entitled "pipelined gramming 〇f Non- Volatile Memories Using Early Data", the entire contents of which is incorporated herein by reference. In addition to the full sequence and upper page_lower page stylization as shown in FIG. 8, according to various embodiments, 'may also be used Other stylized techniques. For example: In one embodiment, a two-stage process technique can be used, which reduces the floating gate to floating gate coupling: for any special port After a single page is written to the previous page of the adjacent memory unit, ^ is a specific page of the unit. In one of the techniques, the code is used, and the four types are used. Data status (for example, non-Gray 2, 2, erased storage 11, stylized states A, B, and c store 01, 10 盥 - /, respectively) store two bit data of the mother memory unit. 125212.d〇i -23- 20083776 0 The parent-body unit stores two pages of data based on the reference destination page and the lower page. In the first process, the lower page is stored in the stylized part page! The memory unit is maintained in state e, material 0 The memory unit rises to the intermediate state B. In the second process: Baye f! 7 is stylized. The memory unit in the upper page where the data 1 is in the state of being erased is maintained in the state E. And to be stored in the upper page: the memory unit in state E is programmed to state A. The memory unit in the intermediate state B where the data 1 is stored in the second part is programmed to In the final state B, the memory unit in the intermediate state B to be stored in the upper page is programmed to the state c.

圖9之表格描述單頁技術(諸如上部頁/下冑頁程式化憤 王序列私式化(-次程式化所有資料狀態)兩者中經程式化 之記憶體單元的各種電容電荷耦合。圖9之表格描述四狀 態式記憶體裝置之效應,但是熟習此項技術者應明白,具 體化之原理可配合具有更多實體狀態之系統予以運用。在 單頁私式化與全序列程式化兩者中,經程式化至第一經程 式化狀態A之記憶體單元遭遇到最多的電容電荷耦合效 應。第一列描述單頁程式化(攔382)以及處於狀態A之記憶 體單兀且其鄰近者經程式化至狀態c (欄38句。處於狀態A 之記憶體單元將歷經其表觀臨限電壓增大,其等於常數α (表示介於浮動閘極之間的耦合係數)乘介於經擦除狀態Ε 與經程式化狀態C之間的電壓變化之乘積。因為當程式化 第一頁時程式化處於狀態Α之記憶體單元,並且稍後當程 式化第二頁時將其鄰近者程式化至狀態C,所以鄰近者的 125212.doc -24- 200837760 于動閘:電[將上升’其上升量為介於經擦除狀態與經程 式m c之間的差異。欄386列出對於記憶體系統中介於 各種狀態之間μ程式化的記憶體單元的示範性臨限電壓 變化。在所提供之實例中,自狀態Ε程式化至狀態C將使臨 限電壓偏移約6伏。處於狀態Α之記憶體單元將歷經其臨限 電壓中之表觀偏移’其等於此差異乘麵合係數之乘積。其 鄰近者經私式化至狀態Β的處於狀態Α之記憶體單元將歷 經表觀偏移,彡等於常數乘介於第-經程式化狀態A與第 一經程式化狀態B之間的電壓變化(例如,2伏)之乘積。其 鄰近者處於狀態C的處於狀態B之記憶體單元將歷經其臨 限電壓中之表觀偏移,其等於常數乘介於第二經程式化狀 悲B與第二經程式化狀態c之間的電壓變化(例如,^伏)之 乘積。 雖然狀態A之記憶體單元在全序列程式化中歷經較小之 電容電荷叙合,但是與處於其它狀態之記憶體單元相比, 狀恶A之記憶體單元仍然歷經更多電容電荷耦合。對於其 鄰近者經程式化至狀態C的處於狀態a之記憶體單元,在 全序列程式化中,在所選記憶體單元被程式化之後,鄰近 者將僅自狀態A上升至狀態C (例如,2伏)。在全序列中, 同時程式化所有狀態。因此,在大多數情況中,所選記憶 體單元將在大約相同於其鄰近者之時間抵達狀態A。因 此’在所選記憶體單元被程式化之後的唯一耦合效應係由 自狀態A移至狀態c的鄰近者所反映。 其鄰近者經程式化至狀態B的處於狀態A之記憶體單元 125212.doc -25- 200837760 將歷經表觀偏移,其等於常數乘介於第一經程式化狀態A 與第二經程式化狀態B之間的電壓變化(例如,2伏)之乘 積。其鄰近者處於狀態C的處於狀態B之記憶體單元將歷 經其臨限電壓中之表觀偏移,其等於常數乘介於第二經程 式化狀恶B與第三經程式化狀態c之間的電壓變化(例如,2 伏)之乘積。 如圖9所示,處於最低經程式化實體狀態之記憶體單元 將遭遇到來自稍後經程式化之鄰近者的最多電容電荷耦 合。彼等不均衡耦合效應可使A狀態分佈加寬並且可能這 成項取錯誤。為了補償處於第一經程式化狀態之記憶體單 元所歷經的較大表觀臨限電壓偏移量,在一項具體實施例 中,用較高精確度程式化第一經程式化狀態以補償額外電 荷耦合效應。在各項具體實施例中,使用技術來達成第一 經程式化狀態之更緊密臨限電壓分佈,以補償其表觀臨限 電壓之後續變化。在一項具體實施例中,當程式化記憶體 單7G至最低可程式狀態時,可對於程式電壓訊號使用較小 遞增值。在另一具體實施例中,當程式化一第一邏輯頁資 料至一組記憶體單元時,遞增值較小;當程式化一第二邏 輯頁資料至該組記憶體單元時,遞增值較大。在管線式架 構中’當記憶體自單頁程式化切換至全序列程式化時,可 增大遞增值。 圖10繪示非揮發性記憶體系統處理與寫入個別邏輯頁資 料之夺序圖’其中先完成一組記憶體單元的第一頁資料, 其後才開始第二頁資料。首先在控制器處接收來自主機裝 125212.doc •26- 200837760 置的貧料。控制器轉遞資料與位址資訊至行控制電路,今 行控”路將資料置於資料暫存器中。在圖1〇中,對於= I己憶體單元’控制器純收_τ部頁資料,其後才接收 :上:頁資料。載入適當的暫存器,並且回應於來自控制 '的-程式命令,在時_,狀態機將開始程式化該下部 頁資料至該組記憶體單元中。 在夺間tl之後,自主機接收用於對應組記憶體單元之上 部頁資料的資料。在時m2,控制器完成載人請至陣列 處的暫存器令。在圖10所示之特定具體實施例中,系統直 至:已程式化下部頁才程式化上部頁資料。據此,狀態機繼 μ私式化下部頁資料至該組記憶體單元中,直到時間u為 止’並且上部頁資料維持在暫存器中。一旦下部頁資料被 程j化,控制器可發出用於上部頁資料之程式命令,並且 狀悲機作出回應而程式化來自暫存器的上部頁資料至陣列 中。如所述,㈣化上部頁包括將記憶體單元自狀態E程 式化至狀態C及自狀態A程式化至狀態B。在—項替代具體 貝施例中,使用如上文所述之:次進程技術,使得程式化 上部頁包括將記憶體單元自狀態味式化至狀態以及自中 間狀態程式化至狀態B與狀態c。 根據本文揭示技術之一項具體實施例,在如圖1〇所示之 一具體實施例中,#在時間t3程式化上部頁時,對於程式 電壓訊號使用-不同之遞增值。圖u繪示根據一具體實施 例,程式電壓tfl號。將記憶體單元自經擦除狀態程式化至 狀態A開始於當施加一第一程式電壓脈衝時之時間u。之 125212.doc -27· 200837760 後只行一驗證以判定被程式化至狀態A的記憶體單元(若有 的話)’施加一額外脈衝,其峰值係以△Vmm遞增。繼續此 處理私序,直到下部頁程式化在時間t3完成。 在程式電壓脈衝數量抵達預先決定最大數量之後,或待 • 被秋式化至狀態A的充分數量記憶體單元經成功驗證為已 抵達狀恶A之後,下部頁程式化可完成。隨著記憶體單元 經耘式化為狀態A,或隨著下部頁程式化已完成,程式電 ζ [遞^值變更為AVPgm2。在一項具體實施例中,為了加速 私式化處理程序,Δνρ㈤大於AVpgml。在-項具體實施例 中’、在進行分析以判定適當值之後選擇AVpgm2,致使儘速 ^式化陣列中的大多數記憶體單元(較大遞增值),但同時 亦確保精確程式化資料(較小遞增值)。之值可係 △VPgm2之經調整值,以補償經程式化至狀態Α之記憶體單 元所歷絰的增大之電荷耦合量。舉例而言,可依據所接收 之額外電容電荷耗合量,自AVpgm2值降低Δνρ㈤。請參考 1/ 圖9舉例而吕,當程式化狀態Α時之ΔνΡΕηι1低於之 里可依據耘式化之後其鄰近者之浮動閘極電壓中之4伏差 異相對於經程式化至狀態Β與C之記憶體單元的2伏差異。 舉非限制性只例而言,在一項具體實施例中,用於下部頁 . ㈤可係約〇.3伏,帛於上部頁或全序列程式化的 △Vpgm2可係約〇·4伏。 圖12繪示非揮發性記憶體系統處理與寫入邏輯頁資料之 時序圖’其中開始對於第一頁資料之程式化,並且接著當 接收第二頁資料時暫停或中斷,以便開始對於所有頁的全 125212.doc •28· 200837760 序列程式化。在控制器處接收來自主機裝置的資料,控制 器轉遞資料與位址資訊至行控制電路,該行控制電路^ 料置於資料暫存器中。在圖12中’對於相同記憶體單元貝 控制器先接收一下部頁資料’其後才接收一上部頁資料。 載入適當的暫存器,並且回應於來自控制器的程式命令, =時間ti ’狀態機開始程式化該下部頁資料至該組記憶體 單元中。The table of Figure 9 depicts the various capacitive charge couplings of a single page technique (such as the upper page/lower page stylized version of the insane sequence (--stylized all data states). The table of 9 describes the effects of the four-state memory device, but those skilled in the art should understand that the principle of materialization can be applied to systems with more physical states. In single page customization and full sequence programming Among them, the memory unit programmed to the first programmed state A encounters the most capacitive charge coupling effect. The first column describes the single page stylization (block 382) and the memory unit in state A and its The neighbor is programmed to state c (column 38. The memory unit in state A will increase in its apparent threshold voltage, which is equal to the constant α (representing the coupling coefficient between the floating gates). The product of the voltage change between the erased state Ε and the stylized state C. Because the stylized memory cell is programmed when the first page is programmed, and later when the second page is programmed Adjacent The program is programmed to state C, so the neighbor's 125212.doc -24-200837760 is on the gate: the electricity [will rise] its rise is the difference between the erased state and the program mc. Column 386 lists An exemplary threshold voltage change for a memory cell that is stylized between states in a memory system. In the example provided, staging from state 状态 to state C will shift the threshold voltage by about 6 volts. The memory cell in state 将 will experience the apparent offset in its threshold voltage, which is equal to the product of this difference multiplied by the face factor. The neighbors are privately conditioned to the state Β memory cell Will experience an apparent offset, 彡 equals the constant multiplied by the product of the voltage change (eg, 2 volts) between the first-stylized state A and the first programmed state B. The neighbor is in state C. The memory cell of state B will experience an apparent offset in its threshold voltage, which is equal to the constant multiplying the voltage change between the second programmed B and the second programmed state c (eg, ^ The product of volts. Although the memory unit of state A In the full sequence of stylization, the smaller capacitive charge is combined, but compared with the memory cells in other states, the memory cell of the A is still more capacitively coupled. For its neighbors, it is programmed to The memory unit of state C in state a, in full sequence stylization, after the selected memory unit is programmed, the neighbor will only rise from state A to state C (eg, 2 volts). In the middle, all states are programmed at the same time. Therefore, in most cases, the selected memory unit will arrive at state A at approximately the same time as its neighbors. Therefore, 'the only coupling after the selected memory unit is programmed The effect is reflected by the neighbors moving from state A to state c. The memory unit 125212.doc -25-200837760 whose state is programmed to state B in state B will experience an apparent offset equal to a constant Multiplies the product of the voltage change (eg, 2 volts) between the first programmed state A and the second programmed state B. The memory cell in state B whose state is in state B will experience an apparent offset in its threshold voltage, which is equal to the constant multiplication between the second programmed B and the third programmed state c. The product of the voltage change (for example, 2 volts). As shown in Figure 9, the memory cells in the lowest programmed entity state will encounter the most capacitive charge coupling from the later programmed neighbors. These unbalanced coupling effects can widen the A-state distribution and may cause errors in this term. To compensate for the larger apparent threshold voltage offset experienced by the memory cells in the first programmed state, in one embodiment, the first programmed state is programmed with higher accuracy to compensate Additional charge coupling effect. In various embodiments, techniques are used to achieve a tighter threshold voltage distribution of the first programmed state to compensate for subsequent changes in its apparent threshold voltage. In one embodiment, a smaller increment value can be used for the program voltage signal when the program memory is 7G to the lowest programmable state. In another embodiment, when the first logical page data is programmed to a set of memory cells, the increment value is smaller; when the second logical page data is programmed to the set of memory cells, the increment value is compared. Big. In a pipelined architecture, the incremental value can be increased when the memory is programmed from a single page to a full sequence. Figure 10 illustrates a sequence diagram of a non-volatile memory system processing and writing individual logical page data. The first page of a set of memory cells is first completed, and then the second page of data is started. The lean material from the main unit 125212.doc •26-200837760 is first received at the controller. The controller forwards the data and the address information to the row control circuit, and the current control device places the data in the data register. In Fig. 1〇, the controller _τ is for the =I memory unit Page data, after which it receives: top: page data. Loads the appropriate scratchpad and responds to the command from the control's program. At time _, the state machine will start to program the lower page data to the group of memory. In the body unit, after the interval t1, the data for the upper page data of the corresponding group memory unit is received from the host. At time m2, the controller completes the register order to the person at the array. In a particular embodiment, the system up to: the programmed lower page is to program the upper page data. Accordingly, the state machine subsequently privateizes the lower page data into the set of memory cells until time u' and The upper page data is maintained in the scratchpad. Once the lower page data is processed, the controller can issue a program command for the upper page data, and the sorrow machine responds to program the upper page data from the scratchpad to In the array. As mentioned, (4) The page includes staging the memory unit from state E to state C and from state A to state B. In the case of the alternative, the use of the sub-process technique, as described above, makes the stylized upper part The page includes staging the memory unit from the state to the state and staging from the intermediate state to state B and state c. According to a particular embodiment of the technology disclosed herein, in one embodiment as shown in FIG. When the upper page is programmed at time t3, the program voltage signal is used with a different increment value. Figure u illustrates the program voltage tfl number according to an embodiment. The memory unit is programmed from the erased state to the state. A begins with the time u when a first program voltage pulse is applied. After 125212.doc -27·200837760, only one verification is performed to determine the memory unit (if any) that is programmed to state A. The extra pulse, whose peak value is incremented by ΔVmm. Continue this process private sequence until the lower page is programmed at time t3. After the program voltage pulse reaches the predetermined maximum number, or wait for • autumn After the sufficient number of memory cells of state A have been successfully verified as having arrived at A, the lower page stylization can be completed. As the memory cells are formatted as state A, or as the lower page is programmed The program value is changed to AVPgm2. In a specific embodiment, Δνρ(f) is greater than AVpgml in order to speed up the privateization process. In the specific embodiment, after the analysis is performed to determine the appropriate value, AVpgm2, which causes most of the memory cells in the array to be optimized (larger increments), but also ensures accurate stylized data (smaller increments). The value can be adjusted to ΔVPgm2 to Compensating for the increased amount of charge coupling experienced by the memory cells programmed to state 。. For example, Δνρ (f) can be reduced from the AVpgm2 value depending on the amount of additional capacitance charge received. Please refer to 1/Fig. 9 for example. When the stylized state is lower than ΔνΡΕηι1, the difference of 4 volts in the floating gate voltage of its neighbors after the simplification can be based on the stylized state. A 2 volt difference in the memory unit of C. By way of non-limiting example, in one embodiment, for the lower page. (5) can be about 〇3 volts, △Vpgm2 can be about 〇·4 volts on the upper page or in the full sequence. . 12 is a timing diagram of processing and writing logical page data of a non-volatile memory system, in which the stylization for the first page of data is started, and then when the second page of data is received, it is paused or interrupted to start for all pages. The full 125212.doc •28· 200837760 sequence stylized. The controller receives the data from the host device, and the controller forwards the data and the address information to the row control circuit, and the row control circuit is placed in the data register. In Fig. 12, 'the next page data is received first for the same memory unit controller', and then an upper page data is received. The appropriate scratchpad is loaded and, in response to a program command from the controller, the =time ti' state machine begins to program the lower page data into the set of memory cells.

C 在時間ti之後,自主機接收用於相同組記憶體單元之上 部㈣料。在時間t2,㈣器完成載人資料至陣列處的暫 存器中。在判定新資料係用於相同記憶體單元的上部頁資 料之後,控制器可中斷程式化操作。在圖12所示之特定具 體實施例中,系統可中斷、暫停或以其它方式停止下部頁 程式化處理程序,以切換至全序列操作。在一項具體實施 例中,控制器僅若下部頁程式化操作未已完成縣充分= 成以保證允許其結束才停止下部頁程式化操作。 70 控制器可在提供上部頁資料至列控制電路中的資料暫存 器之後,發佈一用於全序列程式化之新程式命令。'可實= :驗證操作以鎖定已抵達其目標狀態之記憶體單元= 著,使用全序列程式化繼續該兩頁資料之程式化。 根據一項具體實施例’在如圖12所示之一具體實施例 中’當在時間t2之後制全序列程式化時,對於程式電壓 訊號使用-不同之遞增值。圖12繪示根據_項具體實施例 之程式電壓訊號。將記憶體單元自經擦除狀態程式化至狀 態A開始於當施加一第一擦除電壓脈衝時之時間丨丨。在對 125212.doc -29- 200837760 於下σ卩頁私式化至狀態A完成之前,自主機接收上部頁資 料在時間t2 ’控制器發佈資料至行控制電路,該行控制 電路將資料儲存於暫存器中。 在驗證、鎖定及發佈一新程式命令之後,全序列程式化 開始於時間t2之後的第一脈衝。配合發佈之全序列命令, 在各止下部頁程式化之後,以高於最後程式電壓脈衝的 =vPgm2來遞增下一程式電壓脈衝。Δνρ^2值與值可 隨具體實施例及各種實施方案需求而異。在—項具體實施 例中,如參考圖U之描述來選擇丨與△Vpgw。 圖14繪示根據一項具體實施例之程式化非揮發性記憶體 之方法的流程圖。依據正被程式化之狀態及/或程式化操 作與資料類型,使用一可變遞增值。在步驟5〇2,一"資料 載入"命令係由控制器318予以發出且被輸入至命令電路 314,以允許將資料輸入至資料輸入/輸出緩衝器η〕。輸 入之育料被辨識為一命令,且經由一輸入至命令電路Η# 的命令鎖存訊號(圖中未繪示)而由狀態機316予以鎖存。在 步驟504,從控制器或主機將指定頁位址的位址資料輸入 至列控制器或解碼器3G6。輸人之f料被辨識為頁位址, 且經由狀態機316予以鎖存(受到輸人至命令電路314的位 址鎖存訊號所影響)。在步驟寫,所^址之頁的一頁程式 資料被輸入至資料輸入/輸出緩衝器312以進行程式化。: 例而言,在一具體實施例中可輸入阳個位元組資料。該 貧料被鎖存在用於所選位元線的適#暫存器中。在一此且 體實施例中,該資料亦被鎖存在用於驗證操作使用之所選 125212.doc -30- 200837760C After the time ti, the upper (four) material for the same group of memory cells is received from the host. At time t2, the (4) device completes the manned data into the register at the array. After determining that the new data is used for the upper page information of the same memory unit, the controller can interrupt the stylization operation. In the particular embodiment illustrated in Figure 12, the system can interrupt, pause, or otherwise stop the lower page stylization process to switch to full sequence operation. In one embodiment, the controller stops the lower page stylization operation only if the lower page stylization operation has not completed the county sufficient = to ensure that it is allowed to end. The controller can issue a new program command for full sequence stylization after providing the upper page data to the data register in the column control circuit. 'Complete=: Verify operation to lock the memory unit that has reached its target state = continue to use the full sequence of stylization to continue the stylization of the two pages of data. According to a specific embodiment, 'in a specific embodiment as shown in Fig. 12', when the full sequence is programmed after time t2, a different increment value is used for the program voltage signal. Figure 12 illustrates a program voltage signal in accordance with an embodiment of the invention. Staging the memory cell from the erased state to state A begins at a time 当 when a first erase voltage pulse is applied. Before the completion of the 125212.doc -29-200837760 private state to state A, the upper page data is received from the host at time t2 'the controller issues the data to the row control circuit, and the row control circuit stores the data in In the scratchpad. After verifying, locking, and issuing a new program command, the full sequence stylization begins with the first pulse after time t2. With the full sequence command issued, after the lower page is programmed, the next program voltage pulse is incremented by =vPgm2 higher than the last program voltage pulse. The Δνρ^2 values and values may vary with the particular embodiment and various embodiments. In the specific embodiment, 丨 and ΔVpgw are selected as described with reference to Figure U. 14 is a flow chart of a method of staging non-volatile memory in accordance with an embodiment. A variable increment value is used depending on the state being programmed and/or the stylized operation and data type. At step 5, 2, a "data load" command is issued by controller 318 and input to command circuit 314 to allow data to be entered into data input/output buffer η]. The input feed is recognized as a command and is latched by state machine 316 via a command latch signal (not shown) that is input to command circuit Η#. At step 504, the address data of the specified page address is input from the controller or host to the column controller or decoder 3G6. The input material is recognized as a page address and is latched via state machine 316 (affected by the address latch signal input to command circuit 314). At the step of writing, a page of program data of the page of the address is input to the data input/output buffer 312 for programization. For example, in a specific embodiment, a positive byte data can be entered. The poor material is latched in the appropriate register for the selected bit line. In one embodiment, the data is also latched in the selection for verification operation. 125212.doc -30- 200837760

位元線的第二暫存器中。在一項具體實施例中,資料對應 於儲存額外邏輯頁的一組記憶體單元所儲存的一邏輯頁資 料。该組記憶體單元可被設定以在該組記憶體單元所定義 的只體頁中儲存多個邏輯頁。一實體頁可包括一組記憶體 早元了儲存之所有貧料,諸如一次一列。一實體頁亦可表 不可一次程式化至該組記憶體單元的最大資料量。在步驟 508 ’ 一”程式”命令係由控制器318予以發出且被輸入至資 料輸入7輸出緩衝器312。經由輸入至命令電路314的命令 鎖存訊號而由狀態機3 1 6鎖存該命令。 藉由”程式”命令之觸發,使用圖6所示之施加至適當字 線的步進式脈衝,由狀態機3丨6控制以將在步驟中鎖存 的資料程式化至所選記憶體單元中。在步驟51〇,施加至 所選字線的程式化脈衝電壓位準Vpgm被初始化為開始脈衝 (例如,12伏),並且狀態機3 16所維護的一程式計數器 被初=化為0或另一值。在步驟512,施加第一 脈衝至 所選字線。如果儲存在一特定資料鎖存器中邏輯,,〇”指示 出應程式化相對應之記憶體單元,則相對應之位元線被接 地。另一方面,如果儲存在一特定鎖存器中的邏輯”丨,,指 不出相對應之記憶體單it應維持其現有資料狀態,則相對 應之位元線被連接至VDD以禁止程式化。 隹步驟514 、心、別木1貝踯到 —所選記憶體單元的目標臨限電壓已到達適當位準,則相 對應之資料鎖存器中儲存的資料被變更為邏輯"丨、如2 價測到目標臨限電壓未到達適當位準,則不變更相對:之 125212.doc -31 - 200837760The second register of the bit line. In one embodiment, the data corresponds to a logical page of information stored by a set of memory cells storing additional logical pages. The set of memory cells can be set to store a plurality of logical pages in a body page defined by the set of memory cells. A physical page can include a set of memory that has all the poor materials stored in the early days, such as one column at a time. A physical page can also be programmed to the maximum amount of data in the set of memory cells. At step 508, a "program" command is issued by controller 318 and input to data input 7 output buffer 312. The command is latched by state machine 3 16 via a command latch signal input to command circuit 314. By the trigger of the "program" command, the stepping pulse applied to the appropriate word line as shown in FIG. 6 is controlled by the state machine 3丨6 to program the data latched in the step to the selected memory unit. in. At step 51, the programmed pulse voltage level Vpgm applied to the selected word line is initialized to a start pulse (eg, 12 volts), and a program counter maintained by state machine 316 is initialized to zero or another A value. At step 512, a first pulse is applied to the selected word line. If the logic is stored in a particular data latch, 〇" indicates that the corresponding memory cell should be programmed, then the corresponding bit line is grounded. On the other hand, if stored in a particular latch The logic "丨", which means that the corresponding memory list should maintain its existing data state, the corresponding bit line is connected to VDD to prohibit stylization.隹Step 514, heart, 别木1踯到—The target threshold voltage of the selected memory unit has reached the appropriate level, then the data stored in the corresponding data latch is changed to logic "丨, 2 If the target threshold voltage is not reached to the appropriate level, then the relative value is not changed: 125212.doc -31 - 200837760

:料鎖存器中儲存的資料。在此方式中,在本身相對應之 資料鎖存器中已儲存邏輯,,i π的位元線不需要予以程式 化。當所有資料鎖存器皆正在儲存邏輯”,,時,狀態機知 道已程式化所有所選記憶體單元。在步驟516,檢查是否:有諸鎖存H正儲存邏輯"Γ,。若是,因為所㈣選記fe體單兀皆已予以程式化且已予以驗證經程式化至其目標 狀態,所以程式化處理程序完成且成功。在步驟川,報 告·’通過"(PASS)狀態。 ί. 在步驟516,如果判定非所有資料鎖存器正儲存邏輯 "1”,則程式化處理程序繼續進行。在步驟52〇,比對一程 式限制值來檢查該㈣計㈣pc。―項實例之程式限制值 係20,但是,在各種實施方案中可以使用其他值。如果該 程式計數器PC不小於20,則在步驟526判定尚未成功程式 化的位元數量是否等於或小於預先衫數量。如果未成^ 程式化的位元數量等於或小於預先決定數量,則用旗標將 程式化處理程序標示為已通過,並且在步驟528報告一通 過狀態。在讀取處理程序期間,可使用錯誤修正來修正未 成功程式化的位元。但是’如果未成功程式化的位元數量 大於預先決定數量,則用旗標將程式化處理程序標示為已 失敗,並且在步驟530報告一失敗狀態。如果該^式'計數 器PC小於20 則按步進大小來增大Vpgm位準: Material stored in the material latch. In this mode, the logic has been stored in its own corresponding data latch, and the bit line of i π does not need to be programmed. When all data latches are storing logic, the state machine knows that all selected memory cells have been programmed. At step 516, it is checked if there are latches H are storing logic "Γ, if so, Because the (4) selection fe body has been programmed and verified to be programmed to its target state, the stylization process is completed and successful. In the step, report 'pass' (PASS) status. In step 516, if it is determined that not all of the data latches are storing the logical "1", the stylized processing continues. In step 52, the (four) pc is checked by comparing the one-way limit value. The program limit value for the item instance is 20, however, other values may be used in various embodiments. If the program counter PC is not less than 20, it is determined in step 526 whether the number of bits that have not been successfully programmed is equal to or less than the number of pre-shirts. If the number of unprogrammed bits is equal to or less than a predetermined number, the stylized handler is flagged as passed, and a pass status is reported in step 528. Error corrections can be used to correct unsuccessfully programmed bits during the read process. However, if the number of unsuccessfully stylized bits is greater than a predetermined number, the stylized handler is flagged as failed with a flag, and a failure status is reported in step 530. If the type 'counter PC is less than 20, increase the Vpgm level by the step size.

522遞增該程式計數器pC 並且在步驟 步驟512,以施加下一 Vpgm脈衝。 在步驟522,處理程序迴圈回到 如步驟502’至510’, 當實行圖14所示之任何操作時,記 125212.doc -32 - 200837760 體糸統可接㈣於-或多個㈣的額外資料。㈣η 在任何時候來自主機的資料,而發佈-額外資料載 步驟502’)。步驟游可發生於除圖14具料示以外 八匕位置與時間。可接收來自主機之資料,並且在步驟 如,執行-資料載人命令,以在步驟财於列控制器處載 入位址資m在步㈣6,將㈣程式化轉列處可用 =存μ。在-項具體實施例中,控制器等待直到狀態 機或行控制電路發佈其可在暫存器中接收更多資料的一可 用訊號。 在步驟508,’控制器及狀態機可用各種方式來處理新近 接收之貝料。在一項具體實施例中,控制器停止當前之下 部頁程式化操作。這係藉由自方塊綱,至方塊5〇8之箭頭描 繪。接著’控制器發佈一新程式命令,該新程式命令指示 狀態機使用全序列程式化同時(並行)程式化下部頁鱼上部 頁資料。在步驟510’’狀態機可設定Δνρ^ν_,如此 當以全序列程式化該兩頁時使用一較大遞增值。接著,程 式操作繼續(如上文所述),但是使用較大遞增值 在步驟51〇’所示之一項具體實施例中’在開始全序列之前 重設Vpgm並且PC計數器被重設為〇。ν㈣&丨2可等於、小於 或大於Vinitiall。但是’非在所有具體實施例中重設1。 舉例而言’如果上部頁相對迅速抵達並且僅已施加很少數 私式電壓脈衝’則Vpgm可開始於其最後峰值、開始於大於 其最後峰值之值(以或其它值遞增)或開始於小於其 最後峰值之值。 125212.doc -33- 200837760 圖15繪示更詳細描述圖μ之牛 口丨4之步驟508,中設定程式命522 increments the program counter pC and at step 512, applies the next Vpgm pulse. At step 522, the processing loops back to steps 502' through 510'. When any of the operations shown in FIG. 14 are performed, the 125212.doc -32 - 200837760 system can be connected (four) to - or multiple (four) Additional information. (d) η at any time from the host's data, and the release-extra data is carried in step 502'). The step tour can occur in addition to the information shown in Figure 14 and the position and time of the gossip. The data from the host can be received, and in the step, for example, the execution-data manned command is executed to load the address in the step of the column controller in step (4) 6, and the (four) stylized transition is available = store μ. In the specific embodiment, the controller waits until the state machine or row control circuit issues an available signal that it can receive more data in the scratchpad. At step 508, the controller and state machine can process the newly received bedding in a variety of ways. In one embodiment, the controller stops the current lower page stylization operation. This is drawn by the arrow from the square to the box 5〇8. The controller then issues a new program command that instructs the state machine to use the full sequence to program (simultaneously) program the lower page of the lower page. At step 510'' the state machine can set Δνρ^ν_ such that a larger increment value is used when the two pages are programmed in full sequence. Next, the program operation continues (as described above), but using a larger increment value in a particular embodiment shown in step 51 〇' resets Vpgm before starting the full sequence and the PC counter is reset to 〇. ν(4)&丨2 can be equal to, less than, or greater than Vinitiall. However, it is not reset in all of the specific embodiments. For example, 'if the upper page arrives relatively quickly and only a few private voltage pulses have been applied' then Vpgm can start at its last peak, start at a value greater than its last peak (increment in other values) or start at less than Its final peak value. 125212.doc -33- 200837760 Figure 15 depicts a more detailed description of the step 508 of Figure 牛牛丨4, setting the program life

流程圖。圖15就是-項示範性實施方案,其中對於 程式化使用-較小遞増值,並且對於上部Μ式化或 列程式化使用-較大遞增值。在圖15之具體實施例中,在 發佈程式命令至行控制電路前,在步驟602,控制器比較 新近接收之㈣的位址資料與對於職錢體單元之下部 頁當前正被程式化之資料。在步驟604,如果控制器發現 匹配(其意謂著對於相同記憶體單元正被程式化之資料係 新近接收之資料),則方法繼續進行步驟608。在—項具體 實施例中,狀態機及/或行控制電路實行步驟6()4。如果位 址貝料不匹配’則下部頁程式化繼續進行,例如,藉由在 圖14之步驟512施加下一程式電壓脈衝。除步驟512以外, 圖14之方法亦可在圖15之步驟6〇6之後繼續進行。如果位 址資料不匹配,則在步驟6〇8 ’控制器判定下部頁程式化 是否完成。當已施加預先決定數量之程式電壓脈衝時、當 待被程式化至狀態Α的所有記憶體單元已被成功程式化至 狀態A時,或當預先決定數量之記憶體單元已被成功程式 化至狀態A時,下部頁程式化可完成。 如果下。卩頁知式化完成,則在步驟6 i 〇開始將上部頁資 料程式化至該組記憶體單元,此處由控制器發出上部頁程 式命令。在步驟620,上部頁程式化繼續進行,如圖14中 之步驟510’所示,此處程式電壓訊號被初始化為⑴並 且被設定為AVp^2。當程式化記憶體單元的上部頁 時,使用第二遞增值來增大後續程式化脈衝的大 125212.doc -34- 200837760 鄰近::狀悲A之讀體單元將歸因於隨後程式化之 川,^=純寬之臨限電壓分佈,所以在步驟 頁使用較大頁時使用較小ΔΥρ_,並且對於上部 驟2下Λ頁=化未完成(或未接w 白半 停下部頁程式化操作。在圖14中,藉由 々^化°控制器或狀態機可發佈-強制行控制電路停止 王:、化下部頁資料之命令。於步驟612停止下部頁程式化 :後’在步驟614 ’控制器設定全序列程式資料命令。在 、,員具體實施例中,該命令係經由輸人至命令電路的命令 鎖存訊號而輸入至緩衝器312並且由狀態機予以鎖存。回 應於該程式命令而觸發狀態機,以程式化所選記憶體單 凡、在步驟616,控制器可使行控制電路實行驗證操作, 疋於下邛頁程式化期間已成功程式化的記憶體單元。 、;已抵達其目標位準的任何記憶體單元,控制器鎖定彼 等圮憶體單元以阻止進一步程式化。在鎖定已程式化之記 憶體單元之後,方法繼續進行步驟62〇,其相對應於圖Μ 之步驟5 1 〇’,在此步驟可調整遞增值。 在一項具體實施例中,大於ΔνρΕΐη1。在程式化之 後,為了減小狀態Α分佈寬度,對於下部頁程式化或對於 程式化最低經程式化實體狀態,使用較小之Vpgmi。因為 處於狀態A之記憶體單元歷經額外電容耦合效應,所以使 用較小之AVpgm值’以避免過度程式化及/或造成程式化干 125212.doc -35- 200837760 擾。在一項具體實施例中,步驟510,進一步包括重設、@ 之初始值。舉例而言,如果下部百葙彳 ^ 不^ 口丨貝私式化幾乎完成,則在 施加電壓信號以程式化上部頁或 , |貝次以王序列進行程式化之 别’ Vpgm之值可被重設為一較低值。太 s ^ 权似值在一項具體實施例 中,Vpgm被重設為相同於開始下部頁程式化時的開始值。 在-項具體實施例中,v—被重設為低於其當前值但非如 用於下部頁程式化之原始值一樣低之值。在另一項具體實 施例中,當在完成下部頁程式化之後開始全序列或上部頁 程式化時,始於當程式化下部頁時使用之最後值,以第三 遞增值AVp^3來增大或減小程式電壓訊號。在一項具體實 施例中,在開始上部頁或全序列程式化之前,不重設程式 電壓訊號Vpgm。在一項具體實施例中,以厶¥1)^2增大用於 下部頁程式化的最後電壓脈衝。在另一具體實施例中,對 於第一上部頁或全序列脈衝,以AVpgm3增大用於下部頁程 式化的最後電壓脈衝。在第一額外脈衝之後,如所述,值 係以AVpgm2予以增大。AVpgm3可大於AVpgm2,以提供介於 兩個不同步進大小所致之兩種程式化速率之間的更穩定轉 變0 步驟510’亦可包括重設程式計數器pc。舉例而言,可對 於下部頁與上部頁或全序列程式化,獨立地建置最大數量 之脈衝。當介於下部頁與上部頁之間或介於下部頁與全序 列程式化之間的程式化轉變時,計數器PC可被重設為〇或 另一值。在另一具體實施例中,不重設計數器PC,並且對 於整體程式化操作使用總最大數量之反覆作業或脈衝。根 125212.doc -36- 200837760 據具體實施例,可進行實踐其它變化方案。 在圖1 5中描述之技術的變化方案中,直到程式化最低經 程式化狀態完成才增大用於程式電壓脈衝的遞增值,甚至 在轉變至全序列程式化之後。在步驟608,如果控制器判 疋下部頁程式化完成’則操作進行步驟6丨〇,如上文所 述但疋’如果下部頁程式化尚未完成,則控制器停止下 部頁程式化處理程序,並且轉換至全序列程式化,如步驟 612至618所示。但是,直到對於最低經程式化狀態(例 如狀悲A)之程式化已完成才使程式電壓遞增值變更為 △Vpgm2田所有或預先決定之記憶體單元被驗證為已抵達 狀態A時,程式電壓遞增值可變更為AVpgm2。在一項具體 實施例中’ ϋ擇並且使用贱決定之程式電壓脈衝,以判 疋何日π切換至ΔΥρ"2。舉例而言,在切換至全序列程式化 之後,可在自最初開始下部頁程式化處理程序以來已施加 預先决定數ΐ之程式電壓脈衝之後,將程式電壓遞增值變 更為△VpgW。 在一項具體實施例中,在步驟⑽,對下部頁程式化完 成之判定可絲據預先決定之程式化循環反覆次數 (即’預先決定程式電麗脈衡數量),而不顧及是否記憶體 早凡被實際程式化至適當狀態(例如,狀態Α)。當預先決 定^己憶體單元已抵達適當狀態時,亦可判定下部頁程式 化元成’這可係在驗證成訂部頁程式化正常所需的整组 記憶體單元已抵達適當狀態之前。在任—情況中,在重設 程式電麼訊號並且使用遞增值之後,依據下部頁資 125212.doc •37- 200837760 料’ 一些記憶體單元可需要額外程式化。這可於步驟610 的上部頁程式化期間予以處置,其方式為不鎖定尚未被驗 證為已抵達狀態A的待被程式化至狀態a之彼等記憶體單 元’並且於上部頁程式化期間繼續以狀態A位準進行驗 證。 在Μ —些§己憶體早元可仍然需要被程式化至最低經程式 化狀態時在步驟610轉變至上部頁程式化的一項具體實施 例中,直到對於最低經程式化狀態之程式化完成才將 △Vpgm變更至△Vpgw。將於上部頁程式化期間繼續使用 △Vpgml,直到對於最低經程式化狀態之程式化被驗證。舉 例而§ ’待被程式化至狀態A的預先決定數量之記憶體單 元已抵達狀態A之後,AVpgm可自AVpgml變更為AVpgm2。 在對圖15之另一項變化方案中,如果按步驟608判定下 部頁程式化未完成,則控制器在接收到上部頁資料之後不 立即停止下部頁程式化處理程序及轉換至全序列程式化。 而是,控制器在進行轉變之前等待直到其判定下部頁程式 化完成。 圖16繪示僅如果對於下部頁之程式化完成才在接收到額 外資料之後變更遞增值之方法的流程圖。該技術可在自下 部頁程式化轉變至上部頁程式化或自下部頁程式化轉變至 全序列程式化時切換遞增值。在步驟704接收下部頁資 料。在步驟706設定位址資料,在步驟708輸入程式資料, 並且在步驟710控制器318發出程式命令。在步驟712, Vpgm被設定為其初始值,程式計數器被初始化,並且 125212.doc -38 - 200837760 △ vpgm被設定為ΔΥμμ。在步驟714,下部頁資料被程式 化,舉例而言,如圖14 (步驟512至530)所示。當程式化下 部頁時’在步驟71 6接收上部頁資料。在步驟718設定位址 資料’並且在步驟720輸入程式資料。在發佈一更新程式 命令之前,在步驟722,控制器判定下部頁程式化是否完 成如上文對於圖15之步驟608描述所述,可用各種方式 來只行判疋狀態Α程式化是否完成。可判定是否待被程式 ( 化至狀態A的所有或預先決定數量之記憶體單元已抵達狀 態A。在其他情況中,步驟722可包括判定是否已施加預先 決定數量之脈衝,若是,則判定下部頁程式化完成。 如果下部頁程式化未完成,則在步驟724,控制器繼續 下部頁程式化,並且繼續此迴圈以等待下部頁程式化完 成田下σ卩頁耘式化完成時,方法繼續進行步驟726,此 處控制器發佈一更新程式命令。在步驟8,控制器可重 没Vpgm為一初始值νίη_ΐ2,重設遞增值為,並 I 且重:私式計數器PC。在重設彼等訊號(其可藉由重設一 負貝提仏耘式電壓訊號之數位轉類比轉換器予以實行)之 後,在步驟730,依據新命令之程式化開始。 在一項具體實施例中,在步驟726進行的程式命令設定 用於上部頁程式化。在步驟722,控制器可在下部頁程 式化完成之前等待直到對於狀態A之程式化被成功驗證。 在此情況中,下部頁程式化資料已被程式化,並且在步驟 730僅需要實行對於上部頁之程式化。在此技術之變化方 案中,可在步驟726發佈上部頁程式命令,並且一旦下部 125212.doc -39- 200837760 頁程式化完成,立即在步驟730繼續上部頁之程式化。但 是,直到對於最低經程式化狀態的程式化被驗證為完成才 在步驟728重設△Vpgm。待被程式化至狀態A的已被程式化 至狀悲A之e己彳思體早元將被鎖定以防止進一步程式化,並 且待被程式化至狀態A的尚未抵達狀態A之記憶體單元將 遭受到進一步程式化及驗證。當狀態A程式化被驗證完成 時’程式電壓遞增值可變更為AVpgm2。 在一項具體實施例中,當一些記憶體單元可仍然需要上 升至狀態A時,在步驟722判定下部頁程式化完成(例如, 藉由依據預先決定脈衝數量予以判定),在步驟726,程式 化可轉變至全序列程式化。對於全序列程式化,可在施加 第一程式電壓之前實行驗證。於全序列程式化期間,於下 部頁程式化期間已抵達其最終目標狀態的任何記憶體單元 可被鎖定以阻止進一步程式化。在判定下部頁程式化完成 之日$間尚未抵達狀態A的待被程式化至狀態a之任何記憶 體單元將於全序列程式化期間遭受到額外程式化。全序列 程式化可用狀態A位準實行驗證,所以於全序列程式化期 間,於下部頁程式化期間未抵達其目標A狀態記憶體單元 可被私式化至狀態A。在一項具體實施例中,當下部頁程 式化完成時在步驟726進行轉變至全序列程式化,但是直 到最低經程式化狀態被驗證為完成(如果尚未如此驗證)才 將程式電壓遞增值變更至AVpgm2。 僅在完成下部頁程式化之後轉變至全序列程式化時(或 上部頁程式化)及較大遞增值之具體實施例(如圖16所示)可 125212.doc 200837760 在特定實施方案中提供益處。舉例而言,一些記憶體裝置 將使用一數位轉類比轉換器,以產生用於不同大小之程式 電c脈衝的私式電壓訊號。在一項具體實施例中,該數位 轉類比轉換器可形成列控制電路的部件。一些轉換器依賴 '一數位輸入值,以產生具有所必需峰值之類比程式電壓 脈衝在-些情況中,在程式電壓脈衝之中間改良程式電 塵遞增值可顯現出問題。變更遞增量大小能無法產生具有 ( 特先前脈衝加新遞增值之峰值的下-脈衝。而是,I些 轉換器可產生下一脈衝,猶如自開始程式化該序列以來遞 θ值已在新位準。據此,在使用此類轉換器之—項具體實 A 中纟、、4將等待下部頁程式化完成。當下部頁程式化 70成時’亦可用新遞增值與初始程式㈣脈衝大小重設數 位轉類比轉換器。此將避免程式電壓脈衝值之任何不一致 或大跳躍。 雖然已按照自下部頁程式化轉變至全序列程式化或自下 頁$式化轉變至上部頁程式化來描述具體實施例,但是 具體實施例可應用於單類型程式化操作。舉例而言,在一 貞實把例中,當在全序列程式化中程式化最低經程式 狀〜、時4吏用車又小遞增值Δνρ⑽。當最低經程式化狀態 (例如,狀態A)完成程式化時,可重^Vpgm至較大遞增值 △vPgm2。在此方式中’藉由較小來補償經程式化至狀 態A之記憶體單元所遭遇的增大電容輕合效應。雖然在程 式化方法中未實行任何變更,但是仍然調整遞增值以使最 低位準經程式化狀態更精確。 125212.doc 200837760 在一項具體實施例中,使用統計來預測何時最低經程式 化狀態將完成程式化。在預測較低狀態程式化完成之後, 可變更程式電壓遞增量大小。舉例而言,可在8個脈衝 90%時間之後判定較低狀態完成程式化。程式化演算法可 對於前八個程式電壓脈衝使用較小AVpgmi,並且接著對於 其餘脈衝切換至較大遞增值Δνρρ2。在此方式中,不需要 • 任何用以估定是否下部頁程式化實際上完成並且作出回應 ( 而改變程式電壓遞增值之電路。而是,在指定或預先決定 數量之脈衝或施加程式電壓訊號之後,自動變更遞增值。 此做法可配合前文提及之在切換遞增值之前等待最低經程 式化狀態完成程式化之任何技術一起使用。舉例而言,如 上文所述之在切換至之前施加的預先決定脈衝數量 可係依據何時程式化記憶體單元至狀態Α將完成的統計評 估。 用於當自下部頁程式化轉變至上部頁程式化或全序列程 式化時或當對於第—經程式化狀態之程式化完成時變更遞 增值的技術可與其它程式化技術相組合。在一項具體實施 例中’結合變更之遞增值來利用所謂之粗略/精細程式 - 化。 • P在粗略/精細程式化中,—粗略程式化階段包括嘗試使 2限電壓以較快方式上升,並且相對較不注意到達成緊密 ^ 。豎刀佈。精細程式化階段嘗試使臨限電壓以較慢方 、 x到達目標臨限電壓,因此達成較緊密臨限電壓 刀佈。有關粗略/精細程式化方法論之實例,請參閱美國 125212.doc -42- 200837760 專利案第6,643,188號’該案整份内容以引用方式併入本文 中〇 圖7及圖丨8提供粗略/精細程式化方法論之一項實例之 更詳k二節。圖丨7A及圖丨8 a繪示經施加至控制閘極的程 式化脈衝vpgm。圖17B及圖18B繪示用於正被程式化之記憶 體早疋的位兀線電壓。圖PC及圖18C緣示正被程式化之 忑L、體單το的臨限電壓。此項實例描述使用兩個驗證位準 ( (如圖中VvA1與VvA所示)將記憶體單元程式化至狀態A。最 、、、目私位準係VvA。當記憶體單元的臨限電壓已到達VvA 時,藉由施力口一禁纟電壓至相對應於該記憶體單it的位元 線’以禁止對該記憶體單元進行進—步程式化。舉例而 口位元線電壓可上升至Vinhibit (請參閱圖17B及圖18B)。 但=,當圮憶體單元已到達接近(但小於)目標值VvA的臨限 電壓時,藉由施加一特定偏壓電壓(典型約0.3伏至0.8伏) 至位元線,使在隨後程式化脈衝期間記憶體單元之臨限電 〇 ㈣移減緩。因為於接下來少數程式化脈衝期間減小臨限 電壓偏移速率,所以可使最終臨限電壓份佈較窄。為了實 施此方法,使用一第二驗證位準,其低於Vva之位準。圖 17及圖18中將此第二驗證位準描繪為VvAi。當記憶體單元 .㈣限電壓大於VvA1但仍然小於Vva時,藉由施加一位元 線偏壓Vs (圖18B),對於隨後程式化脈衝,記憶體單元之 臨限電壓偏移將減小。請注意,在此情況中,對於每一狀 態需要兩個驗證操作。對於應用粗略/精細程式化方法論 的每一狀態,以相對應之最終驗證位準(例如,U進行一 125212.doc -43 · 200837760 驗證操作,並且對於每一狀態,以相對應之第二驗證位準 (例如,Vva !)進行一驗證操作。這可增加程式化記憶體單 元所需之總時間。但是,可使用較大之步進大小, 以加速處理程序。 圖17Α、17Β與17C繪示在一個程式化脈衝中其臨限電壓 移動通過VvA〗與VvA的記憶體單元之行為。舉例而言,圖 17C中繪示臨限電壓在“與。之間通過1^與¥仏。因此, 在之前,記憶體單元係處於粗略階段。在“之後,記憶 體單元係處於禁止模式。 圖18A、18B與18C繪示進入粗略與精細程式化階段之記 憶體單元。記憶體單元的臨限電壓在時間“與“之間越過 VvA1。在h之前,記憶體單元係處於粗略階段。在“之後,flow chart. Figure 15 is an exemplary embodiment in which - for a stylized use - a smaller divergence value, and for an upper tiling or column stylization - a larger increment value. In the embodiment of FIG. 15, before issuing the program command to the row control circuit, in step 602, the controller compares the newly received (four) address data with the data currently being programmed by the lower page of the active money unit. . At step 604, if the controller finds a match (which means that the data being newly received by the same memory unit is being stylized), then the method proceeds to step 608. In a specific embodiment, the state machine and/or the row control circuit implements step 6()4. If the address does not match, then the lower page stylization continues, for example, by applying the next program voltage pulse at step 512 of FIG. In addition to step 512, the method of FIG. 14 can also continue after step 6-6 of FIG. If the address data does not match, then at step 6〇8' the controller determines if the lower page stylization is complete. When a predetermined number of program voltage pulses have been applied, when all memory cells to be programmed into state 已 have been successfully programmed to state A, or when a predetermined number of memory cells have been successfully programmed to In state A, the lower page is stylized. If below. After the page is programmed, the upper page data is programmed into the set of memory cells in step 6 i, where the controller issues the upper page command. At step 620, the upper page stylization continues, as shown in step 510' of Figure 14, where the program voltage signal is initialized to (1) and set to AVp^2. When stylizing the upper page of the memory cell, use the second increment value to increase the size of the subsequent stylized pulse. 125212.doc -34- 200837760 Proximity:: The sorrow A reading unit will be attributed to the subsequent stylization Chuan, ^ = pure width threshold voltage distribution, so use a smaller ΔΥρ_ when using a larger page on the step page, and for the upper step 2 = page is not completed (or not connected w white half stop lower page stylized In Figure 14, the controller can be issued by the controller or the state machine - the forced row control circuit stops the command of the lower page data. In step 612, the lower page is programmed: after step 614 The controller sets the full sequence program data command. In the embodiment, the command is input to the buffer 312 via the command latch signal input to the command circuit and is latched by the state machine. The program command triggers the state machine to programmatically select the selected memory. In step 616, the controller can cause the row control circuit to perform a verify operation on the memory unit that has been successfully programmed during the programming of the next page. Has Any memory unit whose target level is reached, the controller locks its memory unit to prevent further stylization. After locking the programmed memory unit, the method proceeds to step 62, which corresponds to the map. Step 5 1 〇 ', in this step the incremental value can be adjusted. In a specific embodiment, greater than Δνρ Εΐ η 1. After stylization, in order to reduce the state Α distribution width, for the lower page stylized or for the stylized minimum Stylize the state of the entity, using a smaller Vpgmi. Because the memory cells in state A have an extra capacitive coupling effect, use a smaller AVpgm value to avoid over-stylization and/or stylize dry 125212.doc -35 - 200837760. In a specific embodiment, step 510 further includes resetting, the initial value of @. For example, if the lower part of the 葙彳 葙彳 ^ ^ 丨 私 私 私 私 私 私 私 私 私The signal is programmed to the upper page or, |Becker is stylized by the king sequence. The value of Vpgm can be reset to a lower value. Too s ^ weight-like value in a specific implementation Vpgm is reset to the same starting value as when starting the lower page stylization. In the specific embodiment, v- is reset to its current value but not to the original value for the lower page stylization. The same low value. In another embodiment, when the full sequence or upper page stylization is started after the completion of the lower page stylization, the last value used when the lower page is programmed is started, with a third increment value. AVp^3 increases or decreases the program voltage signal. In one embodiment, the program voltage signal Vpgm is not reset until the upper page or full sequence is programmed. In a specific embodiment, ¥1)^2 increases the final voltage pulse for the lower page stylization. In another embodiment, for the first upper page or full sequence of pulses, the final voltage pulse for lower page programming is increased by AVpgm3. After the first additional pulse, as described, the value is increased by AVpgm2. AVpgm3 may be greater than AVpgm2 to provide a more stable transition between the two stylized rates resulting from two different step sizes. Step 510' may also include resetting the program counter pc. For example, the maximum number of pulses can be independently built for the lower and upper pages or the full sequence. The counter PC can be reset to 〇 or another value when stylized between the lower page and the upper page or between the lower page and the full sequence stylization. In another embodiment, the counter PC is not reset and the total maximum number of repeated jobs or pulses are used for the overall stylized operation. Roots 125212.doc -36- 200837760 Other variations can be practiced in accordance with specific embodiments. In the variation of the technique described in Figure 15, the increment value for the program voltage pulse is not increased until the programmed minimum programmed state is completed, even after transitioning to full sequence stylization. At step 608, if the controller determines that the lower page is stylized, then the operation proceeds to step 6, as described above, but if the lower page is not yet programmed, the controller stops the lower page stylization process, and Switch to full sequence stylization as shown in steps 612 through 618. However, until the stylization of the lowest programmed state (eg, sadness A) is completed, the program voltage increment value is changed to ΔVpgm2, and all or predetermined memory cells are verified as arrived state A, the program voltage The increment value is more variable than AVpgm2. In a specific embodiment, the program voltage pulse is determined and used to determine the date π is switched to ΔΥρ"2. For example, after switching to full sequence programming, the program voltage increment value can be changed to ΔVpgW after a predetermined number of program voltage pulses have been applied since the initial page programming process. In a specific embodiment, in step (10), the determination of the completion of the lower page stylization can be based on a predetermined number of stylized cycle repetitions (ie, 'predetermine the number of programs, the number of the device's pulse balances, regardless of the memory. It is actually programmed to the appropriate state (for example, state Α). When it is determined in advance that the unit has reached the appropriate state, it can also be determined that the lower page of the program element is 'this can be determined before the entire set of memory cells required to verify that the book page is properly programmed has arrived at the appropriate state. In the case of the case, after resetting the program signal and using the increment value, according to the lower page 125212.doc •37-200837760 material 'some memory units may require additional stylization. This may be handled during the upper page stylization of step 610 by not locking those memory cells that have not been verified as having arrived at state A to be programmed to state a and continuing during the upper page stylization Verification is performed at the status A level. In a particular embodiment where the § 己 体 早 早 can still be programmed to the lowest programmed state, in step 610 to the upper page stylization, until stylized for the lowest programmed state Only when ΔVpgm is changed to ΔVpgw. △Vpgml will continue to be used during the stylization of the upper page until the stylization for the lowest programmed state is verified. For example, § 'After a predetermined number of memory cells to be programmed into state A have arrived at state A, AVpgm can be changed from AVpgml to AVpgm2. In another variation of FIG. 15, if it is determined in step 608 that the lower page stylization is not completed, the controller does not immediately stop the lower page stylization process and convert to full sequence stylization after receiving the upper page data. . Instead, the controller waits until it determines that the lower page is programmed. Figure 16 is a flow chart showing a method of changing the increment value only after receiving the extra data if the stylization of the lower page is completed. This technique toggles the increment value from the next page stylization to the upper page stylization or from the lower page stylization to the full sequence stylization. At step 704, the lower page information is received. The address data is set in step 706, the program data is entered in step 708, and the controller 318 issues a program command in step 710. At step 712, Vpgm is set to its initial value, the program counter is initialized, and 125212.doc -38 - 200837760 Δvpgm is set to ΔΥμμ. At step 714, the lower page material is programmed, for example, as shown in Figure 14 (steps 512 through 530). When the lower page is programmed, the upper page data is received at step 71 6. The address data is set at step 718 and the program data is entered at step 720. Prior to issuing an update program command, in step 722, the controller determines if the lower page stylization is complete. As described above with respect to step 608 of FIG. 15, the various states can be used to determine whether the stylization is complete. It can be determined whether or not the program is to be programmed (all or a predetermined number of memory cells of state A have arrived at state A. In other cases, step 722 can include determining if a predetermined number of pulses have been applied, and if so, determining the lower portion The page is stylized. If the lower page is not programmed, then in step 724, the controller continues the lower page stylization and continues this loop to wait for the lower page to be programmed to complete the TANASY Sigma page. Proceeding to step 726, the controller issues an update program command. In step 8, the controller can reset Vpgm to an initial value of νίη_ΐ2, reset the increment value, and I and heavy: the private counter PC. After the signals (which can be implemented by a digital-to-digital converter that resets a negative Bayesian voltage signal), in step 730, the stylization begins with the new command. In a specific embodiment, The program command set at step 726 is set for upper page stylization. At step 722, the controller can wait until the lower page is stylized until stylized for state A. Successful verification. In this case, the lower page stylized material has been programmed, and only the stylization for the upper page needs to be performed at step 730. In a variation of this technique, the upper page program command can be issued at step 726. And once the lower 125212.doc -39-200837760 page is programmed, the stylization of the upper page is continued immediately in step 730. However, the ΔVpgm is reset in step 728 until the stylization for the lowest programmed state is verified as complete. The e-sense body that has been programmed to state A has been programmed to be sorrowful A will be locked to prevent further stylization, and to be stylized into state A's memory that has not yet reached state A. The unit will suffer further programming and verification. When the state A stylization is verified to be completed, the program voltage increment value can be changed to AVpgm2. In a specific embodiment, when some memory cells can still need to rise to state A. At step 722, it is determined that the lower page is programmed (eg, by determining the number of pulses in advance), and in step 726, the stylization can be converted to full Sequence stylization. For full sequence stylization, verification can be performed before the first program voltage is applied. During full sequence stylization, any memory cells that have reached their final target state during the next page stylization can be locked to block Further stylization. Any memory unit that has not yet arrived in state A to be programmed to state a on the date of completion of the stylization of the lower page will be subject to additional stylization during full sequence stylization. Full sequence stylization is available The status A level performs verification so that during the full sequence of stylization, the target A state memory unit is not allowed to arrive during the lower page stylization period can be privateized to state A. In a specific embodiment, the lower page When the stylization is complete, the transition to full sequence stylization is performed in step 726, but the program voltage increment value is changed to AVpgm2 until the lowest programmed state is verified as complete (if not already verified). A specific embodiment (as shown in Figure 16) that transitions to full sequence stylization (or upper page stylization) and larger increment values after completion of the lower page stylization may be 125212.doc 200837760 provides benefits in a particular implementation . For example, some memory devices will use a digital to analog converter to generate a private voltage signal for a program of different sizes. In a specific embodiment, the digital to analog converter can form part of a column control circuit. Some converters rely on 'a digital input value to generate an analog voltage pulse with the necessary peak value. In some cases, an improved program dust rise value in the middle of the program voltage pulse can present a problem. Changing the increment size can not produce a down-pulse with a peak of the previous pulse plus the new increment value. Instead, some converters can generate the next pulse, as if the value of θ has been new since the program was programmed. According to this, in the case of using such a converter, 纟, 4 will wait for the lower page to be programmed. When the lower page is programmed 70%, the new increment value and the initial program (four) pulse can also be used. Resizing the digital to analog converter. This will avoid any inconsistency or large jump in the program voltage pulse value. Although it has been converted from the lower page to the full sequence or from the next page to the upper page. The specific embodiment is described, but the specific embodiment can be applied to a single type of stylized operation. For example, in a practical example, when the minimum sequence is programmed in the full sequence stylization, The car has a small increment value of Δνρ(10). When the lowest programmed state (for example, state A) is programmed, it can be reset to Vpgm to a larger increment value of ΔvPgm2. In this mode, the program is compensated by a smaller one. The increased capacitance splicing effect encountered by the memory cells in state A. Although no changes have been made in the stylized method, the incremental values are still adjusted to make the lowest level stylized state more accurate. 125212.doc 200837760 In one embodiment, statistics are used to predict when the lowest programmed state will be stylized. After the predicted lower state stylization is completed, the program voltage increment can be changed. For example, it can be 8 pulses at 90. After % time, it is determined that the lower state completes the stylization. The stylized algorithm can use a smaller AVpgmi for the first eight program voltage pulses, and then switch to the larger increment value Δνρρ2 for the remaining pulses. In this way, no • any A circuit that evaluates whether the lower page is stylized and actually responds (and changes the program voltage increment value. Instead, the increment value is automatically changed after a specified or predetermined number of pulses or a program voltage signal is applied.) Stylize with the lowest programmed state before switching the increment value as mentioned above Any technique is used together. For example, the number of predetermined pulses applied before switching to as described above may be based on a statistical evaluation of when the staging memory unit to state will be completed. The technique of changing the increment value when transitioning to the upper page stylization or full sequence stylization or when stylized for the first stylized state can be combined with other stylization techniques. In one embodiment, 'incorporating changes The increment value takes advantage of the so-called coarse/fine program--. • In coarse/fine stylization, the rough stylization phase involves attempting to make the 2-limit voltage rise in a faster manner and relatively less noticeable. Vertical knives. The fine stylization phase attempts to bring the threshold voltage to the target threshold voltage at a slower side, thus achieving a tighter threshold voltage knives. For an example of a rough/fine stylized methodology, see U.S. Patent No. 125,212.doc-42-200837760, Patent No. 6,643,188, the entire disclosure of which is hereby incorporated by reference herein in An example of a methodology is more detailed in Section II. Figure 7A and Figure 8a illustrate the programmed pulse vpgm applied to the control gate. 17B and 18B illustrate the bit line voltage for the early memory of the memory being programmed. The graph PC and Fig. 18C show the threshold voltage of the 忑L and the body το which are being programmed. This example describes the use of two verification levels (as shown in VvA1 and VvA in the figure) to program the memory unit to state A. The most, and the private level VvA. When the threshold voltage of the memory unit When the VvA has been reached, the voltage is applied to the bit line corresponding to the memory unit by the force-carrying port to prohibit the step-by-step programming of the memory unit. For example, the bit line voltage can be Rise to Vinhibit (see Figure 17B and Figure 18B). But =, when the memory cell has reached a threshold voltage close to (but less than) the target value VvA, by applying a specific bias voltage (typically about 0.3 volts) Up to 0.8 volts to the bit line, causing the threshold of the memory cell to shift during the subsequent stylized pulse. Because the threshold voltage shift rate is reduced during the next few stylized pulses, the final The threshold voltage distribution is narrow. To implement this method, a second verification level is used, which is lower than the Vva level. This second verification level is depicted as VvAi in Figures 17 and 18. When the memory unit (4) When the voltage limit is greater than VvA1 but still less than Vva, by applying One bit line bias voltage Vs (Fig. 18B), the threshold voltage offset of the memory cell will decrease for subsequent stylized pulses. Note that in this case, two verify operations are required for each state. Apply each state of the coarse/fine stylized methodology to the corresponding final verification level (eg, U performs a 125212.doc -43 · 200837760 verification operation, and for each state, the corresponding second verification bit A verification operation (eg, Vva!) is performed. This increases the total time required to program the memory unit. However, a larger step size can be used to speed up the processing. Figure 17Α, 17Β, and 17C In a stylized pulse, its threshold voltage moves through the behavior of the VvA and VvA memory cells. For example, the threshold voltage is shown in Figure 17C as being between 1 and ¥. Previously, the memory cell was in a rough phase. After "the memory cell is in the disable mode. Figures 18A, 18B and 18C show the memory cell entering the coarse and fine stylization phase. Voltage between the time "and" crossed Vval. Prior H, system memory cell in the coarse phase. "Behind,

t3與t4之間,記憶體單元的臨限電壓越過VM。因此,藉由Between t3 and t4, the threshold voltage of the memory unit crosses the VM. Therefore, by

一步程式化。 nhibit ’禁止對該記憶體單元進行進Step by step. Nhibit ‘ prohibits the memory unit from entering

的一項具體實施例中,可 細耘式化。但是,在變更遞增值 可結合遞增值之切換來使用粗略/ 125212.doc -44 - 200837760 精、’田耘式化。舉例而言’舉例而言 使用單個最終驗證位準, 式化下#頁時可 態A。當在辦大# 疋°己體早元已抵達狀 頁進行之後進行全序㈣式化或對於上部 實施例中,心=動粗略/精細程式化。在一項具想 能你 、序]或上部頁程式化期間,對於每一狀 心,使用粗略及精細驗證位 對於所㈣〜土 +长”匕具體實施例中,僅 …吏用粗略及精細驗證位準。舉例而言,可僅 證位準^與狀SC’或僅對於狀紅’制粗略及精細驗 土^圖解及說明的目,前文已提出本發明的實施方式。 ^非意欲詳盡說明本發明或使本發明限定於揭示的確切形 2可按照前面的講授進行許多修改及變化。選取的具體 A例係為了最佳地解說本發明的原理及其實務應用,使 熟悉此項技術者以各種具體實施例最佳地運用本發明,並 且各種修改皆適用於所考量的特定料。本發明範•擬藉 由隨附的申請專利範圍予以定義。 9 【圖式簡單說明】 圖1繪示NAND串的俯視圖。 圖2%示圖12Nand串的同等電路圖。 圖3繪示三個NAND串的電路圖。 圖4繪示非揮發性記憶體系統之具體實施例的方塊圖。 圖5%示記憶體陣列的示範性組織。 圖6緣示一組示範性程式電壓訊號。 圖7繪示一組示範性臨限電壓分佈及一全序列程式化處 125212.doc -45- 200837760 理程序。 圖8繪示一組示範性臨限 & 壓分佈及-兩切程(two- pass)處理程序。 、 圖9繪示用以描述在一頊In a specific embodiment, it can be refined. However, the change in the increment value can be combined with the increment value to use the rough / 125212.doc -44 - 200837760 fine, 'field. For example, by using a single final verification level, the # page can be used to determine the state A. The full-sequence (four) formula is performed after the process has been reached. For the upper embodiment, the heart = dynamic coarse/fine stylization. During a stylization of a program, a sequence, or an upper page, for each centroid, use a rough and fine verification bit for the (four) ~ soil + long" 匕 specific examples, only ... use coarse and fine For example, the embodiment of the present invention may have been proposed in the foregoing, and the foregoing description has been made for the purpose of illustration and description only for the form of SC' or only for the shape of red. The invention may be modified or modified in accordance with the foregoing teachings. The specific example A is chosen to best explain the principles and practical applications of the present invention. The present invention is best utilized in various embodiments, and various modifications are applicable to the specific materials considered. The invention is intended to be defined by the scope of the accompanying claims. 9 [Simple Description] Figure 1 Figure 2 is a circuit diagram of three NAND strings. Figure 4 is a block diagram of a specific embodiment of a non-volatile memory system. Memory Exemplary Organization of Body Arrays Figure 6 illustrates an exemplary set of program voltage signals. Figure 7 illustrates an exemplary set of threshold voltage distributions and a full sequence of stylized sections 125212.doc -45-200837760. A set of exemplary threshold & pressure distribution and two-pass processing procedures are shown. Figure 9 is depicted for use in a

員不乾性非揮發性記憶體系統中 電容電荷耦合效應的表袼。 μ Ά T 圖1 〇繪示用於在相異時問„ κ 一 # 子間間隔期間程式化下部頁 上部頁資料之時序圖。 只貝抖與 圖11緣示根據一項具體眘 體貝施例之程式電壓訊號,見可用 於如圖10所示之程式化下部頁資料與上部頁資料,、 圖12繪示用於程式化下部 1貝貝枓並且接著當接收到上部 頁資料時同時程式化下部頁眘 I貝貝枓與上部頁資料之時序圖。 圖13繪示根據-項具體實施例之程式電壓訊號,盆可用 於如圖12所示之程式化下部頁資料與上部頁資料。 圖晴示根據-項具體實施例之程式化非揮發性記憶體 之方法的流程圖。 圖15繪示用於當接收到用於當前正被程式化之一組,己憶 體單元的額外資料時設定—程式命令之方法的流程圖。〜 ,圖16繪示用於#下部頁程式化完成時轉變至使用較大遞 增值進行全序列程式化時戋 了 Α上邛頁耘式化之方法的流程 圖。 圖與17C繪示程式化處理程序之一項具體實施 例,該程式化處理程序係作為粗略/精細程式化之部分。 圖似、⑽與會示程式化處理程序之一項具體實施 例,該程式化處理程序係作為粗略/精細程式化之部分。 125212.doc -46- 200837760 【主要元件符號說明】 100, 102, 104, 106 電晶體 100CG,102CG, 控制閘極 104CG,106CG 100FG,102FG, 浮動閘極 104FG,106FG 120 第一選擇閘極 120CG 控制閘極 122 第二選擇閘極 122CG 控制閘極 126 位元線 128 源極線 202, 204, 206 NAND 串 220, 230, 240, 250 選擇電晶體(選擇閘極) 222, 224,226, 228,記憶體單元 242, 244, 246, 248, 252 302 記憶體單元之臨限電壓位準增 加至狀態A(圖8) 304 臨限電壓增加至狀態B範圍(圖 8) 306 臨限電壓增加至狀態C範圍内 (圖8) 302 記憶體單元陣列(圖4) 125212.doc -47- 200837760 304 行控制電路(圖4) 306 列控制電路(列控制器或解 器)(圖4) 308 P井控制電路 310 共同源極線控制電路 312 資料輸入/輸出緩衝器 * 314 命令電路 315 ( 控制電路 316 狀態機 318 控制器 330, 332 程式化脈衝 334 第一驗證脈衝 336 弟一驗證脈衝 338 第三驗證脈衝 BLe 偶數位元線 / BLo \ ^ 奇數位元線 PC 程式計數器 SGD 選擇線(選擇閘極汲極線) • SGS 選擇線(選擇閘極源極線) WLO, WL1,WL2, 字線 WL3… A,B,C 臣品限電壓分佈(經程式化狀態) E ^限電壓分佈(經擦除狀態) V inhibit 禁止電壓 碼 125212.doc 48- 200837760The appearance of capacitive charge-coupled effects in non-dry non-volatile memory systems. μ Ά T Figure 1 〇 shows the timing diagram for stylizing the upper page of the lower page during the interval between the „ κ and # 子 intervals. Only the Bayer and Figure 11 are based on a specific caution. For example, the program voltage signal can be used for the stylized lower page data and the upper page data as shown in FIG. 10, and FIG. 12 shows the program for the lower part 1 beibe and then when the upper page data is received. The timing chart of the lower page and the upper page data is shown in Fig. 13. Fig. 13 shows the program voltage signal according to the specific embodiment, and the basin can be used for the stylized lower page data and the upper page data as shown in FIG. A flow chart of a method for stylizing non-volatile memory according to the specific embodiment is shown in Figure 15. Figure 15 is a diagram showing additional data for receiving a memory unit for a group that is currently being stylized. Time setting - a flow chart of the method of the program command. ~ , Figure 16 shows the method for changing the page layout to the upper page when using the larger incremental value for the full sequence programming when the lower page is programmed. Flow chart. Figure and 17C A specific embodiment of the programming program, the stylized processing program is part of a rough/fine stylization. Figure 1, (10) and a specific embodiment of the stylized processing program, the stylized processing program is Rough/fine stylized part. 125212.doc -46- 200837760 [Main component symbol description] 100, 102, 104, 106 Transistor 100CG, 102CG, control gate 104CG, 106CG 100FG, 102FG, floating gate 104FG, 106FG 120 first selection gate 120CG control gate 122 second selection gate 122CG control gate 126 bit line 128 source line 202, 204, 206 NAND string 220, 230, 240, 250 select transistor (select gate) 222, 224, 226, 228, memory unit 242, 244, 246, 248, 252 302 The threshold voltage level of the memory unit is increased to state A (Fig. 8) 304 The threshold voltage is increased to the state B range (Fig. 8 306 The threshold voltage is increased to the range of state C (Fig. 8) 302 Memory cell array (Fig. 4) 125212.doc -47- 200837760 304 row control circuit (Fig. 4) 306 column control circuit (column controller or solution) (Fig. 4) 308 P well control circuit 310 common source line control circuit 312 data input/output buffer * 314 command circuit 315 (control circuit 316 state machine 318 controller 330, 332 stylized pulse 334 first verify pulse 336 brother one verification pulse 338 third verification pulse BLe even bit line / BLo \ ^ odd bit line PC program counter SGD selection line (select gate dipole line) • SGS selection line (select gate source line) WLO , WL1, WL2, word line WL3... A, B, C Price limit voltage distribution (stylized state) E ^ Limit voltage distribution (erased state) V inhibit Prohibited voltage code 125212.doc 48- 200837760

Vinitiall, Vinitiai2 程式電壓訊號Vinitiall, Vinitiai2 program voltage signal

Vpgm 程式電壓Vpgm program voltage

VrA5 VrB, VrC 讀取參考電壓 Vva,Vvb,Vvc 驗證參考電壓 VvA1,Vva 驗證位準(圖17,圖18) V v Pgm 程式化脈衝電壓位準 ^Vpgmi5 ^Vpgm2, △Vpgm3程式電壓遞增值 125212.doc 49-VrA5 VrB, VrC read reference voltage Vva, Vvb, Vvc verify reference voltage VvA1, Vva verify level (Figure 17, Figure 18) V v Pgm Stylized pulse voltage level ^Vpgmi5 ^Vpgm2, △Vpgm3 program voltage increment value 125212 .doc 49-

Claims (1)

200837760 十、申請專利範圍: 1 · 一種非揮發性記憶體系統,其包括: 一組多狀態非揮發性儲存元件; 管理電路,其與該複數個非揮發性儲存元件 ^ 管理電路以下列方式程式化該複數個非針該 件·· 丨土减存元 非揮發性儲存元件 接收一程式化資料至該組多狀態 之請求,200837760 X. Patent application scope: 1 · A non-volatile memory system, comprising: a set of multi-state non-volatile storage elements; a management circuit, and the plurality of non-volatile storage elements ^ management circuit are programmed in the following manner The plurality of non-needle pieces of the non-volatile storage element receive a stylized data to the group of multi-state requests, 施加預先決定數量之程式電壓脈衝至該組多狀態非 揮發性儲存元件’以程式化該f料至該等非揮發性儲存 凡件’該施加包括以-第-遞增值來增大該等程式電壓 脈衝之每一者之一大小,直到抵達該預先決定數量,^ 施加一或多個額外程式電壓脈衝至該組多狀態非揮 發性儲存元件,以完成該資料之程式化,該施加一或多 個額外程式電壓脈衝包括以一第二遞增值來增大該一或 多個額外程式電壓脈衝之每一者之一大小。 2 ·如請求項1之非揮發性記憶體系統,其中·· 該施加預先決定數量之程式電壓脈衝包括使待被程式 化至一第一經程式化狀態、一第二經程式化狀態及一第 三經程式化狀態之非揮發性儲存元件的一臨限電壓上 升; 該施加一或多個額外程式電壓脈衝包括使待被程式化 至该第二經程式化狀態及該第三經程式化狀態之非揮發 性儲存元件的一臨限電壓上升。 125212.doc 200837760 3 ·如請求項1之非揮發性記憶體系統,其中該第二遞增值 大於該第一遞增值。 4·如請求項1之非揮發性記憶體系統,其中: 該管理電路包括一數位轉類比轉換器,該數位轉類比 轉換器提供一程式電壓訊號,該程式電壓訊號包括該預 先決定數量之程式電壓脈衝及該一或多個額外程式電壓 脈衝;及 該官理電路在施加該一或多個額外擦除電壓脈衝之前 重設該程式電壓訊號。 5 ·如請求項4之非揮發性記憶體系統,其中: 該管理電路在施加該一或多個額外擦除電壓脈衝之 前,藉由降低該程式電壓訊號之一峰值至一初始值來重 设该程式電壓訊號。 6 ·如請求項1之非揮發性記憶體系統,其中: 施加該一或多個額外程式電壓脈衝包括施加該等額外 程式電壓脈衝中之一第一額外程式電壓脈衝; 施加該預先決定數量之程式電壓脈衝包括施加該預先 決定數量之程式電壓脈衝中之一最後程式電壓脈衝;及 該第一額外程式電壓脈衝具有一峰值,該峰值實質上 等於該最後程式電壓脈衝之一峰值與該第二遞增值之一 總和。 7.如請求項1之非揮發性記憶體系統,其中: 施加該一或多個額外程式電壓脈衝包括施加該等額外 程式電壓脈衝中之一第一額外程式電壓脈衝; 125212.doc 200837760 、,加,縣決定數量之程式電麵衝包括施加該預先 、、數里之私式電壓脈衝中之一最後程式電壓脈衝;及 第額外轾式電壓脈衝具有一峰值,該修值實質上 等於該最後程式·脈衝之—峰值與—第三遞增值之一 總和。 8_如請求項7之非揮發性記憶體系統,其中·· 該第三遞增值大於該第二遞增值。 9.如請求項7之非揮發性記憶體系統,其中: 該組多狀態非揮發性儲存元件係一組多狀態nand型 快閃記憶體裝置。 10· —種程式化非揮發性儲存裝置之方法,其包括: 接收一帛式化資料至一組多狀態非揮發性儲存元件之 請求; 施加預1決定數量之程式電壓脈衝至一組非揮發性儲 存元件,以程式化該資料至該等非揮發性儲存元件該 施加包括以一第一遞增值來增大該等程式電壓脈衝之每 一者之一大小,直到抵達該預先決定數量; 施加一或多個額外程式電壓脈衝至該組非揮發性儲存 元件,以完成該資料之程式化,該施加一或多個額外程 式電壓脈衝包括以一第二遞增值來增大該一或多個額外 程式電壓脈衝之每一者之一大小。 11·如請求項10之方法,其中·· 該施加預先決定數量之程式電壓脈衝包括使待被程式 化至一第一經程式化狀態、一第二經程式化狀態及一第 125212.doc 200837760 一經程式化狀態之非揮發性儲存元件的一臨限電壓上 升; 該,加-或多個額外程式電壓脈衝包括使待被程式化 至该第二經程式化狀態及該第三經程式化狀態之非揮發 性儲存元件的一臨限電壓上升; 12·如請求項10之方法,其中該第二遞增值大於該第一遞增 值。 13 ·如睛求項1 〇之方法,其中·· 該預先決定數量之程式電壓脈衝及該一或多個額外程 式電壓脈衝係藉由一共同程式電壓訊號予以提供;及 該方法進一步包括在施加該一或多個額外擦除電壓脈 衝之前重設該程式電壓訊號。 14·如請求項1 3之方法,其中·· 重設該程式電壓訊號包括:在施加該一或多個額外擦 除電壓脈衝之前,降低該程式電壓訊號之一峰值至一初 始值。 1 5 ·如請求項1 〇之方法,其中: 施加u亥或夕個額外程式電壓脈衝包括施加該等額外 程式電壓脈衝中之一第一額外程式電壓脈衝; 施加该預先決定數量之程式電壓脈衝包括施加該預先 決定數量之程式電壓脈衝中之一最後程式電壓脈衝;及 該第一額外程式電壓脈衝具有一峰值,該峰值實質上 等於該最後程式電壓脈衝之一峰值與該第二遞增值之一 總和。 125212.doc 200837760 16 ·如請求項1 〇之方法,其中·· 施加该-或多個額外程式電壓脈衝包括施加該等額外 程式電壓脈衝中之一第一額外程式電壓脈衝; 施加該預先決定數量之程式電壓脈衝包括施加該預先 決定數量之程式電壓脈衝中之一最後程式電壓脈衝;及 忒第一額外程式電壓脈衝具有一峰值,該峰值實質上 4於該最後程式電壓脈衝之一峰值與一第三遞增值之一 總和。 17.如請求項10之方法,其中·· 該第三遞增值大於該第二遞增值。 1 8 ·如請求項1 〇之方法,其中·· 該組多狀態非揮發性儲存元件係一組多狀態NAND型 快閃記憶體裝置。 125212.docApplying a predetermined number of program voltage pulses to the set of multi-state non-volatile storage elements 'to program the material to the non-volatile storage items'. The application includes increasing the programs by a -first increment value One of each of the voltage pulses, up to the predetermined number, applying one or more additional program voltage pulses to the set of multi-state non-volatile storage elements to complete the stylization of the data, the application of one or The plurality of additional program voltage pulses includes increasing a size of each of the one or more additional program voltage pulses by a second increment value. 2. The non-volatile memory system of claim 1, wherein the applying a predetermined number of program voltage pulses comprises causing the program to be programmed to a first programmed state, a second programmed state, and a a threshold voltage rise of the third programmed state non-volatile storage element; the applying one or more additional program voltage pulses comprising causing to be programmed to the second programmed state and the third programmed A threshold voltage of the non-volatile storage element of the state rises. 125. The method of claim 1, wherein the second incremental value is greater than the first incremental value. 4. The non-volatile memory system of claim 1, wherein: the management circuit comprises a digital to analog converter, the digital to analog converter providing a program voltage signal, the program voltage signal including the predetermined number of programs a voltage pulse and the one or more additional program voltage pulses; and the logic circuit resets the program voltage signal prior to applying the one or more additional erase voltage pulses. 5. The non-volatile memory system of claim 4, wherein: the management circuit resets by decreasing a peak value of the program voltage signal to an initial value before applying the one or more additional erase voltage pulses The program voltage signal. 6. The non-volatile memory system of claim 1, wherein: applying the one or more additional program voltage pulses comprises applying one of the first additional program voltage pulses of the additional program voltage pulses; applying the predetermined number of The program voltage pulse includes applying a last program voltage pulse of the predetermined number of program voltage pulses; and the first additional program voltage pulse has a peak value substantially equal to one of the peak value of the last program voltage pulse and the second The sum of one of the increment values. 7. The non-volatile memory system of claim 1, wherein: applying the one or more additional program voltage pulses comprises applying one of the first additional program voltage pulses of the additional program voltage pulses; 125212.doc 200837760,, Adding, the county determines the number of programmable electrical impulses including applying one of the pre-, a few of the private voltage pulses of the last program voltage pulse; and the additional 轾 voltage pulse has a peak value, the repair value is substantially equal to the last The sum of the program's pulse-peak and - third increment values. 8_ The non-volatile memory system of claim 7, wherein the third incremental value is greater than the second incremental value. 9. The non-volatile memory system of claim 7, wherein: the set of multi-state non-volatile storage elements is a set of multi-state nand type flash memory devices. 10. A method of staging a non-volatile storage device, comprising: receiving a request for a data to a set of multi-state non-volatile storage elements; applying a predetermined number of programmed voltage pulses to a set of non-volatile a storage element for staging the data to the non-volatile storage elements, the applying comprising increasing a size of each of the program voltage pulses by a first increment value until the predetermined amount is reached; One or more additional program voltage pulses are applied to the set of non-volatile storage elements to complete programming of the data, the applying one or more additional program voltage pulses comprising increasing the one or more by a second incremental value One of the extra program voltage pulses of each size. 11. The method of claim 10, wherein the applying a predetermined number of program voltage pulses comprises causing the program to be programmed to a first programmed state, a second programmed state, and a 125212.doc 200837760 A threshold voltage of the non-volatile storage element in a programmed state is increased; the addition or the plurality of additional program voltage pulses includes a program to be programmed to the second programmed state and the third programmed state A threshold voltage of the non-volatile storage element rises; 12. The method of claim 10, wherein the second incremental value is greater than the first incremental value. 13) The method of claim 1 wherein: the predetermined number of program voltage pulses and the one or more additional program voltage pulses are provided by a common program voltage signal; and the method further comprises applying The program voltage signal is reset before the one or more additional erase voltage pulses. 14. The method of claim 13 wherein resetting the program voltage signal comprises decreasing a peak value of the program voltage signal to an initial value prior to applying the one or more additional erase voltage pulses. The method of claim 1, wherein: applying an additional voltage pulse comprises applying a first additional program voltage pulse of the additional program voltage pulses; applying the predetermined number of program voltage pulses Including applying a last program voltage pulse of the predetermined number of program voltage pulses; and the first additional program voltage pulse has a peak value substantially equal to one of a peak value of the last program voltage pulse and the second increment value A sum. The method of claim 1, wherein applying the one or more additional program voltage pulses comprises applying one of the first additional program voltage pulses of the additional program voltage pulses; applying the predetermined amount The program voltage pulse includes applying a last program voltage pulse of the predetermined number of program voltage pulses; and the first additional program voltage pulse has a peak value substantially equal to one of the last program voltage pulses and one peak The sum of one of the third increment values. 17. The method of claim 10, wherein the third incremental value is greater than the second incremental value. 1 8 The method of claim 1, wherein the multi-state non-volatile storage element is a multi-state NAND type flash memory device. 125212.doc
TW096137701A 2006-10-10 2007-10-08 Non-volatile memory system and a method of programing non-volatile storage TWI385665B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US11/548,267 US7450426B2 (en) 2006-10-10 2006-10-10 Systems utilizing variable program voltage increment values in non-volatile memory program operations
US11/548,264 US7474561B2 (en) 2006-10-10 2006-10-10 Variable program voltage increment values in non-volatile memory program operations

Publications (2)

Publication Number Publication Date
TW200837760A true TW200837760A (en) 2008-09-16
TWI385665B TWI385665B (en) 2013-02-11

Family

ID=39046765

Family Applications (1)

Application Number Title Priority Date Filing Date
TW096137701A TWI385665B (en) 2006-10-10 2007-10-08 Non-volatile memory system and a method of programing non-volatile storage

Country Status (3)

Country Link
KR (1) KR101013200B1 (en)
TW (1) TWI385665B (en)
WO (1) WO2008045805A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI455134B (en) * 2009-03-23 2014-10-01 Toshiba Kk Nonvolatile semiconductor memory device

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8953386B2 (en) 2012-10-25 2015-02-10 Sandisk Technologies Inc. Dynamic bit line bias for programming non-volatile memory
EP4181131B1 (en) * 2021-11-16 2024-04-03 Samsung Electronics Co., Ltd. Operation method of memory device

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5903495A (en) * 1996-03-18 1999-05-11 Kabushiki Kaisha Toshiba Semiconductor device and memory system
EP1249842B1 (en) * 2001-04-10 2009-08-26 STMicroelectronics S.r.l. Method for programming nonvolatile memory cells with program and verify algorithm using a staircase voltage with varying step amplitude
ATE449465T1 (en) * 2001-04-12 2009-12-15 Juniper Networks Inc ACCESS NOISE CANCELLATION IN A DIGITAL RECEIVER
US6522580B2 (en) * 2001-06-27 2003-02-18 Sandisk Corporation Operating techniques for reducing effects of coupling between storage elements of a non-volatile memory operated in multiple data states
US7177197B2 (en) * 2001-09-17 2007-02-13 Sandisk Corporation Latched programming of memory and method
TWI292914B (en) * 2002-01-17 2008-01-21 Macronix Int Co Ltd
US7136304B2 (en) * 2002-10-29 2006-11-14 Saifun Semiconductor Ltd Method, system and circuit for programming a non-volatile memory array
US6937520B2 (en) * 2004-01-21 2005-08-30 Tsuyoshi Ono Nonvolatile semiconductor memory device
US7092290B2 (en) * 2004-11-16 2006-08-15 Sandisk Corporation High speed programming system with reduced over programming
US7173859B2 (en) * 2004-11-16 2007-02-06 Sandisk Corporation Faster programming of higher level states in multi-level cell flash memory
US7301817B2 (en) * 2005-10-27 2007-11-27 Sandisk Corporation Method for programming of multi-state non-volatile memory using smart verify

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI455134B (en) * 2009-03-23 2014-10-01 Toshiba Kk Nonvolatile semiconductor memory device

Also Published As

Publication number Publication date
TWI385665B (en) 2013-02-11
KR20090089310A (en) 2009-08-21
WO2008045805A1 (en) 2008-04-17
KR101013200B1 (en) 2011-02-10

Similar Documents

Publication Publication Date Title
EP1812932B1 (en) High speed programming system with reduced over programming
EP1866928B1 (en) Faster programming of higher level states in multi-level cell flash memory
KR101632367B1 (en) Multi-pass programming for memory with reduced data storage requirement
EP1738374B1 (en) Variable programming of non-volatile memory
JP4931915B2 (en) Method for initiating program voltage shift as non-volatile memory is repeated
KR101595045B1 (en) Multi-pass programming for memory using word line coupling
KR101600551B1 (en) Programming non-volatile memory with high resolution variable initial programming pulse
KR101020812B1 (en) Programming defferently sized margins and sensing with compensations at select states for improved read operations in non-volatile memory
US7450426B2 (en) Systems utilizing variable program voltage increment values in non-volatile memory program operations
TWI337746B (en) System and method for non-real time reprogramming of non-volatile memory to achieve tighter distribution of threshold voltages
JP5174829B2 (en) Reading of non-volatile memory cell in consideration of storage state of adjacent memory cell
CN101405813A (en) Method for non-real time reprogramming of non-volatile memory to achieve tighter distribution of threshold voltages
CN102138183A (en) Selective erase operation for non-volatile storage
JP2013524400A (en) Saw-shaped multi-pulse programming for reducing program noise in memory
EP1971984B1 (en) Continued verification in non-volatile memory write operations
EP2067142B1 (en) Faster programming of highest multi-level state for non-volatile memory
JP4995264B2 (en) Reduction of program disturbance during reading
TW200837760A (en) Systems utilizing variable program voltage increment values in non-volatile memory program operations
JP4995265B2 (en) Reduction of program disturbance during reading
JP4820879B2 (en) Method for reading non-volatile memory by effectively controlling non-selected word lines

Legal Events

Date Code Title Description
MM4A Annulment or lapse of patent due to non-payment of fees