TW502171B - Data transmitting system - Google Patents

Data transmitting system Download PDF

Info

Publication number
TW502171B
TW502171B TW90106210A TW90106210A TW502171B TW 502171 B TW502171 B TW 502171B TW 90106210 A TW90106210 A TW 90106210A TW 90106210 A TW90106210 A TW 90106210A TW 502171 B TW502171 B TW 502171B
Authority
TW
Taiwan
Prior art keywords
bus
data
aforementioned
address
data transmission
Prior art date
Application number
TW90106210A
Other languages
Chinese (zh)
Inventor
Takayuki Hiroya
Original Assignee
Casio Computer Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Casio Computer Co Ltd filed Critical Casio Computer Co Ltd
Application granted granted Critical
Publication of TW502171B publication Critical patent/TW502171B/en

Links

Landscapes

  • Bus Control (AREA)
  • Information Transfer Systems (AREA)

Abstract

I/O bus 20 which is a bus structure that combines channel type/bus type consists of I/O common bus that is connected to I/O device 21-23 and of channel interface signal line that is independently installed by each of I/O device. Each of the channel interface signal line comprises clock pulse SCLK. The data transmitting between each of I/O device and bus arbitrator 15 and the clock pulse SCLK is carried out simultaneously. In the I/O common bus, commands, addresses, and data transmit CAD bus with 8 bit width based on a time distribution and comprises, in the CAD bus, signal line used as distinguishing command from data. Because there is installed with a distinguishing signal for the command/data, I/O device can independently require data transmitting so as to achieve flexible access control.

Description

502171 五、發明説明(1 ) [發明所屬之技術範疇] 本發明是有關使用改良的匯流排構造的資料傳輸系 統。 [習知的技術] 習知在攜帶式資訊終端機等的電子裝置所使用的I/O 裝置用的匯流排構造,一般爲眾所周知的有(υ匯流排 型、(2i通道型的2種。 匯流排型爲複數的I/O裝置,以共通的匯流排來連接 的構造,除了位址匯流排、資料匯流排之外,對於時鐘 信號線而言也是在複數的I/O裝置之間共用。因此可以 減少信號配線數量,但是裝置間共用了各信號的緣故, 爲了改善所產生的信號波型的不穩定性,而必須採取配 線的對策,對於頻率的提升及消耗電力上造成不利。 一方面,通道型對於各I/O裝置在位址匯流排、資料 匯流排、時鐘信號線等爲獨立配設的構造。各個I/O裝 置可以進行彈性的存取控制,而且與匯流排型比較之下 ,雜訊對策較爲簡單,也很容易在性能上的提升的特徵 存在。可是相反的,具有信號配線數增加的問題存在。 攜帶式資訊終端機等的小型電子機器中,因爲部件的 實際組裝面積有限制,因此對信號配線數有需要極力的 抑制在最少。可是,因爲如此而使用上述的匯流排型的 配線構造時’全部的裝置上須經常的供給時鐘脈衝的要 因,所以電力消耗量造成增大的情形。因此即使要採用 通道型的配線構造時’爲了使信號線配線數可以減少, 502171 五、發明説明(2 ) 因而要求新的匯流排構造。 以減少匯流排的信號配線數的手法而言,將位址及資 料在同一匯流排上,採取時間分配爲眾所周知的事實。 可是,習知在位址與資料的輸出順序及該輸出時序以預 先決定,因此必須經常以該時序來存取。所以,由於雜 訊產生序列偏差的情形時,可以會引起錯誤動作的問題 產生。且,位址相位與資料相位要求必須連續,所以 用以存取將某個I/O裝置的位走相位與資料相位予以分 割,在這之間,執行對於其他的I/O裝置的存取,或是 接受來自其他的I/O裝置的存取要求的控制上很困難。 [發明槪要] 本發明乃鑑於本情況而尋求解答的內容,經由實現新 的匯流排構造,提供予要求小型及低消耗電力的攜帶式 資料終端機等的小型電子機器適合的資料傳輸系統爲目 本發明的資料傳輸系統,' P在匯流排仲裁器與複數的 I/O裝置間傳送資料的資料傳輸系統,其特徵爲,在前 述的各個I/O裝置,設置了要求匯流排所有權的需求裝置; 對前述匯流排仲裁器中,設置了允許前述匯流排所有權 的接地裝置;其中前述匯流排仲裁器與前述個I/O裝置 中具有:指令及位址及數據資料在同一匯流排中輸出的 裝置;使用前述匯流排傳送資料時,向對方傳送時鐘脈 衝的裝置;接收由對方傳來時鐘脈衝再接收由對方傳送 而來的資料之後的受信裝置;前述指令與位址及資料在 »4- 502171 五、發明説明(3 ) 同 — 匯 流 排 輸 出 時 將 別 在 此匯 流 排 所輸 出的是指令 或 資 料 的 信 號 輸 出 的 裝 置 〇 對 於 此 資 料 傳 輸 裝 置 採 用 在同 匯 流排 上輸出指令 及 位 址 及 貝 料 的 架 構 在 匯 流 排仲 裁 器 與複 數的I/O裝 置 間 的 資 料 傳 送 所 使 pzf 用 的 匯 流 排上 指 令與 位址與資料 以 時 間 分 配 來 傳 达 〇 此 種 情 形 下, 在 匯 流排 所輸出的是 指 令 或' 料 係 以 由 傳 送 處 輸 出區 別 的 專用 信號的方式 來 指 定 〇 因 此 各 個 指 令 及 資 料 即 使 未能 以 習 知所 決定的時序 來 傳 送 時 也 可 以 正 常 的 進 行 資料 傳 送 ,所 以對於雜訊 的 序 列 偏 差 可 以 提 升 耐 性 〇 而 且, 位址 相位 和資料相位 並 不 需 要 連 續 的 執 行 , 所 以 分 割用 於 存 取某 個I/O裝置 的 位 址 相 位 和 資 料 相 位 在 這 之間 和 其他 的I/O裝置 之 間 的 資 料 傳 送 也 可 以 執 行 〇 而 且 要 區 別 爲 指 令 或 資 料 的信 號 是 由I/O裝置的輸 出 5 該 I/O 裝 置 輸 出 到 匯 流 排 的內 容 對於 匯流排仲裁 器 可 以 明 顯 的 標 示 通 知 爲 指 令 或資 料 I/O裝置可以自 行 的 發 行 指 令 進 行 資 料 的 傳 送 〇特 別 是 ,在 本資料傳送 系 統 中 接 收 端 與 從 傳 送 端 傳 送而 來 的 時鐘 脈衝同步以 接 收 資 料 的 通 道 型 構 造 , 因 爲 是在 指 令 及位 址及資料在 同 一 匯 流 排 上 輸 出 的 匯 流 排 型 架構 的 組 合方 式來使用, 所 以 從 I/O 裝 置 的 資 料 傳 送 是起 因 於 對該 I/O裝置讀 取 資 料 所 形 成 的 5 或 是 從 I/O 裝置 白 己 本身 的指令所形 成 的 需 要 判 斷 的 機 構 才 可 以 • 5- ,所 以 採 用了 依專用的信 71 五、發明説明(4 ) 號,從I/O裝置通知位址仲裁器端的架構,可以更有彈 性的實現存取控制。 [圖面的簡單說明] 第1圖顯示有關本發明的實施型態使用資料傳輸系統 的資訊處理裝置的系統架構方塊圖。 第2圖顯示同樣的實施型態匯在匯流排仲裁器及1/0 裝置的~端子規格圖。 第3圖說明在相同實施型態所使用的3時鐘脈衝命令 及4時鐘脈衝命令的圖形。 第4 _顯示在相同實施型態所使用的指令的構造。 第5圖由於在相同實施型態所使用的指令,顯示可指 定的資料傳輸種類。 第6圖顯示在相同實施型態所使用的傳輸電路部位的 具體電路架構圖。 第7圖對於相同實施型態,顯示CPU對I/O裝置寫入 資料動作時的時序流程圖。 第8圖在相同實施型態,作爲說明要求重試時序的時 序流程圖。 第9圖在相同實施型態,CPU從I/O裝置讀取資料時 ,顯示動作的時序流程圖。 第1 0圖在相同實施型態,依來自I/O裝置的讀取需 求,顯示資料傳輸動作時的時序流程圖。 第11圖在相同實施型態,依來自I/O裝置的寫入需 求,顯示資料傳輸動作時的時序流程圖。 五、發明説明(5 ) 弟1 2 Η在相问貫施型態,顯示資料傳輸途中停止動 作的時序流程圖。 第1 3圖在相同實施型態,顯示ί/〇裝置切換動作的 時序流程圖。 第1 4圖在相同實施型態,顯示1/()裝置切換動作之 其他範例的時序流程圖。 [實施森例] 在第1圖中,顯示是有關本發明之實施型態,使用資 料傳輸系統的資訊處理裝置的系統架構。此資訊處理是 作爲攜帶式資訊終端機等使用,如圖所示,具備了 CPU 1 1、RAM 12、ROM 13、閘極陣列(Gate Array) 14、及複 數的I/O裝置21〜23。 I/O裝置21〜23爲,例如磁碟機(FDD)、硬碟機(HDD) 、音響裝置、數位相機、通信控制器等的各種周邊裝 置。 閘極陣列(Gate Array) 14爲’主機端的記憶體匯流排 10與I/O裝置21〜23端的I/O匯流排20的雙方面連接, 因應從CPU 11來的存取要求以進行I/O裝置21〜23的存 取控制,或是因應來自I/O裝置21〜23的記憶體存取需 求,而進行記憶體存取等的控制。對於此閘極陣列(Gate Array)14作爲周邊裝置,例如驅動液晶顯示裝置(圖中 未顯示)的顯示驅動電路一起組裝在其中。而且閘極陣 列14對於I/O裝置21 ~23,包含了控制匯流排切換的匯 流排仲裁器(Bus Arbitrator)電路15。 -7- 502171 五、發明説明(6 ) 主機端的記億體匯流排1 〇 ’如圖所示’包含32位元 寬的資料匯流排、26位元寬的位址匯流排、與控制信號 。控制信號包含晶片選擇信號CS,與讀取信號Read, 與3條寫入信號Write。Write信號中顯示,1位元組寫 入、2位元組寫入、4位元組寫入。 CPU 1 1藉由使用記憶體匯流排1 0來發出記憶體位 址,而存取RAM 12、ROM 13。CPU1 1對於I/O裝置 2 1〜23,也是使用匯流排記憶體1 0來發出記億體位址而 存取。亦即,從CPU 1 1來看全部的I/O裝置21〜23,成 爲在記憶體地圖上所展開的硬體架構,各I/O裝置作爲 記憶體地圖I/O來使用。 I/O裝置21〜23所連接的I/O匯流排20爲,前述通道 型/匯流排型倂用的匯流排架構,係由對I/O裝置21〜23 共通連接的I/O共通匯流排,與每個I/O裝置由獨立設 置的通道介面信號線所構成。記憶體匯流排丨〇的資料 傳輸寬度爲32位元,相對的I/O共通匯流排的資料傳輸 寬度爲8位元。亦即,在I/O共通匯流排中,包含以時 間分配來傳送指令與位址與資料之8位元寬的CAD匯流 排,而與各;[/0裝置之間的資料傳輸係介由CAD來實 行。 因此,在CPU 11和各I/O裝置之間的資料傳輸之 際,執行將2 6位元寬的位址及3 2位元寬的平行資料各 以8位元單位分割,再以序列的傳送到i/C)裝置端作序 歹〖J /平行變換處理’或是從I/O裝置端以8位元資料單 502171 五、發明説明(7 ) 位’將序列轉送的位址或資料彙整爲32位元寬而轉送 到記憶體匯流排1 0端以執行序列/平行變換處理。 匯流排仲裁器(Bus Arbitrator)電路15中,準備了通道 1〜3的3個通道。各通道藉由通道介面信號線連接所對 應的I/O裝置。 匯流排仲裁器電路1 5及各I/O裝置的端子規格如第2 圖所示'亦即,I/O共通匯流排爲13條、8條的CAD信 號線CAD[7:〇]、1條的指令/位址狀態信號線CA、2條 的承認信號線ACK[ 1:0]、1條的OFF信號線OFF、1條 重設信號線RESETB所構成。 指令/位址狀態信號線CA,作爲區別目前在CAD上所 輸出的爲指令或資料的信號。指令或位址在輸出的期間 中、以其輸出端(主控端)將狀態信號CA設定爲作用 狀態"H"。而資料輸出期間中將狀態信號CA設定爲非作 用狀態。 承認信號線ACK[ 1:0]爲表示指令、位址或資料接收端 (slave)的收信狀態的回應信號、ACK = ”01"爲表示指令、 位址> 或資料的循環正常結束(肯定回應Acknowledge ) ,而且,ACK" 10"爲表示指令,位址或資料的循環重送 需求(Retry)。此承認信號線ACK〔l:〇]在第幾個循環輸 出,係根據各個位址相位及資料相位來固定。 OFF信號線0FFB及重設信號線RESETB爲從各個匯 流排仲裁器電路15往I/O裝置端輸出的信號,表示I/O 裝置的切斷及重設。 502171 Ί 五、發明説明(δ ) 而且對於各通道介面信號線設置有源極同步時鐘脈撞? 信號SCLK、插入信號INTB、匯流排使用權需求信號 BREQ、匯流排使用權許可信號BGNT各1條。 源極同步時鐘脈衝信號SCLK爲輸出指令、位址、或 資料時的轉送用同步時鐘脈衝,使用於CAD匯流排上 的資料(指令、位址、資料)樣本。源極同步時鐘脈衝 信號StLK爲從主控制輸出。收信端的從動裝置之資料 收信,是依來自主控端的源極同步時鐘脈衝信號SCLK 而開始,以與該源極同步時鐘脈衝信號SCLK同步地接 收指令、位址、或資料。 匯流排使用權需求信號BREQ爲對於匯流排仲裁器電 路15要求1〇共通匯流排的使用權之信號,係從1/〇裝 置對匯流排仲裁器電路1 5發送。匯流排仲裁器電路! 5 爲調停來自I/O裝置2 1〜23各個的使用權需求信號 BREQ ’以對應賦予匯流排使用權的;[/〇匯流排發送匯流 排使用權許可信號BGNT。匯流排使用權需求信號BREQ 及匯流排使用權許可信號BGNT、各個皆爲負邏輯的信 號。 因此形成匯流排仲裁器電路15端爲13 + (3x4) = 25根 ’而各I/O裝置端爲13 + 4=17根的端子。 在匯流排仲裁器電路15與I/O裝置21〜23之間,資料 的交換經常以主控制端從動裝置的2個關係來產生,在 傳輸資料的輸出端確保匯流排權之後控制轉送用的時鐘 脈衝SCLK而進行資料轉送。;[/0裝置具有各個匯流排 -1 〇- 502171 五、發明説明(9 ) 需求(BREQ)電路,只有取得匯流排仲裁器電路15許可 的I/O裝置可以輸出資料。 <指令與位址> 其次參考第3或5圖,對於以8位元寬的CAD[7:0]轉 送的指令及位址予以說明。 對於本實施型態的匯流排系統,準備了如第3 (a)圏的 3時鐘脈衝命令,與如第3 (b)_的4時鐘脈衝命令。3 時鐘脈衝命令是由8位元的指令,及其後續分爲2次轉 送的位址(A)[15:8]、A[7:0]所構成。通常使用這3個時 鐘脈衝命令,以3個時鐘脈衝命令所表現的記憶體位址 空間成爲存取對象。4時鐘脈衝命令是在將記憶體位址 空間予以擴充的場合時所使用的命令,係由8位元的指 令,及其後續分爲3次轉送的位址(A[23:16]、Α[15:8]、 Α[7:0])所構成。 要使用3時鐘脈衝命令或使用4時鐘脈衝命令,是由 8位元的指令內所定義的擴充位元來指定。第4圖顯示 8位元的指令構造。 CAD[6]〜CAD[3]的4位元,係作爲指示資料轉送種類 的指令部(CM[3]~CM[0])。CM[3]〜CM[0]的組合與指令內 容之間的關係範例如第5圖所示。在執行2位元組以上 的連續資料轉送時,通常在指令收信端每1個位元組的 讀/寫的位址各+ 1的自動增加,但在本實施型態中,也 準備了在位址沒有增加的情況下,執行2位元組以上的 連續資料轉送命令。而旦,由CM[3]〜CM[0],也可以指 »11- 502171 五、發明説明(10 ) 指示中止執行中的資料轉送。 CAD[2]爲指示讀/寫的存取種類者,caD[2] = "0”表示 閱讀存取,CAD[2] = "1"表示寫入存取。 CAD[1]對位址擴充的有無,亦即指示現在的指令及位 址爲3時鐘脈衝命令,或爲4時鐘脈衝命令的指示位址 擴充位元。CAD[l] = "〇”表示3時鐘脈衝命令,CAD[1] = ” r爲f時鐘脈衝命令。 C A D [ 0 ]係作爲指不接繞於指令後之位址先頭位元的內 容’在3時鐘脈衝命令時作爲A [ 1 6 ],或是在4時鐘脈 衝命令時作爲A[24]來使用,因此,實際上,8位元的指 令形成也包含了位址資訊。 <電路架構> 其次,參考第6圖,對於本匯流排系統所使用的資料 轉送電路部的具體電路架構予以說明。由於1/0裝置21 〜23的匯流排介面都是相同的,所以在第6圖中以I/O 裝置21爲代表來表示。而且,通道介面信號線之中, 對於插入信號予以省略。 在匯流排仲裁器電路1 5中除了匯流排調停功能之外 ,包含了控制與I/O裝置之間資料轉送的功能。亦即, 匯流排仲裁器電路1 5如圖所示,具備了 FIFO緩衝器 151、指令/位址/資料合成電路(CAD合成電路)152、 轉送控制電路153、及DMA電路154。 FIFO緩衝器1 5 1是與記憶體匯流排1 〇之間的介面用 緩衝器,爲了實現前述平行/序列變換、序列/平行變 -12- 502171 五、發明説明(11 ) 換’執行來自記憶體匯流排1 〇的控制信號、位址、資 料的收信及積蓄、以及執行從CAD匯流排的指令、位 址、資料的收信及積蓄。例如,CPU 1 1對I/O裝置執行 寫入存取時,位址與資料與寫入信號成爲保持在FIFO 緩衝器1 5 1中。 CAD合成電路152是使保持在FIFO緩衝器151中的 儲存資辛f搭乘在CAD匯流排上,以變換爲8位元序列資 料的內容者,以此CAD合成電路152與FIFO緩衝器151 來構成前述的平行/序列變換的介面。CPU 1 1對I/O裝 置執行32位元的資料寫入存取時,CAD合成電路152, 依據寫入信號與位址,產生前述的3時鐘脈衝或4時鐘 脈衝命令(CA)的同時,將32位元寬的資料分割成爲每 、一個8位元而變換爲D0〜D3的3位元組的序列資料。 傳輸控制電路1 53係執行與各I/O裝置之間的實際資 料傳輸控制。對於此傳輸控制電路1 53,如圖所示,設 置了匯流排接地(BGNT)電路201、時鐘脈衝產生電路 202、CAD轉送電路203、指令解釋電路204、及錯誤偵 測器205。 匯流排接地(BGNT)電路201係具有每個I/O裝置 2 1〜23及匯流排接地(BGNT)電路201之間,以一對一連 接的匯流排使用權需求信號BREQ1〜3,及匯流排使用權 作許可信號BGNT 1~3,使用這些的使用權需求信號 BREQ1〜3及匯流排使用權許可信號BGNT 1〜3來執行I/O 共通匯流排的使用權調停。對於I/O共通匯流排的匯流 -13 - 502171 五、發明説明() 共通匯流排的使用權調停。對於1/()共通匯流排的匯流 排解放時機’以指令/位址相位、資料相位在哪個循環 解放已經固定化’如有其他匯流排使用權需求信號 BREQ的話’則在該時機執行匯流排權的切換。 時鐘脈衝產生電路202係於資料輸出時,用於產生前 述之源極同步時鐘脈衝信號SCLK,此源極同步時鐘脈 衝信號SCLK也可以作爲CAD轉送電路203的動作時鐘 脈衝來利用。源極同步時鐘脈衝信號SCLK的頻率可以 配合對方的I/O裝置的轉送處理速度設定爲可變。而 且,時鐘脈衝產生電路202之內部具有切換電路,其在 收信時’把從I/O裝置所接收的源極同步時鐘脈衝信號 SCLK,給予CAD傳輸電路202或錯誤偵測器205以作 爲資料載入時鐘脈衝。 CAD傳輸電路203是藉由CAD匯流排以執行與I/O裝 置之間的資料轉送之用,其擁有前述的指令/位址狀態 信號CA的收送信介面。當CAD匯流排上輸出指令及位 址時,CAD轉送電路203設定狀態信號CA爲"H"位準。 指令解釋電路204爲解釋從I/O裝置發出的指令,因 應其結果對DMA電路154設定執行DMA轉送的參數。 藉此使來自I/O裝置的記憶體讀取/記憶體寫入的需求形 成由DMA轉送來處理。 錯誤偵測器205由I/O裝置接收以固定的時序返回的 承認信號ACK,因應其結果判斷I/O裝置的接收信狀態 。當RETRY需求返回時,對CAD傳輸電路203指示RETRY。 -14- 502171 五、發明説明(i3 ) 各I/O裝置如圖所示,具備了匯流排需求(BREQ)電路 301、時鐘脈衝產生電路302、CAD傳輸電路303、指令 解釋電路304、錯誤偵測器305、FIFO緩衝器306、資 料讀取控制電路307、記錄器群308、及DMA控制器 (DMAC)309。 匯流排需求(BREQ)電路301,爲使用匯流排使用權需 求信號BREQ而對匯流排仲裁器電路1 5的匯流排接地 (BGNT)電路201要求匯流排使用權者,當接收到匯流排 使用權許可信號BGNT時,允許以CAD轉送電路303將 資料輸出。 時鐘脈衝產生電路302、CAD傳輸電路303、指令解 釋電路304、及錯誤偵測器305等各自與匯流排仲裁器 電路15的時鐘脈衝產生電路202、CAD傳輸電路203、 指令解釋電路204及錯誤偵測器205具有相同的功能。 但是,對於時鐘脈衝產生電路302,並無特剀需要,如 匯流排仲裁器電路1 5的時鐘脈衝頻率可變功能。 資料讀取寫入控制電路307爲因應來自匯流排仲裁器 電路15的存取要求,而執行記錄器群308的讀/寫。讀/ 寫資料爲藉由FIFO緩衝器306在與CAD轉送電路303 之間傳送接收。FIFO緩衝器306爲顯示出記錄器中空閒 狀況,如是滿的時候,錯誤偵測器205則返回RETRY 的信號。如無問題時則返回肯定的回應。如無接收資料 時則無回應。 DMA控制器(DMAO309以I/O裝置作爲匯流排主控端 *-15- 502171 五、發明説明(14 ) 來動作,當設定起始位址及轉送位元數時,因應的指令 及位址則由C A D轉送電路3 0 3輸出。 其次’對於本匯流排系統的具體動作予以說明。 < CPU將資料寫入I/O裝置的場合時> 首先,參考第7及8圖’從CPU將32位元的資料寫 入I/O裝置21的場合時予以說明。 在此1昜合時,CPU 1 1將「4位元寫入信號」與「位址 (25位元)」及「資料(32位元)」輸出至記憶體匯流排1〇 上。這些資料則被輸入且記憶在匯流·排仲裁器電路1 5 的FIFO緩衝器151。 在指令/位址/資料合成電路(CAD合成電路)1 52中 ,從FIFO緩衝器1 5 1中取出資料以搭載本實施型態的 匯流排系統般地變換爲指令/位址(CA)/及資料D0-D3。 在此,CA成爲指令+位址, 指令爲,4位元轉送+寫入+4時鐘脈衝命令+A[24] 位址爲,A[23:16]、A[15:18]、A[7:0] 資料爲,D0 + D1+D2.D3 ( 8位元x4 = 32位元)。 這些資料通過CAD匯流排而寫入I/O裝置21。此時的 資料轉送控制動作如下。 (1) 匯流排仲裁器15確保匯流排權成爲主控端(Bus Owner)。 (2) 匯流排仲裁器15對存取對象的I/O裝置21,將 SCLK 1以配合該I/O裝置21的轉送速度的頻率來輸 出。 -16- 502171 五、發明説明(i5 ) (3) 在SCLK 1的升起點上’ !/〇裝置21辨識自己所選擇 的內容,並準備資料的接收。 (4) 匯流排仲裁器15以同步於SCLK的升起點,而對 C A K [ 7 : 0 ]匯流排以指令、位址、資料的順序來輸出。 在輸出指令及位址的指令/位址相位中,匯流排仲裁 器15將前述的指令/位址狀態信號CA設爲"H"。 (5) I/O裏置2丨同步於SCLK 1的下降點,以讀取 CAD[7:0]。接著’解釋指令及起始位址以準備寫入動 作。 (6) I/O裝置21將4時鐘脈衝作爲1循環,以該第3時 鐘脈衝的T3時序對ACK[1:0]信號輸出狀態(3時鐘 脈衝命令的場合,以T2時序對ACK[1:0]輸出狀 態)。 (7) 在這之後所指定的轉送位元數的資料爲由匯流排仲 裁器15來輸出。 (8) I/O裝置21接收傳送資料,將其寫入所指定的記憶 區域。 (9) 在資料相位中,I/O裝置21以其起始的T1時序對 ACK[1 :0]端子輸出狀態。當有Retry要求的場合時, 轉送端亦即匯流排仲裁器1 5判斷該循環(4或3時 鐘脈衝)的轉送失敗,就再度的傳送(第8圖)。 (10) 資料傳輸與時鐘之衝停止的同時一起結束。 而在資料相位中A C K的時序爲D 0 (T 1)的情形是,是 否可接收資料爲由FIFO的空閒狀況察看便可馬上了解 502171 五、發明説明(16 ) ’所以可以由D0的時序來送出。可是指令的場合時, 須解釋爲何種指令的緣故’所以無法在最初的時鐘脈衝 時送出。 < CPU從I/O裝置讀取資料的場合時> 其次’參考第9圖,CPU 1 1從I/O裝置21執行8位 元組資料的讀取場合時予以說明** 在此場合’ C P U 1 1在記憶體匯流排1 〇上產生續取信 號及位址。匯流排仲裁器電路1 5的FIFO緩衝器1 5 1接 收讀取信號時,在I/O完成準備爲止會送出等待信號 WaU。接著,從匯流排仲裁器電路15對I/O裝置21轉 送讀取指令、位址,而由I/O裝置開始資料的讀取。此 時的資料轉送控制動作如下。 (1) 匯流排仲裁器1 5對存取對象的I/O裝置2 1,將 SCLK 1以配合該I/O裝置21的轉送速度頻率來輸 出。 (2) 在SCLK 1的升起點上,I/O裝置21辨識自己所選擇 的內容,並選擇資料的接收。 (3) 匯流排仲裁器15同步於SCLK 1的升起點,對 CAD[7:0]匯流排以指令、位址、資料的順序來輸出。 在輸出指令及位址的指令/位址相位中,匯流排仲裁 器15將前述的指令/位址狀態信號CA設爲。502171 V. Description of the invention (1) [Technical category to which the invention belongs] The present invention relates to a data transmission system using an improved bus structure. [Known Technology] There are generally known two types of bus structures for I / O devices used in electronic devices such as portable information terminals, such as (υ bus type and (2i channel type). The bus type is a plural I / O device, and the structure is connected by a common bus. In addition to the address bus and the data bus, the clock signal line is also shared among the plural I / O devices. Therefore, the number of signal wiring can be reduced, but because the signals are shared between the devices, in order to improve the instability of the generated signal waveform, wiring countermeasures must be taken, which is disadvantageous to the increase of frequency and power consumption. In terms of the channel type, each I / O device is independently configured in the address bus, data bus, clock signal line, etc. Each I / O device can perform flexible access control and is compared with the bus type Below, the noise countermeasures are relatively simple, and it is easy to improve the performance. However, on the contrary, there is a problem of increasing the number of signal wirings. For portable information terminals, etc. In small electronic devices, because the actual assembly area of the components is limited, the number of signal wirings must be minimized. However, when using the above-mentioned bus-type wiring structure, all the equipment must be frequently used. The reason for supplying the clock pulse is that the power consumption may increase. Therefore, even when a channel-type wiring structure is used, 'in order to reduce the number of signal line wiring, 502171 V. Description of the invention (2) Therefore, a new bus is required. Structure. In order to reduce the number of signal wirings of the bus, it is a well-known fact that the address and data are on the same bus, and the time is allocated. However, it is known that the output order of the address and data and the output timing It is determined in advance, so it must always be accessed at this timing. Therefore, when a sequence deviation occurs due to noise, a problem that can cause erroneous actions can occur. In addition, the address phase and data phase requirements must be continuous, so Access separates the bit phase and data phase of an I / O device. It is difficult to control access to other I / O devices, or to accept access requests from other I / O devices. [Invention Summary] The present invention seeks to solve the problem in view of this situation. A data transfer system suitable for a small electronic device such as a portable data terminal that requires a small size and low power consumption is realized by realizing a new bus structure, and the data transfer system of the present invention is a data transfer system of the present invention. The data transmission system for transmitting data between I / O devices is characterized in that each of the aforementioned I / O devices is provided with a demand device that requires ownership of the bus; the aforementioned bus arbiter is provided to allow the aforementioned bus ownership Grounding device; wherein the aforementioned bus arbiter and the aforementioned I / O device have: a device in which instructions, addresses and data are output in the same bus; when using the aforementioned bus to transmit data, a clock pulse is transmitted to the other party Device; a trusted device after receiving a clock pulse from the other party and then receiving data from the other party; the aforementioned instructions, addresses, and data are in » 4- 502171 V. Description of the invention (3) Same—When the bus is output, it will be different from the output of the bus that is outputting instructions or signal signals. For this data transmission device, output instructions and bits on the same bus are used. The structure of the address and the data is transmitted by the data transfer between the bus arbiter and the plural I / O devices. The instructions on the bus used by pzf and the address and data are transmitted by time allocation. In this case, in the case of the bus, The output is a command or 'material is specified by a special signal output by the transmission place. Therefore, even if each command and data can not be transmitted at the timing determined by the conventional, data transmission can be performed normally, Therefore, the sequence deviation of noise can improve the tolerance. Moreover, the address phase and data phase do not need to be performed continuously, so the division is used to access a certain I / O. Data phase between the set address phase and data phase and other I / O devices can also be performed. Also, the signal to be distinguished as a command or data is output by the I / O device. 5 The I / O The content output from the device to the bus can be clearly marked as a notice to the bus arbiter as a command or a data I / O device can issue instructions on its own to transmit data. In particular, in this data transmission system, the receiving end and the slave transmitting end The transmitted clock pulses are synchronized to receive the channel-type structure of the data. Because it is used in a combination of the bus-type architecture in which commands, addresses, and data are output on the same bus, data is transmitted from the I / O device It is because of the mechanism that needs to be judged because 5 is formed by reading data from the I / O device, or it is formed by the instruction of the I / O device itself. 71 dedicated letter V. invention is described in No. (4), the I / O end notification address arbitration device architecture, the access control can be achieved more elastic properties. [Brief description of the drawings] Fig. 1 is a block diagram showing a system architecture of an information processing device using a data transmission system according to an embodiment of the present invention. Fig. 2 shows the ~ terminal specifications of the same implementation type in the bus arbiter and 1/0 device. Fig. 3 illustrates a pattern of a 3-clock pulse command and a 4-clock pulse command used in the same embodiment. Section 4_ shows the structure of the instructions used in the same implementation type. Figure 5 shows the types of data transfer that can be specified because of the instructions used in the same implementation. Fig. 6 shows a specific circuit architecture diagram of a transmission circuit portion used in the same embodiment. Fig. 7 shows a timing flow chart when the CPU writes data to the I / O device for the same embodiment. Fig. 8 is a flowchart showing the timing of the retry sequence in the same embodiment. Fig. 9 shows the timing flow chart of the display operation when the CPU reads data from the I / O device in the same embodiment. Fig. 10 shows a timing flow chart during the data transmission operation according to the read request from the I / O device in the same implementation mode. Fig. 11 shows a timing flow chart of the data transmission operation according to the write request from the I / O device in the same implementation mode. V. Description of the invention (5) Brother 1 2 时序 The sequence diagram showing the stopping of operations during data transmission in a consistent application mode. Fig. 13 shows a timing chart of the switching operation of the I / O device in the same embodiment. Fig. 14 is a timing chart showing other examples of the switching operation of the 1 / () device in the same embodiment. [Embodiment example] In Fig. 1, a system architecture of an information processing device using a data transmission system according to an embodiment of the present invention is shown. This information processing is used as a portable information terminal, etc. As shown in the figure, it includes a CPU 11, a RAM 12, a ROM 13, a gate array 14, and a plurality of I / O devices 21 to 23. The I / O devices 21 to 23 are various peripheral devices such as a magnetic disk drive (FDD), a hard disk drive (HDD), an audio device, a digital camera, and a communication controller. Gate Array 14 is a two-way connection between the memory bus 10 on the host side and the I / O bus 20 on the I / O devices 21 ~ 23. The I / O bus 20 is connected according to the access request from the CPU 11 for I / O. The access control of the O devices 21 to 23 or the memory access control is performed in response to the memory access requirements from the I / O devices 21 to 23. The gate array 14 is assembled as a peripheral device, such as a display driving circuit that drives a liquid crystal display device (not shown). In addition, the gate array 14 includes a bus arbitrator circuit 15 for controlling I / O devices 21 to 23 to control bus switching. -7-502171 V. Description of the invention (6) The host-side memory bus 1 0 'as shown in the figure' includes a 32-bit wide data bus, a 26-bit wide address bus, and control signals. The control signals include a chip selection signal CS, a read signal Read, and three write signals Write. The Write signal indicates that 1-byte write, 2-byte write, and 4-byte write. The CPU 11 accesses the RAM 12 and the ROM 13 by using the memory bus 10 to issue a memory address. For the I / O devices 2 1 to 23, the CPU 1 1 also uses the bus memory 10 to issue a memory address and access it. That is, from the perspective of the CPU 11, all the I / O devices 21 to 23 become a hardware architecture developed on a memory map, and each I / O device is used as a memory map I / O. The I / O bus 20 to which the I / O devices 21 to 23 are connected is the bus structure used for the aforementioned channel / bus type, which is a common I / O bus connected to the I / O devices 21 to 23 Rows, and each I / O device are made up of channel signal lines that are set independently. The data transfer width of the memory bus 丨 〇 is 32 bits, and the data transfer width of the corresponding I / O common bus is 8 bits. That is, the common I / O bus includes an 8-bit-wide CAD bus that transmits instructions and addresses and data by time allocation, and each; [/ 0 data transmission between devices is via CAD to implement. Therefore, when data is transferred between the CPU 11 and each I / O device, a 26-bit wide address and 32-bit wide parallel data are each divided into 8-bit units, and then serialized. (Send to i / C) Device side for sequence: "J / parallel conversion processing" or 8-bit data sheet 502171 from I / O device side 5. Description of invention (7) bit 'Address or data for sequence transfer The aggregate is 32 bits wide and forwarded to the memory bus 10 to perform the sequence / parallel transform process. In the bus arbitrator circuit 15, three channels of channels 1 to 3 are prepared. Each channel is connected to the corresponding I / O device through the channel interface signal line. The bus arbiter circuit 15 and the terminal specifications of each I / O device are as shown in Figure 2. That is, the I / O common bus is 13 and 8 CAD signal lines CAD [7: 〇], 1 It consists of two instruction / address status signal lines CA, two acknowledgement signal lines ACK [1: 0], one OFF signal line OFF, and one reset signal line RESETB. The command / address status signal line CA serves as a signal to distinguish the command or data currently output on the CAD. During the output period of the instruction or address, the status signal CA is set to the active status "H" by its output terminal (master terminal). The status signal CA is set to the inactive state during the data output period. Acknowledge signal line ACK [1: 0] is a response signal indicating the receiving status of the command, address, or data receiver, and ACK = ”01 " is a signal indicating that the command, address, or data cycle has ended normally ( Acknowledge), and ACK "10" is a cyclic retransmission request for instructions, addresses, or data. This acknowledges that the signal line ACK [l: 〇] is output in several cycles, and is based on each address. The phase and the data phase are fixed. The OFF signal line 0FFB and the reset signal line RESETB are signals output from each bus arbiter circuit 15 to the I / O device side, and indicate the disconnection and reset of the I / O device. 502171 Ί V. Explanation of the invention (δ) And for each channel interface signal line, a source synchronous clock pulse is set? Signal SCLK, insert signal INTB, bus usage right request signal BREQ, bus usage right permission signal BGNT, one each. Source The pole synchronous clock pulse signal SCLK is a synchronous clock pulse used for transferring instructions, addresses, or data. It is used for data (command, address, and data) samples on the CAD bus. Source synchronous clock pulse The signal StLK is output from the master control. The data reception of the slave device at the receiving end starts according to the source synchronous clock pulse signal SCLK from the master, and receives the instruction in synchronization with the source synchronous clock pulse signal SCLK. Address, or information. The bus use right request signal BREQ is a signal requesting 10 common bus use rights for the bus arbiter circuit 15, and is sent from the 1/0 device to the bus arbiter circuit 15. The bus Bus arbiter circuit! 5 In order to mediate the usage right request signal BREQ 'from each of the I / O devices 2 1 to 23 to correspond to the bus usage right; [/ 〇The bus sends the bus usage right permission signal BGNT. Bus The right-of-use demand signal BREQ and the bus-use-right permission signal BGNT are each a negative logic signal. Therefore, the 15 terminals of the bus arbiter circuit are 13 + (3x4) = 25 'and each I / O device terminal is 13 + 4 = 17 terminals. Between the bus arbiter circuit 15 and the I / O devices 21 ~ 23, the data exchange is often generated by the two relationships between the master control slave device and the output end of the transmission data. Indeed After the bus power is controlled, the clock pulse SCLK for transfer is controlled to perform data transfer. [/ 0 device has each bus-1 〇- 502171 V. Description of the invention (9) Demand (BREQ) circuit, only the bus arbiter circuit is obtained 15 permitted I / O devices can output data. ≪ Instructions and Addresses > Next, referring to Figures 3 or 5, the instructions and addresses transferred in 8-bit wide CAD [7: 0] will be explained. For the bus system of this embodiment type, a three-clock pulse command such as 3 (a) 圏 and a four-clock pulse command such as 3 (b) _ are prepared. 3 The clock pulse command is composed of an 8-bit instruction and its subsequent address (A) [15: 8], A [7: 0] divided into two transfers. These three clock pulse commands are usually used, and the memory address space represented by the three clock pulse commands becomes the access target. The 4 clock pulse command is a command used when the memory address space is expanded. It is an 8-bit command and its subsequent divided into 3 transfer addresses (A [23:16], Α [ 15: 8], Α [7: 0]). To use the 3-clock pulse command or to use the 4-clock pulse command, it is specified by the extension bit defined in the 8-bit instruction. Figure 4 shows the 8-bit instruction structure. The four digits of CAD [6] to CAD [3] are command sections (CM [3] ~ CM [0]) that indicate the type of data transfer. An example of the relationship between the combination of CM [3] to CM [0] and the instruction content is shown in Figure 5. When continuous data transfer of more than 2 bytes is performed, the read / write address of each byte of the instruction receiver is usually increased by +1 automatically. However, in this embodiment, it is also prepared. When the address is not increased, a continuous data transfer command of more than 2 bytes is executed. Once, CM [3] ~ CM [0] can also refer to »11- 502171 V. Description of the invention (10) instructs to suspend the transfer of data during execution. CAD [2] indicates the access type of read / write, caD [2] = " 0 "indicates read access, and CAD [2] = " 1 " indicates write access. CAD [1] alignment The presence or absence of address expansion, which indicates that the current instruction and address are 3-clock pulse commands, or the 4-bit pulse command indicates the address expansion bit. CAD [l] = " 〇 "means 3-clock pulse command, CAD [1] = ”r is f clock pulse command. CAD [0] is used to refer to the content of the first bit of the address that is not wrapped around the instruction. 'As A [1 6] when 3 clock pulse command, or The 4-clock pulse command is used as A [24]. Therefore, in fact, 8-bit instruction formation also includes address information. ≪ Circuit Architecture > Second, referring to Figure 6, for the bus system The specific circuit architecture of the used data transfer circuit section will be explained. Since the bus interface of the 1/0 device 21 to 23 is the same, the I / O device 21 is used as a representative in FIG. 6. In the interface signal line, the insertion signal is omitted. In addition to the bus mediation in the bus arbiter circuit 15 In addition, it includes the function of data transfer between control and I / O devices. That is, the bus arbiter circuit 15 is provided with a FIFO buffer 151, an instruction / address / data synthesis circuit (as shown in the figure) CAD synthesis circuit) 152, transfer control circuit 153, and DMA circuit 154. The FIFO buffer 1 51 is an interface buffer with the memory bus 1 0. In order to realize the aforementioned parallel / sequence conversion, serial / parallel conversion -12- 502171 V. Description of the invention (11) Change 'execute control signal, address, data receipt and accumulation from memory bus 10, and execute instruction, address, data receipt from CAD bus For example, when the CPU 11 performs write access to the I / O device, the address, data, and write signal are held in the FIFO buffer 1 5 1. The CAD synthesis circuit 152 is held in the FIFO buffer. The storage data stored in the device 151 is boarded on a CAD bus to be converted into the contents of 8-bit sequence data, and the CAD synthesis circuit 152 and the FIFO buffer 151 are used to form the aforementioned parallel / sequence conversion interface. CPU 1 1 32-bit implementation of I / O devices When writing and accessing data, the CAD synthesizing circuit 152 generates the aforementioned 3-clock pulse or 4-clock pulse command (CA) according to the write signal and address, and divides the 32-bit wide data into 8 bits each. Bits are converted into 3-byte sequence data from D0 to D3. Transmission control circuit 1 53 performs actual data transmission control with each I / O device. For this transmission control circuit 1 53, as shown in the figure A bus ground (BGNT) circuit 201, a clock pulse generating circuit 202, a CAD transfer circuit 203, an instruction interpretation circuit 204, and an error detector 205 are provided. The bus ground (BGNT) circuit 201 has a busbar use right demand signal BREQ1 to 3 and a bus connection between each I / O device 2 1 to 23 and the bus ground (BGNT) circuit 201 in a one-to-one connection. The usage rights are used as the permission signals BGNT 1 ~ 3, and these usage right demand signals BREQ1 ~ 3 and the bus usage right permission signals BGNT 1 ~ 3 are used to perform the mediation of the usage rights of the I / O common bus. For the I / O common bus -13-502171 V. Description of the invention () The mediation of the right to use the common bus. For the bus liberation timing of the 1 / () common bus, 'the cycle in which the command / address phase and data phase are liberated has been fixed.' If there are other bus usage rights request signals BREQ, then the bus is executed at this timing Right switch. The clock pulse generating circuit 202 is used for generating the aforementioned source synchronous clock pulse signal SCLK during data output. This source synchronous clock pulse signal SCLK can also be used as the operation clock pulse of the CAD transfer circuit 203. The frequency of the source synchronous clock pulse signal SCLK can be set to be variable according to the transfer processing speed of the counterpart I / O device. In addition, the clock pulse generating circuit 202 has a switching circuit therein, and when receiving a signal, the source synchronization clock pulse signal SCLK received from the I / O device is given to the CAD transmission circuit 202 or the error detector 205 as data. Load clock pulse. The CAD transmission circuit 203 is used for data transfer between the CAD bus and the I / O device. It has the aforementioned sending / receiving interface of the command / address status signal CA. When a command and an address are output on the CAD bus, the CAD transfer circuit 203 sets the status signal CA to the " H " level. The instruction interpretation circuit 204 interprets the instruction issued from the I / O device, and sets a parameter for executing the DMA transfer to the DMA circuit 154 in accordance with the result. This allows the memory read / write request from the I / O device to be processed by DMA transfer. The error detector 205 receives the acknowledgement signal ACK returned by the I / O device at a fixed timing, and judges the reception status of the I / O device according to the result. When the RETRY demand returns, the CAD transmission circuit 203 instructs RETRY. -14-502171 V. Description of the Invention (i3) Each I / O device is provided with a bus request (BREQ) circuit 301, a clock pulse generation circuit 302, a CAD transmission circuit 303, an instruction interpretation circuit 304, and an error detection circuit, as shown in the figure. A tester 305, a FIFO buffer 306, a data read control circuit 307, a recorder group 308, and a DMA controller (DMAC) 309. The bus demand (BREQ) circuit 301 requests the bus arbiter circuit 15 for the bus grounding (BGNT) circuit 201 to use the bus right request signal BREQ. The bus grounder (BGNT) circuit 201 requires the bus right to use it. The permission signal BGNT allows data to be output by the CAD transfer circuit 303. The clock pulse generation circuit 302, the CAD transmission circuit 303, the instruction interpretation circuit 304, and the error detector 305 are each connected to the clock pulse generation circuit 202, the CAD transmission circuit 203, the instruction interpretation circuit 204, and the error detection circuit of the bus arbiter circuit 15. The tester 205 has the same function. However, there is no special need for the clock pulse generating circuit 302, such as the clock pulse frequency variable function of the bus arbiter circuit 15. The data read / write control circuit 307 executes read / write of the recorder group 308 in response to an access request from the bus arbiter circuit 15. The read / write data is transmitted to and received from the CAD transfer circuit 303 through the FIFO buffer 306. The FIFO buffer 306 is used to display the idle status in the recorder. If it is full, the error detector 205 returns a RETRY signal. If there is no problem, a positive response is returned. If no data is received, there is no response. DMA controller (DMAO309 uses I / O device as the bus master * -15- 502171 V. Description of invention (14) to act, when setting the start address and the number of transferred bits, the corresponding instruction and address Then, it is output by the CAD transfer circuit 303. Next, 'the specific operation of the bus system will be explained. ≪ When the CPU writes data to the I / O device > First, refer to Figures 7 and 8' from the CPU A description will be given of the case where 32-bit data is written into the I / O device 21. When this 1 is combined, the CPU 11 writes the "4-bit write signal" with the "address (25-bit)" and " Data (32-bit) ”is output to the memory bus 10. The data is input and stored in the FIFO buffer 151 of the bus · arbiter circuit 15. In the instruction / address / data synthesis circuit (CAD In the synthesis circuit) 1 52, the data is taken out from the FIFO buffer 1 51 and converted into instructions / addresses (CA) / and data D0-D3 like the bus system equipped with this embodiment. Here, CA becomes Instruction + address, instruction is, 4-bit transfer + write + 4 clock pulse command + A [24] address is, A [23:16], A [15:18], A [7: 0] The data is D0 + D1 + D2.D3 (8 bits x 4 = 32 bits). These data are written into the I / O device 21 through the CAD bus. The data transfer control operation at this time is as follows. (1) The bus arbiter 15 ensures that the bus right becomes the master (Bus Owner). (2) The bus arbiter 15 transfers SCLK 1 to the I / O device 21 of the access object to match the transfer of the I / O device 21 The frequency of the speed is output. -16- 502171 V. Description of the invention (i5) (3) At the starting point of SCLK 1 '! / 〇 The device 21 recognizes the content selected by itself and prepares for the reception of the data. (4) Confluence The bank arbiter 15 synchronizes with the rising point of SCLK, and outputs the CAK [7: 0] bus in the order of instruction, address, and data. In the command / address phase of the output instruction and address, the bus The arbiter 15 sets the aforementioned instruction / address status signal CA to " H ". (5) I / O is set 2 and synchronized with the falling point of SCLK 1, to read CAD [7: 0]. Then ' Interpret the instruction and the start address to prepare for the write operation. (6) The I / O device 21 uses 4 clock pulses as one cycle, and signals ACK [1: 0] at the T3 timing of the third clock pulse. Output status (in the case of a 3-clock pulse command, the ACK [1: 0] output status is output at T2 timing.) (7) The data of the number of transfer bits specified after that is output by the bus arbiter 15. 8) The I / O device 21 receives the transmitted data and writes it into the designated memory area. (9) In the data phase, the I / O device 21 outputs a state to the ACK [1: 0] terminal at its initial T1 timing. When there is a request for Retry, the transfer end, that is, the bus arbiter 15 determines that the transfer of the cycle (4 or 3 clock pulses) has failed, and retransmits it (Figure 8). (10) The data transfer ends at the same time as the clock pulse stops. In the data phase, the timing of the ACK is D 0 (T 1). Whether the data can be received is based on the idle state of the FIFO. You can immediately understand 502171. 5. Description of the invention (16) 'So it can be determined by the timing of D0. Submit. However, in the case of a command, it is necessary to explain what kind of command it is, so it cannot be sent out at the first clock pulse. < When the CPU reads data from the I / O device > Secondly, referring to FIG. 9, when the CPU 11 reads 8-byte data from the I / O device 21, it will be described. ** In this case 'CPU 1 1 generates the fetch signal and address on the memory bus 10. When the FIFO buffer 15 of the bus arbiter circuit 15 receives a read signal, it waits for a wait signal WaU until the I / O is ready. Next, the bus arbiter circuit 15 transfers a read command and an address to the I / O device 21, and the I / O device starts reading the data. The data transfer control operation at this time is as follows. (1) The bus arbiter 15 outputs SCLK 1 to the I / O device 21 to be accessed in accordance with the transfer speed frequency of the I / O device 21. (2) At the rising point of SCLK 1, the I / O device 21 recognizes the content selected by itself and selects the reception of the data. (3) The bus arbiter 15 is synchronized with the rising point of SCLK 1, and outputs the order of command, address, and data to the CAD [7: 0] bus. In the command / address phase of the output command and address, the bus arbiter 15 sets the aforementioned command / address status signal CA.

(4) I/O裝置21同步於SCLK 1的下降點,讀取CAD(4) I / O device 21 is synchronized with the falling point of SCLK 1 and reads CAD

[7 : 〇 ]。接著,解釋指令及起始位址以準備寫入動作。 (5) 匯流排仲裁器1 5因到目前爲止的資料轉送,來成爲 -18- 502171 五、發明説明(17 ) 來自1/◦裝置21之待命輸出。 (6) 在這之後I / 〇裝置2 1係對匯流排仲裁器1 5輸出 BREQ 1並要求匯流排權。 (7) 匯流排仲裁器1 5判斷匯流排要求的優先順序,並對 於I/O裝置21發出BGNT 1給予匯流排權。 (8) I/O裝置21確保匯流排權後,將SCLK 1以配合該 I/O裝置21的轉送速度之頻率來輸出。 (9) I/O裝置21同步於SCLK丨的升起點,對CAD匯流 排所指定的轉送位元組數目的資料以位元組單位順序 輸出。 (10) 指定次數的轉送結束時,I/O裝置21使BREQ 1成 爲非作用狀態,而匯流排仲裁器1 5接收該內容使 B G N T 1成爲非作用狀態。 <依來自I/O的讀取要求之DMA轉送> 其次參考第10圖,依來自I/O裝置21的讀取要求而 執行DMA傳輸時的場合予以說明其動作。 (1) I/O裝置21發出BREQ 1以要求匯流排權。 (2) 匯流排仲裁器15發出BGNT 1以給予uo裝置21匯 流排權。 (3) I/O裝置21將SCLK 1以配合該I/O裝置的轉送速度 頻率予以輸出,接著與SCLK 1同步在CAD匯流排上 將讀取指令與起始位址順序輸出。在輸出指令及位址 的指令/位址相位中,I/O裝置2 1將前述的指令/位址 狀態信號CA設爲"H”。 -19- 502171 五、發明説明(18 ) (4) 因信號CA爲"H"的緣故,匯流排仲裁器15與SCLK 1的下降點同步,將CAD [7:0]作爲來自I/O裝置21 的指令/位址來讀取。接著’解釋指令及起始位址以 準備DMA傳輸。 (5) 1/0裝置21因到目前爲止的資料轉送,而成爲來自 匯流排仲裁器15的待命輸出,使BREQ 1成爲非作 用狀~態以解放匯流排。 (6) 以後,匯流排仲裁器15獲得匯流排權而對I/O裝置 21輸出SCLK 1。接著,被CAD匯流排指定的傳輸位 元組的資料以位元組單位順序輸出。 <依來自I/O的寫入要求之DMA轉送> 其次參考第11圖,依來自I/O裝置21的寫入要求, 以執行DMA傳輸的場合予以說明其動作。 (1) 1/0裝置21發出BREQ 1以要求匯流排權。 (2) 匯流排仲裁器15發出BGNT 1以給予I/O裝置21匯 流排權。 (3) 1/0裝置21將SCLK 1以配合該I/O裝置的轉送速度 頻率予以輸出,接著與SCLK 1同步在CAD匯流排上 將寫入指令與起始位址順序輸出。在輸出指令及位址 的指令/位址相位中,I/O裝置2 1將前述的指令/位址 狀態信號CA設爲ΠΗ"。 (4) 因信號CA爲"Η"的緣故,匯流排仲裁器15與SCLK 1的下降點同步,將CAD [7:0]作爲指令/位址來讀 取。接著,解釋指令及起始位址以準備DMA轉送。 -20- 502171 五、發明説明(19 ) 在此準備期間,匯流排使用權係由I/O裝置2 1暫時 解放。 (5) I/O裝置21發出BREQ 1以要求匯流排權。 (6) 匯流排仲裁器15當DMA轉送準備完成時,發送 B G N T 1以給予I / 〇裝置2 1匯流排權。 (7) I/O裝置21將SCLK 1以配合該I/O裝置的轉送速度 頻率_予以輸出。 (8) I/O裝置21與SCLK 1的下降點同步,對CAD匯流 排所指定的傳輸位元組的資料以位元組單位順序輸 出。 <資料轉送的中途停止> 其次參考第12圖,其係例示CPU 1 1在對I/O裝置21 寫入資料時,對於資料轉送途中之停止動作予以說明。 (1) 匯流排仲裁器15對I/O裝置21將SCLK 1以配合該 I/O裝置的轉送速度頻率予以輸出。 (2) 匯流排仲裁器15同步於SCLK 1的升起點,對CAD [7:0]匯流排以8位元組寫入的指令、位址、資料的順 序來輸出。在輸出指令及位址的指令/位址相位上, 匯流排仲裁器1 5將前述的指令/位址狀態信號CA設 爲”H" 〇 (3) 匯流排仲裁器15,在某種原因而中途中斷轉送的場 合時,將指令/位址狀態信號CA重新設爲"H",發行 新指令或對I/O裝置發出停止指令。由此方式,可以 將執行中的資料轉送中途停止。 -21- 502171 五、發明説明(2Q ) <資料轉送的切換i > 其次參考第13圖,在對I/O裝置2 1的資料寫入的轉 送途中’將匯流排周期切換成來自I/O裝匱22的資料讀 取動作時予以說明。在此處,假設爲由1/0裝置22預先 接受讀取指令,在相對於此之來自〗/〇裝置22的讀取資 料的轉送前,先行對I/O裝置2 1開始寫入存取的場合。 (1) 匯流排仲裁器15對於I/O裝置21,將SCLK 1以配 合該I/O裝置21的轉送速度頻率予以輸出。 (2) 匯流排仲裁器15與SCLK 1的上升點同步,對CAD [7:0]匯流排將指令、位址、資料予以順序輸出。[7: 〇]. Next, the instruction and the start address are interpreted in preparation for a write operation. (5) The bus arbiter 15 becomes -18- 502171 due to the data transfer so far. 5. Description of the invention (17) Standby output from 1 / ◦ device 21. (6) After this, the I / 〇 device 2 1 outputs BREQ 1 to the bus arbiter 15 and requests the bus right. (7) The bus arbiter 15 judges the priority order of the bus request, and gives the bus right to the BGNT 1 issued by the I / O device 21. (8) After the I / O device 21 ensures the bus power, SCLK 1 is output at a frequency that matches the transfer speed of the I / O device 21. (9) The I / O device 21 is synchronized with the rising point of SCLK, and outputs the data of the number of bytes specified by the CAD bus in byte units. (10) At the end of the specified number of transfers, the I / O device 21 makes BREQ 1 inactive, and the bus arbiter 15 receives this content to make B G N T 1 inactive. < DMA transfer according to read request from I / O > Next, referring to Fig. 10, the operation will be described in the case where DMA transfer is performed according to the read request from I / O device 21. (1) The I / O device 21 issues a BREQ 1 to request the bus flow right. (2) The bus arbiter 15 issues BGNT 1 to give the uo device 21 the bus right. (3) I / O device 21 outputs SCLK 1 in accordance with the transfer speed and frequency of the I / O device, and then synchronizes with SCLK 1 and outputs the read command and the start address on the CAD bus in sequence. In the instruction / address phase of the output instruction and address, the I / O device 21 sets the aforementioned instruction / address status signal CA to " H ". -19- 502171 V. Description of the invention (18) (4 ) Because the signal CA is " H ", the bus arbiter 15 is synchronized with the falling point of SCLK 1, and reads CAD [7: 0] as the instruction / address from the I / O device 21. Then, ' Interpret the instruction and the start address to prepare for DMA transfer. (5) 1/0 device 21 becomes the standby output from the bus arbiter 15 due to the data transfer so far, making BREQ 1 inactive ~ (6) Later, the bus arbiter 15 obtains the bus right and outputs SCLK 1 to the I / O device 21. Then, the data of the transmission byte designated by the CAD bus is sequentially output in byte units. ≪ DMA transfer according to write request from I / O > Next, referring to FIG. 11, the operation will be described in the case of performing DMA transfer based on the write request from I / O device 21. (1) 1 The / 0 device 21 issues a BREQ 1 to request the bus right. (2) The bus arbiter 15 issues a BGNT 1 to give the I / O device 21 a bus right. (3) The 1/0 device 21 outputs SCLK 1 in accordance with the transfer speed frequency of the I / O device, and then synchronizes with SCLK 1 to sequentially output the write instruction and the start address on the CAD bus. In the output instruction In the command / address phase of the address, the I / O device 21 sets the aforementioned command / address status signal CA to ΠΗ ". (4) Because the signal CA is " Η ", the bus arbiter 15 is synchronized with the falling point of SCLK 1, and CAD [7: 0] is read as the instruction / address. Then, the instruction and the start address are explained to prepare for DMA transfer. -20- 502171 V. Description of the invention (19) During this preparation period, the bus use right is temporarily liberated by I / O device 21 1. (5) I / O device 21 issues BREQ 1 to request the bus right. (6) The bus arbiter 15 is ready when the DMA transfer is ready At this time, BGNT 1 is sent to give I / 〇 device 21 a bus right. (7) I / O device 21 outputs SCLK 1 to match the transfer speed frequency of the I / O device. (8) I / O device 21 is synchronized with the falling point of SCLK 1, and the data of the transmission byte specified by the CAD bus is output in the order of byte unit. ≪ Stopping on the way> Next, referring to FIG. 12, it illustrates that when the CPU 11 writes data to the I / O device 21, it will explain the stopping operation during data transfer. (1) The bus arbiter 15 performs I / O The device 21 outputs SCLK 1 in accordance with the transfer speed frequency of the I / O device. (2) The bus arbiter 15 is synchronized with the rising point of SCLK 1, and outputs the order of the instructions, addresses, and data written in 8 bytes to the CAD [7: 0] bus. On the command / address phase of the output instruction and address, the bus arbiter 15 sets the aforementioned command / address status signal CA to "H " 〇 (3) The bus arbiter 15 for some reason When the transfer is interrupted halfway, reset the command / address status signal CA to "H", issue a new command, or issue a stop command to the I / O device. In this way, the data transfer in progress can be stopped halfway. -21- 502171 V. Description of the invention (2Q) < Switching of data transfer i > Next, referring to FIG. 13, in the process of transferring data to the I / O device 21, 'switch the bus cycle to I The / O device 22 will explain the data reading operation. Here, it is assumed that the 1/0 device 22 accepts the read command in advance, and before the read data from the 〖/ 〇 device 22 is transferred, When writing access to I / O device 21 is started first. (1) Bus arbiter 15 outputs SCLK 1 to I / O device 21 in accordance with the transfer speed and frequency of the I / O device 21. (2) The bus arbiter 15 is synchronized with the rising point of SCLK 1. For CAD [7: 0] the bus will refer to , Address, data to be sequentially output.

(3) 在對I/O裝置21的資料轉送中,來自I/O裝置22的 BREQ 2成爲主動時,匯流排仲裁器15配合預先決定 的匯流排開放時序(此時爲D3結束點)輸出BGNT 2 ’對I/O裝置22賦予匯流排使用權。同時,SCLK 1 的輸出停止,對I/O裝置21的寫入資料轉送停止。 (4) I/O裝置22將SCLK 2以配合該I/O裝置22的轉送速 度頻率予以輸出。 (5) I/O裝置22與SCLK 2的上升點同步,對CAD將指 定的傳輸位元組資料以位元組單元開始順序輸出。 (6) 匯流排仲裁器電路1 5係以預先決定的匯流排開放時 序,將BGNT 2設爲非作用而解放匯流排,接著再次 獲得匯流排使用權。然後,輸出SCLK 1,再開始對 I/O裝置2 1之剩餘的資料。藉以上反覆實施以結束轉 送。 -22- 五、發明説明(21 ) <資料轉送的切換2 > 其次參考第14圖,在來自I/O裝置21的資料讀取的 轉送途中,將匯流排周期切換成來自I/O裝置22的資料 讀取動作時序以說明。在此處,假設爲由I/O裝置22預 先接受讀取指令,在相對於此之來自I/O裝置22的讀取 資料的轉送前,先行對I/O裝置2 1開始讀取存取的場 合。 ‘ (1) I/O裝置21獲得匯流排權之後,將SCLK 1以配合該 I/O裝置21的轉送速度頻率予以輸出。 (2) I/O裝置21與SCLK 1的上升點同步,對CAD所指 定的傳輸位元組資料以位元組單位開始順序輸出。 (3) 來自I/O裝置21的資料轉送中,在來自I/O裝置22 的BREQ 2成爲主動時,匯流排仲裁器15配合預先 決定的匯流排開放時序(此時爲D7結束點)使 BGNT 1爲不作用,使BGNT 2爲主動。 (4) I/O裝置22獲得匯流排權之後,將SCLK 2以配合該 I/O裝置22的轉送速度頻率予以輸出。 (5) I/O裝置22與SCLK 2的上升點同步,對CAD所指 定的轉送位元組資料以位元組單位開始順序輸出。 (6) 來自I/O裝置22的資料轉送中,在來自I/O裝置21 的BREQ 1成爲不作用時,匯流排仲裁器15配合預 先決定的匯流排開放時序(此時爲D3結束點)使 BGNT 2爲不作用,使BGNT 1爲主動。 (7) I/O裝置21輸出3(:]11^1以開始殘餘的資料轉送。藉 »23- 502171 五、發明説明(22 ) 以上反覆的動作來結束轉送。 如以上所述’對於本發明相關的匯流排系統, A·採用通道/匯流排倂用型的匯流排構造,而且附加 了區別指令/位址與資料的信號CA,因此, A-1)例如依第1〇圖的1/〇之read要求之DMA轉送所 說明的’不僅是來自匯流排仲裁器丨5的要求(命令) ’對應於來自I/O的獨立要求(DMA要求)爲可行。 A-2)而且在第12圖的寫入轉送途中停止的說明中, 即使在資料轉送途中,由於匯流排仲裁器.丨5端某種的 理由想要終止轉送時,將CA信號設定爲主動,可以使 資料沒有轉送的情形令I/O端辨識,以新訂定的指令停 止最近的順序,或是由停止指令予以中途停止。 A-3)又例如在第7圖的I/O寫入動作的說明中,由於 可以辨識指令/位址的相位,所以在序列途中因爲某種 原因發生時,也可以重新建立序列。 B.而且,例如,在第8圖的RETRY動作說明中,由 於固定了返回ACK時序, B-1)不需要發送用於鎖住ACK的時鐘脈衝,而可以削 減 p i η 〇 Β-2)ACK產生的時序因爲使用送信端的時鐘脈衝,所 以ACK的回應內容(承認、重試、錯誤)的判斷可以 同步設計,以得到電路的簡單化° B-3)特別是由於資料相位的ACK回應時序設在前端, ACK的接收端在很早時間可以得到資料轉送的回答,所 -24- 502171 五 '發明説明(23 ) 以對應容易可以使電路簡易化。 C. 例如在第14圖的I/O讀取切換的說明中’指令/位 址的相位將匯流排開放時序固定在第3周期或第4周 期,而資料相位將匯流排開時序固定在第2周期或第4 周期, C-1)不需要通知可以開放的匯流排時序就可以刪除該 通信信ϋ。 C-2)由於固定化所以從I/O 21到I/O 22的切換時機, 在實際切換1時鐘脈衝以上之前可以判斷,所以切換的 損失可以在最小。 D. 而且對於指令可以具有位址擴充位元,所以通常在 短周期來存取以實現高速動作,當要使用大記憶體容量 時,使用擴充位元可以擴充記憶體區域。 Ε.而且通常規定轉送資料數目時,此時在I/O端的位 址+1,因位址相同由於附加資料傳送數可以變更之功 能,因此對FIFO等的同一位址,也可以連續的讀寫資 料。 因此,依據本發明之匯流排系統,可以減少信號線配 線數,而且可以實現與通道型相同的彈性地存取控制。 符號說明(3) In the data transfer to the I / O device 21, when the BREQ 2 from the I / O device 22 becomes active, the bus arbiter 15 cooperates with a predetermined bus open timing (in this case, the D3 end point) to output BGNT 2 ′ grants the right to use the bus to the I / O device 22. At the same time, the output of SCLK 1 is stopped, and the transfer of write data to the I / O device 21 is stopped. (4) The I / O device 22 outputs SCLK 2 in accordance with the transfer speed frequency of the I / O device 22. (5) The I / O device 22 is synchronized with the rising point of SCLK 2 and outputs the specified transmission byte data to the CAD in sequential order starting from the byte unit. (6) The bus arbiter circuit 15 uses the predetermined bus open timing to set the BGNT 2 to non-active to liberate the bus, and then obtains the bus use right again. Then, SCLK 1 is output, and the remaining data for the I / O device 21 is started again. Repeat the above to end the transfer. -22- V. Description of the invention (21) < Switching of data transfer 2 > Next, referring to FIG. 14, in the process of transferring data from the I / O device 21, the bus cycle is switched to I / O The data reading operation timing of the device 22 will be described. Here, it is assumed that the I / O device 22 accepts the read command in advance, and before the read data from the I / O device 22 is transferred, read access to the I / O device 21 is started. The occasion. ‘(1) After the I / O device 21 obtains the bus power, it outputs SCLK 1 in accordance with the transfer speed frequency of the I / O device 21. (2) The I / O device 21 synchronizes with the rising point of SCLK 1, and outputs the byte data of the transmission designated by CAD in the order of byte units. (3) In the data transfer from the I / O device 21, when the BREQ 2 from the I / O device 22 becomes active, the bus arbiter 15 cooperates with a predetermined bus opening sequence (the end point of D7 at this time) to enable BGNT 1 is inactive, making BGNT 2 active. (4) After the I / O device 22 obtains the bus power, it outputs SCLK 2 in accordance with the transfer speed frequency of the I / O device 22. (5) The I / O device 22 synchronizes with the rising point of SCLK 2 and sequentially outputs the byte data for the transfer specified by CAD in byte units. (6) In the data transfer from the I / O device 22, when the BREQ 1 from the I / O device 21 becomes inactive, the bus arbiter 15 cooperates with a predetermined bus opening sequence (in this case, the D3 end point) Make BGNT 2 inactive and BGNT 1 active. (7) I / O device 21 outputs 3 (:) 11 ^ 1 to start the transfer of the remaining data. Borrow »23- 502171 V. Description of the invention (22) The above actions are repeated to end the transfer. As described above 'for this The invention relates to a bus system. A. A channel / bus general bus structure is used, and a signal CA that distinguishes instructions / addresses and data is added. Therefore, A-1) For example, according to 1 in FIG. 10 The DMA transfer of the read request of / 〇 stated that 'not only the request (command) from the bus arbiter 5' corresponds to the independent request (DMA request) from the I / O. A-2) Furthermore, in the description of the stop of the write transfer in Fig. 12, even during the data transfer, the bus arbiter is used due to some reason on the 5th end, and the CA signal is set to active. It can make the I / O terminal recognize the situation where the data is not transferred, stop the latest order with the newly set command, or stop it halfway by the stop command. A-3) For example, in the description of the I / O write operation in FIG. 7, since the phase of the instruction / address can be identified, the sequence can be re-established when it occurs for some reason during the sequence. B. For example, in the RETRY operation description in FIG. 8, since the return ACK timing is fixed, B-1) does not need to send a clock pulse for locking the ACK, and can reduce pi η 〇 2-2) ACK The generated timing uses the clock pulse of the transmitting end, so the judgment of the response content (acknowledge, retry, and error) of the ACK can be designed synchronously to simplify the circuit ° B-3) Especially due to the ACK response timing setting of the data phase At the front end, the receiving end of the ACK can get the answer of the data transfer at an early time, so -24-502171 5 'invention description (23) can simplify the circuit to make it easy to respond. C. For example, in the description of the I / O read switching in FIG. 14 'the phase of the instruction / address fixes the bus open timing at the 3rd or 4th cycle, and the data phase fixes the bus open timing at the 3rd or 4th cycle. 2 cycle or 4 cycle, C-1) The communication signal can be deleted without notifying the timing of the open bus. C-2) Due to the fixation, the switching timing from I / O 21 to I / O 22 can be judged before the actual switching is more than 1 clock pulse, so the switching loss can be minimized. D. In addition, the instruction can have an address expansion bit, so it is usually accessed in a short cycle to achieve high-speed operation. When a large memory capacity is to be used, the expansion bit can be used to expand the memory area. Ε. Usually, when the number of data to be transferred is specified, the address on the I / O side is +1. Because the address is the same and the number of additional data transfers can be changed, the same address of the FIFO can also be read continuously. Write information. Therefore, according to the bus system of the present invention, the number of signal lines can be reduced, and the same flexible access control as the channel type can be realized. Symbol Description

6…控制信號 10…記憶體匯流排 1 1 - CPU 12 …RAM -25- 502171 五、發明説明(24 )6… control signal 10… memory bus 1 1-CPU 12… RAM -25- 502171 V. Description of the invention (24)

13 …ROM 14…閘極陣列 15…Bus仲裁器電路 20…I/O匯流排 2L·" I/O 裝置 1 22…I/O裝置2 23:1/0 裝置 3 26…位址匯流排 32…資料匯流排13… ROM 14… Gate Array 15… Bus Arbiter Circuit 20… I / O Bus 2L · " I / O Device 1 22 ... I / O Device 2 23: 1/0 Device 3 26 ... Address Bus 32… data bus

151、306··· FIFO 152…CAD合成電路 153…轉送控制電路 154“· DMA 電路 201…BGNT電路 2 0 2…時鐘脈衝產生電路 203…CAD傳輸電路 204…指令解釋電路 205、305…錯誤偵測器 307…資料讀取寫入控制電路 308…記錄器群151, 306 ... FIFO 152 ... CAD synthesis circuit 153 ... Transfer control circuit 154 "· DMA circuit 201 ... BGNT circuit 2 0 2 ... Clock pulse generation circuit 203 ... CAD transmission circuit 204 ... Instruction interpretation circuit 205, 305 ... Error detection Tester 307 ... Data reading and writing control circuit 308 ... Recorder group

309 …DMAC -26-309… DMAC -26-

Claims (1)

502171 六、申請專利範圍 1 · 一種資料傳輸系統’係在匯流排仲裁器丨5與複數的 I/O裝置之間,傳輸資料的資料傳輸系統,其特徵爲在 前述各I/O裝置,設置了要求匯流排所有權的要求裝 置,對於前述匯流排仲裁器,設置了許可前述匯流排 的所有權之接地裝置,其中前述匯流排仲裁器與前述 各個I/O裝置具有:指令與位址與資料在同一匯流排 上輸甲的裝置;使用前述匯流排傳輸資料時,向對方 傳送時鐘脈衝的裝置;接收由對方來的時鐘脈衝再接 收由對方所傳輸的資料之後的收信裝置;前述指令與 位址與資料在同一匯流排上輸出時,將區別在此匯流 排上所輸出的是指令或資料的信號輸出的裝置。 2·如申請專利範圍第1項之資料傳輸系統,其中匯流排 仲裁器與前述各I/O裝置更具備藉由同一匯流排接收 由對方傳來的指令與位址及資料時,以固定化的時序 將接收端的收信狀態通知對方的通知裝置。 3·如申請專利範圍第1項之資料傳輸系統,其中在前述 指令與位址與資料的傳輸所使用的匯流排開放時序爲 固定,前述的匯流排仲裁器,在固定的匯流排開放時 序控制匯流排的所有權分配。 4 ·如申請專利範圍第1項之資料傳輸系統,其中對於前 述指令,包含了顯示接續在後面的位址位元長度有無 擴充的位址擴充位元,前述仲裁器與前述各I/O裝置 係依前述位址擴充位元來確定接續在前述指令的位址 位元長度。 -27- 502171 六、申請專利範圍 5·如申請專利範圍第丨項之資料傳輸系統,其中前述仲 裁器與前述各I/O裝置修具有將位址遞增同時執行連 續的資料傳輸的第1資料傳輸模式,與不遞增位址而 執行連續資料傳輸的第2資料傳輸模式,更具備有依 前述指令’使用前述第1及第2的任何一項資料傳輸 模式的指定裝置。 6.如申-請專利範圍第1至5項中任一項之資料傳輸系 統’其中前述仲裁器’從CPU端傳送的位址,資料予 以分割,變換爲搭載於匯流排序列資料的裝置。 7 · 一種資料傳輸系統’係具有:將指令、位址、及資料 以時間分配來傳送的匯流排,和連接於前述匯流排的 複數I/O裝置,及連接前述匯流排,將和前述各1/() 裝置間的資料傳輸藉由前述匯流排來執行的資料傳輸 控制裝置的資料傳輸系統,其特徵爲在前述匯流排定 義了相位區別信號線,用於表示在前述匯流排上需將 前述指令及前述資料的任何一個輸出,其中前述資料 傳輸控制裝置與前述各I/O裝置具有:藉由在其間配 置的時鐘脈衝信號線,接收由對方傳送的時鐘脈衝信 號,從該時鐘脈衝信號的傳送端是藉由前述匯流排傳 送的指令、位址、或是資料接收的開始裝置;與前述 時鐘脈衝信號同步在前述匯流排上輸出指令、位址、 或是資料之際,對前述匯流排輸出的爲指令或資料, 使用前述相位區別信號線向對方通知的裝置。 8.如申請專利範圍第7項之資料傳輸系統,其中對於前 502171 六、申請專利範圍 述匯流排,定義了在接收端將指令或資料的收信狀態 通知對方的回應信號線,前述回應信號的驅動時序’ 在前述匯流排上指令所輸出的指令相位,與前述匯流 排上資料所輸出的資料相位,規定每一個皆爲固定。 9 ·如申請專利範圍第7項之資料傳輸系統,其中前述資 料傳輸控制裝置’乃是因應於前述個I/O裝置而來的 匯流排權要求,具有調停匯流排所有權的匯流排調停 裝置,前述匯流排的解放時序,在前述匯流排上指令 所輸出的指令相位,與前述匯流排上資料所輸出的資 料相位,規定每一個皆爲固定,前述匯流排調停裝置, 因應新的匯流排權要求,以固定的時序來進行將實行 中的指令相位及資料相位中的各個解放。 -29-502171 6. Scope of patent application1. A data transmission system is a data transmission system for transmitting data between the bus arbiter 5 and a plurality of I / O devices, which is characterized in that the aforementioned I / O devices are provided with A device for requesting the ownership of the bus is provided. For the aforementioned bus arbiter, a grounding device is provided to allow the ownership of the aforementioned bus. The aforementioned bus arbiter and each of the I / O devices have: instructions, addresses and data in A device for transmitting A on the same bus; a device that transmits a clock pulse to the other party when using the aforementioned bus to transmit data; a receiving device that receives the clock pulse from the other party and then receives the data transmitted by the other party; the aforementioned instructions and bits When the address and data are output on the same bus, the device that outputs the command or data signal output on this bus will be distinguished. 2. If the data transmission system of item 1 of the patent application scope, the bus arbiter and each of the aforementioned I / O devices are further equipped with the same bus to receive instructions, addresses, and data from the other party. The timing of the notification of the receiving state of the receiving end to the notification device of the other party. 3. The data transmission system of item 1 in the scope of patent application, in which the bus opening sequence used in the aforementioned instruction, address and data transmission is fixed, and the aforementioned bus arbiter is used to open the timing control in a fixed bus. Allocation of bus ownership. 4. If the data transmission system of the first item of the patent application scope, wherein the foregoing instructions include the address extension bit indicating whether there is an extension of the subsequent address bit length, the aforementioned arbiter and the aforementioned I / O devices The length of the address bit following the instruction is determined according to the address extension bit. -27- 502171 6. Application scope 5. If the data transmission system of item 丨 of the scope of application for patent, the aforementioned arbiter and each of the aforementioned I / O devices have the first data to increase the address while performing continuous data transmission The transmission mode and the second data transmission mode in which continuous data transmission is performed without incrementing the address, are further provided with a designated device that uses any one of the aforementioned first and second data transmission modes in accordance with the aforementioned instruction. 6. As claimed-please apply for the data transmission system of any one of items 1 to 5 of the patent, where the aforementioned arbiter's address transmitted from the CPU, the data is divided, and it is converted into a device mounted on the bus sequence data. 7 · A data transmission system is provided with a bus that transmits instructions, addresses, and data in time distribution, a plurality of I / O devices connected to the bus, and a connection to the bus, and 1 / () Data transmission between devices The data transmission system of the data transmission control device performed by the aforementioned bus is characterized in that a phase difference signal line is defined on the aforementioned bus, which is used to indicate that Any one of the foregoing instruction and the foregoing data, wherein the foregoing data transmission control device and each of the I / O devices have: receiving a clock pulse signal transmitted by the other party through a clock pulse signal line disposed therebetween, and receiving the clock pulse signal from the clock pulse signal The transmitting end is a command, address, or data receiving start device transmitted through the bus; when the command, address, or data is output on the bus in synchronization with the clock pulse signal, the bus is The output device is a command or data, and the device notifies the other party by using the phase difference signal line. 8. If the data transmission system of the 7th scope of the patent application, for the first 502171 VI, the bus of the scope of the patent application, defines the response signal line at the receiving end to notify the other party of the status of the instruction or data, the aforementioned response signal The driving phase of the command is that the command phase output by the command on the bus and the data phase output by the data on the bus are required to be fixed. 9 · If the data transmission system of item 7 of the patent application scope, wherein the aforementioned data transmission control device 'is a bus mediation device that has ownership of the mediation bus in response to the request for the bus right from the aforementioned I / O device, The liberation timing of the bus, the command phase output by the command on the bus, and the data phase output by the data on the bus are required to be fixed. The bus mediation device corresponds to the new bus right. It is required to release each of the command phase and the data phase in execution at a fixed timing. -29-
TW90106210A 2000-03-17 2001-03-16 Data transmitting system TW502171B (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP2000077078A JP2001265711A (en) 2000-03-17 2000-03-17 Device for transferring data and bus system

Publications (1)

Publication Number Publication Date
TW502171B true TW502171B (en) 2002-09-11

Family

ID=18594710

Family Applications (1)

Application Number Title Priority Date Filing Date
TW90106210A TW502171B (en) 2000-03-17 2001-03-16 Data transmitting system

Country Status (2)

Country Link
JP (1) JP2001265711A (en)
TW (1) TW502171B (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7209998B2 (en) 2004-02-04 2007-04-24 Qualcomm Incorporated Scalable bus structure
US7174403B2 (en) * 2005-02-24 2007-02-06 Qualcomm Incorporated Plural bus arbitrations per cycle via higher-frequency arbiter

Also Published As

Publication number Publication date
JP2001265711A (en) 2001-09-28

Similar Documents

Publication Publication Date Title
US7058748B1 (en) ATA device control via a packet-based interface
TWI320142B (en) Bus connection system
US5600793A (en) Method and system of bi-directional parallel port data transfer between data processing systems
CN101911000B (en) Control bus for connection of electronic devices
US8626979B2 (en) Common protocol engine interface for a controller interface
US20040044811A1 (en) System and method for transferring data over a communication medium using double-buffering
JPH0775016B2 (en) Data processing system and data communication bus system
JP4837659B2 (en) Bus controller for processing split transactions
JP2001345828A (en) General-purpose serial bus functioning device and operating method therefor
KR20140045324A (en) Serial interface
JP2008501178A (en) Bus connection device
JP3592547B2 (en) Information processing apparatus and signal transfer method
JPH11238033A (en) Bus for information processor, and method and unit for bus control
US6622191B1 (en) Computer system
TW502171B (en) Data transmitting system
JP3576647B2 (en) Computer system
CN111221769A (en) Single wire read-write communication method
TW449698B (en) Control chipsets and data exchange method among them
US20070131767A1 (en) System and method for media card communication
TW419923B (en) Data transmission apparatus and method thereof
JPH08314854A (en) Data transfer system and device relating to the same
JPH07200432A (en) Data communication method and system-linking device
CN213715913U (en) Micro control system
JP2988879B2 (en) Bus converter
JPH1188381A (en) Communication system, usb function device, communication system control method and medium recording communication condition display program

Legal Events

Date Code Title Description
GD4A Issue of patent certificate for granted invention patent
MM4A Annulment or lapse of patent due to non-payment of fees