TW200822531A - Very-high-speed low-power low-hold pedestal fully differential CMOS sample-and-hold circuit - Google Patents
Very-high-speed low-power low-hold pedestal fully differential CMOS sample-and-hold circuit Download PDFInfo
- Publication number
- TW200822531A TW200822531A TW95140562A TW95140562A TW200822531A TW 200822531 A TW200822531 A TW 200822531A TW 95140562 A TW95140562 A TW 95140562A TW 95140562 A TW95140562 A TW 95140562A TW 200822531 A TW200822531 A TW 200822531A
- Authority
- TW
- Taiwan
- Prior art keywords
- node
- coupled
- circuit
- gold
- complementary
- Prior art date
Links
Landscapes
- Amplifiers (AREA)
Abstract
Description
200822531 九、發明說明: 【發明所屬之技術領域】 本發明係有關於一種取樣保存電路(Sample_and-Hold Circuit)。該種取樣保存電路不僅可以超高速 (very-high-speed)操作,同時具有低保存基底 (1〇\¥-11〇1(1-卩6(1631&1),為低功率(1(^卩0^^1*)的全差動式取樣 保存電路(fully differential sample-and-hold circuit)。 _ 【先前技術】 第1圖為一種傳統的米勒電容取樣保存電路 100(Miller-Capacitance_based Sample-and_Hold circuit,詳 細内容詳述於以下兩篇論文中:/兄J.Cz>CM治, ν〇1· 26, ρρ· 643-651,April 1991 與 /现五 7>ω似.⑽ CfrcWb iSys/ew-/, vol. 45, ρρ· 198-201,February 1998),其中包 括兩個取樣用的金氧半(MOS)開關102與l〇4、兩個保存電 φ 容1〇6與1〇8、——個增益為A的單端放大器ιι〇(其中a為 一數值)、以及一高輸入阻抗單一增益缓衝器(high input-impedance unity-gain buffer) 112。第 1 圖之電容 114 與116分別為該等保存電容108與1〇6之寄生電容。該傳 統米勒電容取樣保存電路〗00在一取樣模式(sample m〇de) 與一保存模式(hold mode)時會具有不同的等效保存電容 值。在該取樣模式時,該等效保存電容值較小,有利於快 速取樣。在該保存模式時,該等效保存電容值較大,可確 保該電路100所保存的電壓具有較低的保存基底㈣d 0955-A21636TWF(N2);g|〇ri〇uSJien 5 200822531 • pedestal),目此’ 5亥傳統米勒電容取樣保存電路⑽可快 速取樣並且具有相當的精確度。 然而’上述MOS開關⑽與1〇4)之導通電阻 (〇n._e)的非線性效應、開關寄生電容⑽滅 P_tlC Capacitor)、以及信號相_關電荷注入效應 (^ignal dependent SWltch charge 皆 米勒電容取樣保存電路⑽的取樣失細咖随)。上述 Μ⑽噴1G2)之麵性效錢姉樣保存電 •的^樣^與穩定時間(穩%㈣的影響尤其顯 電阻必須很小並且為樣開關之導通 社槿勒電絲樣絲_1GG_單端式 :構:、=源雜訊以及電荷注入效應_ 上,亚沒有顯著的功能。 【發明内容】 籲本發明提出-種新穎的超高速低功率低保存基底全差 ’互補式金氧半取樣保存電路,其巾制—種線性化開 I。乍為取樣用的開關’並且採用—全差動式結構卿乂 ⑽此al structure)以有效地消除共模保存基底 h* pedest傘及雜訊。該種線性 超高速低功率低伴存芙底入m 動式…構曾使該 式互補式金氧半取樣保存 二的保保存電容較上述傳統取樣保存電路刚所使 ,、存电谷小。與傳統取樣保存電路1〇〇相較,本發明200822531 IX. Description of the Invention: [Technical Field of the Invention] The present invention relates to a sample storage circuit (Sample_and-Hold Circuit). This kind of sampling and holding circuit can not only operate very-high-speed, but also has a low-storage substrate (1〇\¥-11〇1 (1-卩6(1631&1), low power (1(^ Fully0^^1*) fully differential sample-and-hold circuit. _ [Prior Art] Figure 1 shows a traditional Miller-Capacitance_based Sample 100 -and_Hold circuit, the details are detailed in the following two papers: / brother J.Cz> CM rule, ν〇1·26, ρρ·643-651, April 1991 and / now five 7>ωlike. (10) CfrcWb iSys /ew-/, vol. 45, ρρ· 198-201, February 1998), which includes two gold-oxide half (MOS) switches 102 and l〇4 for sampling, two holding electric φ capacitances 1〇6 and 1 〇8, a single-ended amplifier with a gain of A (where a is a value), and a high input-impedance unity-gain buffer 112. Capacitance of Figure 1. 114 and 116 are the parasitic capacitances of the storage capacitors 108 and 1 〇 6. The conventional Miller capacitance sampling and holding circuit 00 is in a sampling mode (sample M〇de) and a hold mode have different equivalent storage capacitor values. In this sampling mode, the equivalent storage capacitor value is small, which is convenient for fast sampling. In the save mode, The value of the equivalent storage capacitor is large, which ensures that the voltage stored in the circuit 100 has a lower storage base (4) d 0955-A21636TWF(N2); g|〇ri〇uSJien 5 200822531 • pedestal), the goal is 5 The Le Capacitance Sample Preservation Circuit (10) can be quickly sampled and has considerable accuracy. However, the non-linear effect of the on-resistance (〇n._e) of the above MOS switches (10) and 1〇4), the parasitic capacitance of the switch (10), P_tlC Capacitor, And the signal phase_off charge injection effect (^ignal dependent SWltch charge is the sampling of the Miller capacitance sampling and holding circuit (10). The above-mentioned Μ(10) spray 1G2) is used to save the electricity of the ^^^ and the stabilization time (the stability of the (four) effect, especially the apparent resistance must be small and the conductive switch of the sample switch is the 电 电 wire _1GG_ Single-ended: structure: = source noise and charge injection effect _, sub-Asia has no significant function. [Summary] The present invention proposes a novel ultra-high-speed low-power low-storage substrate all-different 'complementary metal oxygen The half-sampling save circuit, which is a linearized open I. The switch is used for sampling' and uses the fully-differential structure (10) to effectively eliminate the common mode storage substrate h* pedest umbrella and Noise. The linear ultra-high-speed, low-power, low-accommodating, and low-powered-in-the-middle-mode-constructed-type gold-oxygen half-sampling storage has a smaller storage capacity than the conventional sampling and storage circuit. Compared with the conventional sampling and holding circuit 1 , the present invention
〇955-A21636TWF(N2);glorj〇us_tien X 6 200822531 :=的取樣保存電路具有更快速的取樣速度以及更好的 广月準又a種新親的取樣保存電路可應用於低功率範圍。 々月所揭路之超高速低功率低保存基底全差動 ί早存^ =半取樣保存電路屬於全差動核構,該取樣 保存电路之輪入端乃由—輪入信號正端以 端所組成,並且該取樣保 號負 -P ^ ^ ^ ^ 予私路之輸出柒乃由一輪出信號 輸出㈣負端所組成。該取樣保存電路包括一 全差動式放大器,該全差動4访士抑目士 书峪包括 垚動式放大态具有一正輪入端、一 負輸入知、一正輸出端、以及一負輸出端。在該 電路中,—第"電絲接於該全差動式放大器正輪出κ端以 及-第-節點之間,一第二電容耦接於該全差 負輸入端以及該第-節點之間,一第三電容搞接於該入為 動式放大器負輸出端以及_第二節點之間,並且王: 容輕接於該全差動式放大器正輸人端以及該第y =電 間。該第一電容與該第三電容具有相同電容值,並且該= 二電容與該第四電容具有相同電容值。此外,本發明= 括一第一線性化開關、一第一互補式傳輸開關、—第 性化開關、以及一第二互補式傳輸開關。該第—與哕第: 線性化開關之導通電阻近似_很小定值。該第一線=化^ 關耦接於該輸入信號正端以及該第一節點之間。該第一 補式傳輸開關耦接於該全差動式放大器正輸出端以及兮入 差動式放大器負輸入端之間。該第二線性化開關耦接 輸入信號負端以及該第二節點之間。該第二互補式傳輸「 關耦接於該全差動式放大器負輸出端以及該全差動式:開 0955-A21636TWF(N2);glorious_tien 一 7 200822531 =輸二端之間。該取樣保存電路尚包—… 第〜緩衝器。該第一緩衝器的輪 姜衝态以 接該第1點與該輸出錢正端。驾·與輪^分別輕 與輸出物_第二節點與該:出的輪入端〇 955-A21636TWF (N2); glorj〇us_tien X 6 200822531 := The sampling and saving circuit has a faster sampling speed and better. The new sampling system can be applied to the low power range. The ultra-high-speed, low-power, low-storage substrate of the road that is revealed by Haoyue is fully differential. The pre-sampling circuit is a fully differential core. The wheel-in terminal of the sample-save circuit is driven by the positive end of the signal. The composition, and the sample protection number negative -P ^ ^ ^ ^ to the private circuit output 柒 is composed of a round of signal output (four) negative end. The sample-storage circuit includes a fully differential amplifier, and the fully-differential 4-visitor 抑 目 峪 includes a positive wheel input terminal, a negative input terminal, a positive output terminal, and a negative Output. In the circuit, a -the "wire" is connected between the full-shift amplifier and the -th node, and a second capacitor is coupled to the all-signal negative input and the first node A third capacitor is connected between the negative output of the input amplifier and the second node, and the king: the light is connected to the positive input of the fully differential amplifier and the first y = between. The first capacitor has the same capacitance value as the third capacitor, and the second capacitor has the same capacitance value as the fourth capacitor. Furthermore, the present invention includes a first linearization switch, a first complementary transfer switch, a parametric switch, and a second complementary transfer switch. The first and the 哕: the on-resistance of the linearized switch is approximately _ small fixed value. The first line = the coupling is coupled between the positive end of the input signal and the first node. The first complementary transfer switch is coupled between the positive output of the fully differential amplifier and the negative input of the differential amplifier. The second linearization switch is coupled between the negative end of the input signal and the second node. The second complementary transmission is "coupled" to the negative output of the fully differential amplifier and the fully differential: open 0955-A21636TWF (N2); glorious_tien-7 200822531 = between the two ends of the input. Still package -... The first buffer is the first buffer of the wheel ginger to pick up the first point and the output money positive end. The drive and the wheel ^ respectively light and output the object _ the second node and the: out Wheeled end
遠超高速低功率低保存基底全差:、/:。 樣健電路可在—取樣模式以及—保金氧半取 述弟-與第二線性化開關以及上述第—邀:木作。當上 開關皆為導通時,該取樣保存電路處於:;:;互補式傳輸 述第-與第二線性化開關以及上述第樣模式。當上 開關皆為不導通時,該取樣保存電路處傳輸 為讓本發明之上述和其他目的、特汽子果式。 顯易懂,下文特舉出較佳實施例,並配;所能= 細况明於以下實施方式中。 。式作评 【實施方式】 、第2圖為本案之超高速低功率低保絲底全差動式互 ,式金氧半取樣健電路的—實施例。超高速低功率低保 子基底王差動式互補式金氧半取樣保存電路2⑻包括一輸 入L唬正jr而Vin+、一輸入信號負端Vin•、一輸出信號正端 V〇m+、一輸出信號負端Vout_、一全差動式放大器2〇2、一 第一電谷Cl、一第二電容C2、一第三電容C3、一第四電 谷c4、一第一線性化開關Sn、一第二線性化開關s,2、一 第一互補式傳輸開關Sn、一第二互補式傳輸開關Si2、一 第一缓衝器204、以及一第二緩衝器2〇6。 該全差動式放大器202之增益值為a,並且具有一正 0955-A21636TWF(N2);g|〇ri〇us_tien 8 200822531 、 輸入端Va_in+、一負輸入端Va_in_、一正輸出端Va_〇+、以及 一負輸出端Va_。、。 遠弟一電谷C〗與該第三電容c3具有相同的電容值。 e亥弟一電谷C2與该第四電容C4具有相同的電容值。該第 一電容C〗耦接於該全差動式放大器202正輸出端Va_〇+以 及一第一節點Vx+之間。該第二電容c2耦接於該全差動式 放大器202負輸入端va_mJ及該第一節點vx+之間。該第 三電容C3麵接於該全差動式放大器202負輸出端Va_·。—以 • 及一第二節點vx-之間。該第四電容C4耦接於該全差動式 放大器202正輸入端va_m+a及該第二節點vx_之間。在第 2圖中,電容C1B、C2B、C3B以及C4B分別為上述電容Ci、 C2、C3 以及 C4 之寄生底板電容(parasitic b〇tt〇n>piate capacitance) ° 該第一線性化開關sn麵接於該輸入信號正端vin+以及 該第一節點Vx+之間。該第二線性化開關S/2耦接於該輸入 信號負端vin_以及該第二節點vx_之間。該第一互補式傳輸 ⑩開關_接於該全差動式放大器202正輪出端va。+以及 該全差動式放大器202負輸入端端Va_in_之間。該第二互補 式傳輸開關Sc |馬接於該全差動式放大器202負輸出端 Va_。-以及該全差動式放大器202正輸入端Va 之間。 該第一緩衝器204的輸入端與輸出端分別耦接該第一 節點vx+與該輸出信號正端v。·。該第二緩衝器206的輸 入端與輸出端分別耦接該第二節點Vx-與該輸出信號負端 V011t_。上述弟一與弟^一緩衝器可為南輸入阻抗單一增益缓 0955-A21636TWF(N2);glorious_tien 9 200822531 衝器(high input-impedance unity-gain buffer)。上述第一或 弟一緩衝器可為一源極隨輕器(source follower)。 當上述開關(Sn、S,2、Sn、以及Sc)皆為導通時,該取 松保存電路200處於一取樣模式(sampie m〇(je)。該全差動 式放大裔202的動作會如同一單一增益(unity_gain)電路,Far more than high speed, low power, low storage base, all poor:, /:. The sample circuit can be in the -sampling mode as well as - the gold-supplied oxygen-half-the second linearization switch and the above-mentioned first - invitation: wood. When the upper switches are all turned on, the sample preserving circuit is at:;;; complementary transmission of the first-and second linearization switches and the first mode described above. When the upper switch is non-conducting, the sample preserving circuit is transmitted for the above and other purposes of the present invention. It is to be understood that the preferred embodiments are set forth below, and that they can be used in the following embodiments. . Evaluation [Embodiment], Fig. 2 is an embodiment of the ultra-high-speed, low-power, low-protection, silk-bottom fully differential mutual-type gold-oxygen half-sampling circuit of the present invention. Ultra-high speed low power low security sub-base differential differential gold-oxygen half-sampling save circuit 2 (8) includes an input L唬正jr and Vin+, an input signal negative terminal Vin•, an output signal positive terminal V〇m+, an output a signal negative terminal Vout_, a full differential amplifier 2〇2, a first electric valley C1, a second capacitor C2, a third capacitor C3, a fourth electric valley c4, a first linearization switch Sn, A second linearization switch s, 2, a first complementary transfer switch Sn, a second complementary transfer switch Si2, a first buffer 204, and a second buffer 2〇6. The fully differential amplifier 202 has a gain value of a and has a positive 0955-A21636TWF (N2); g|〇ri〇us_tien 8 200822531, an input terminal Va_in+, a negative input terminal Va_in_, and a positive output terminal Va_〇 +, and a negative output Va_. ,. The far brother, the electric valley C, has the same capacitance value as the third capacitor c3. The e-Hidden Valley C2 has the same capacitance value as the fourth capacitor C4. The first capacitor C is coupled between the positive output terminal Va_〇+ of the fully differential amplifier 202 and a first node Vx+. The second capacitor c2 is coupled between the negative input terminal va_mJ of the fully differential amplifier 202 and the first node vx+. The third capacitor C3 is connected to the negative output terminal Va_· of the fully differential amplifier 202. - between • and a second node vx-. The fourth capacitor C4 is coupled between the positive input terminal va_m+a of the fully differential amplifier 202 and the second node vx_. In Fig. 2, capacitors C1B, C2B, C3B, and C4B are the parasitic bottom capacitances of the above-mentioned capacitors Ci, C2, C3, and C4, respectively (the parasitic b〇tt〇n>piate capacitance) ° the first linearization switch sn surface Connected between the positive terminal vin+ of the input signal and the first node Vx+. The second linearization switch S/2 is coupled between the input signal negative terminal vin_ and the second node vx_. The first complementary transmission 10 switch is connected to the positive terminal va of the fully differential amplifier 202. + and between the negative input terminal Va_in_ of the fully differential amplifier 202. The second complementary transfer switch Sc | is connected to the negative output terminal Va_ of the fully differential amplifier 202. - and between the positive input terminal Va of the fully differential amplifier 202. The input end and the output end of the first buffer 204 are respectively coupled to the first node vx+ and the output signal positive terminal v. ·. The input end and the output end of the second buffer 206 are respectively coupled to the second node Vx- and the output signal negative terminal V011t_. The above-mentioned brothers and brothers can be a single input impedance single gain slow 0955-A21636TWF (N2); glorious_tien 9 200822531 (high input-impedance unity-gain buffer). The first or second buffer may be a source follower. When the above switches (Sn, S, 2, Sn, and Sc) are all turned on, the take-off save circuit 200 is in a sampling mode (sampie m〇(je). The action of the fully differential amplifying 202 is like a single gain (unity_gain) circuit,
在該全差動式放大器202之輸入與輸出端之間提供一虛短 路(virtual short)。該取樣保存電路2〇〇會將該輸入信號正 端vm+與該輸入信號負端Vm_間的差動輸入電壓取樣至該 等電容(Ci、C2、C3以及Q)上,然後經由上述第一與第二 緩衝器204與206傳遞至該輸出信號正端v。修以及該輸: 信號負端VQut_間。 當上遍開關(Sn、So、Sn、以及Sw)皆不導通時,該取 樣保存電路200處於一保存模式(h〇ld m〇de)。上述電容 (q、C2、c:3以及Co以及該全差動式放大器2〇2會組成一 回授放大器。基於一米勒(Miller)回授效應,該取樣保存電 路200之等效保存電容值會遠大於該取樣模式時的等效2 存電容值。由於上述線性化關S/i與、之導通電阻报小、 並且近似一定值,該取樣保存電路2〇〇的高頻取樣失真可 獲得改善。上述線性化開關8/1與&2可減少信號相關^ 庄入效應(signal-dependent charge injection)。 可 由於該取樣保存電路2〇〇採用全差動式架構,因電^ 注入效應(charge injection)以及時脈滲透效應。『何 feedthrough)所產生的誤差電壓可在該取樣保存電ς' 差動輸出端(V〇ut+,V〇ut_)抵銷。因此,傳統取樣保路之 0955-A21636TWF(N2);gl〇ri〇us^tien 10 200822531 - 保存基底與雜訊問題皆可大幅改善。該取樣保存電路2〇〇 所需要的電容也因此小於傳統取樣保存電路。 在本發明一實施方式中,上述之全差動式放大器2〇2 更包括一放大電路、一電容中和電路、一共模回授電路、 以及一輸出緩衝電路。該放大電路乃用來放大上述正輪入 令而va in+與負輸入端va」n_的信號。放大後的信號會分別傳 送至一第三節點與一第四節點。該電容中和電路耦接該放 大電路,乃用以消除該放大電路在該全差動式放大器2〇2 # 之輸入端所產生的米勒電容(Miller capacitance)效應。該共 模迴授電路乃用來控制該第三節點與該第四節點的一共模 電壓之準位。該輸出緩衝電路具有高輸入阻抗以及低輪出 阻抗,乃用來將該第三節點以及該第四節點之信號分別傳 送至該全差動式放大器202之負輸出端Va—0_以及正輸出端A virtual short is provided between the input and output of the fully differential amplifier 202. The sampling and holding circuit 2 samples the differential input voltage between the positive terminal vm+ of the input signal and the negative terminal Vm_ of the input signal to the capacitors (Ci, C2, C3, and Q), and then passes through the first And the second buffers 204 and 206 are delivered to the positive end v of the output signal. Repair and the input: signal negative terminal VQut_. When the upper pass switches (Sn, So, Sn, and Sw) are not turned on, the sample save circuit 200 is in a save mode (h〇ld m〇de). The above capacitors (q, C2, c: 3 and Co and the fully differential amplifier 2〇2) constitute a feedback amplifier. Based on a Miller feedback effect, the equivalent storage capacitor of the sample preserving circuit 200 The value will be much larger than the equivalent 2 storage capacitor value in the sampling mode. Since the linearization off S/i and the on-resistance are small and approximate to a certain value, the high-frequency sampling distortion of the sampling and holding circuit 2〇〇 can be Improved. The above linearization switches 8/1 and & 2 can reduce signal-dependent charge injection. Since the sample-and-hold circuit 2 uses a fully differential architecture, The charge injection and the clock penetration effect. The error voltage generated by the "feedthrough" can be offset by the sample-storing power supply 'differential output terminal (V〇ut+, V〇ut_). Therefore, the traditional sampling of the road to the 0955-A21636TWF (N2); gl〇ri〇us^tien 10 200822531 - can save the substrate and noise problems can be greatly improved. The capacitance required for the sample holding circuit 2 is therefore smaller than that of the conventional sample holding circuit. In an embodiment of the invention, the fully differential amplifier 2〇2 further includes an amplifier circuit, a capacitor neutralization circuit, a common mode feedback circuit, and an output buffer circuit. The amplifying circuit is for amplifying the signals of the positive wheel input va in+ and the negative input terminal va"n_. The amplified signals are transmitted to a third node and a fourth node, respectively. The capacitor neutralization circuit is coupled to the amplification circuit to eliminate the Miller capacitance effect generated by the amplification circuit at the input of the fully differential amplifier 2〇2#. The common mode feedback circuit is used to control the level of a common mode voltage of the third node and the fourth node. The output buffer circuit has a high input impedance and a low turn-off impedance for transmitting signals of the third node and the fourth node to the negative output terminal Va_0_ and the positive output of the fully differential amplifier 202, respectively. end
Va_0+。 — 第3圖為一互補式金氧半(CMOS)反相器300,其中包 括一 NMOS電晶體Mn以及一 pM〇s電晶體。該等電晶 脰Mp與Mn之笔何遷移率charge carrier)分別為 ~與μρ、氧化層單值面積電容分別為c_與、有效通 逞寬度/長度分別為Wn/Ln與Wp/Lp L^^.Wn/Ln、 /V=/vCoxp’Wp/Lp。當A為時,該CMOS反相器300的小 信號電壓增益為固定值。該CMOS反相器工作於AB類, 故消耗較低功率。 第4圖為上述放大電路與電容中和電路的一實施例。 該放大電路可由一第一 CM〇s反相器Ιηνι以及一第二 0955-A21636TWF(N2);g!ohous_tien 200822531 CMOS反相器InV2實現。上述CMOS反相器Iim以及Inv2 的小信號電壓增益為固定值。該電容中和電路包括一第一 NM0S電晶體Mnl、一第二NM0S電晶體Mn2、一第一 PMOS 電晶體Mpl、以及一第二pm〇S電晶體Mp2。上述全差動 式放大器200之正輪入端va_m+耦接該第一 CMOS反相器 Invii輸入端、負輸入端Va」n_耦接該第二CMOS反相器 Inv2之輸入端。該第一 nm〇S電晶體Mnli源極與汲極以 及該第一 PM0S電晶體Mpl之源極與汲極皆耦接該第二 CMOS反相器lnV2之輸出端。該第一 nm〇S電晶體Mnl 與該第一 PM0S電晶體Mpl2閘極耦接該正輸入端va」n+。 該第二NM0S電晶體mii2之源極與汲極以及該第二pm〇S 電晶體MP2之源極與汲極皆耦接該第一 CM〇s反相器Ιηνι 之輸出端。該第二NM0S電晶體與該第二PMOS電晶 體Mp22閘極耦接該負輪入端%心。該第一 NM〇s — I曰曰 體Mnl之寬長比(W/L)為該第一 CM〇s反相器InVi之Nm〇s 電晶體之寬長比的一半。該第:NM〇s電晶體Mu之寬長 比為該第二CMOS反相器、之NM〇s電晶體之寬長比 的-半。該第-PM0S電晶體、之寬長比為該第—C]y[〇s 反相器_之PM〇S電晶體之寬長比的一半。該第二PM〇s 電晶體Mp2之寬長比為該第二CM〇s反相器—之心 電晶體之寬長比的-半。上述電晶體HA以及 Mp2所組成的電容中和電路可大幅改善保存基底誤差。 第5圖為本發明之全差動式放大器的一實施方式。一 全呈動式放大器5GG採用f 4圖所示之放大電路以及電容 0955-A21636TWF(N2);gloriousJien 200822531Va_0+. - Figure 3 is a complementary MOS inverter 300 comprising an NMOS transistor Mn and a pM 〇s transistor. The charge carriers of the electric crystals Mp and Mn are respectively - and μρ, the single-value area capacitance of the oxide layer is c_ and the effective overnight width/length are Wn/Ln and Wp/Lp L, respectively. ^^.Wn/Ln, /V=/vCoxp'Wp/Lp. When A is , the small signal voltage gain of the CMOS inverter 300 is a fixed value. The CMOS inverter operates in class AB and therefore consumes less power. Fig. 4 is an embodiment of the above amplifying circuit and capacitor neutralizing circuit. The amplifying circuit can be implemented by a first CM 〇s inverter Ιηνι and a second 0955-A21636TWF(N2); g!ohous_tien 200822531 CMOS inverter InV2. The small signal voltage gains of the above CMOS inverters Iim and Inv2 are fixed values. The capacitor neutralization circuit includes a first NMOS transistor Mn1, a second NMOS transistor Mn2, a first PMOS transistor Mpl, and a second pmS transistor Mp2. The positive input terminal va_m+ of the fully differential amplifier 200 is coupled to the first CMOS inverter Invii input terminal, and the negative input terminal Va"n_ is coupled to the input terminal of the second CMOS inverter Inv2. The source and the drain of the first nm 〇S transistor Mnli and the source and the drain of the first PMOS transistor Mpl are coupled to the output of the second CMOS inverter lnV2. The first nm 〇S transistor Mn1 is coupled to the first PMOS transistor Mpl2 to the positive input terminal va"n+. The source and the drain of the second NMOS transistor mii2 and the source and the drain of the second pmS transistor MP2 are coupled to the output of the first CM〇s inverter Ιηνι. The second NMOS transistor and the second PMOS transistor Mp22 are coupled to the negative wheel terminal. The aspect ratio (W/L) of the first NM 〇 s - I 体 body Mn1 is half of the width-to-length ratio of the Nm 〇s transistor of the first CM 〇 s inverter InVi. The width-to-length ratio of the first:NM〇s transistor Mu is -half of the aspect ratio of the NM〇s transistor of the second CMOS inverter. The width-to-length ratio of the first-PMOS transistor is half of the aspect ratio of the PM-S transistor of the first-C]y[〇s inverter_. The width-to-length ratio of the second PM〇s transistor Mp2 is -half of the aspect ratio of the second CM〇s inverter. The capacitance neutralization circuit composed of the above transistors HA and Mp2 can greatly improve the preservation substrate error. Figure 5 is an embodiment of a fully differential amplifier of the present invention. A full-embedding amplifier 5GG uses the amplification circuit shown in Figure 4 and the capacitance 0955-A21636TWF(N2); gloriousJien 200822531
中和電路。該全差動式放大器500還包括一共模迴授電路 502,其中包括一第三CMOS反相器Inv3、一第四CMOS 反相器I11V4、一第五CMOS反相器Inv5以及一第六CMOS 反相器Inv6。該共模迴授電路502與該第一 CMOS反相器 Iirv〗之輸出端柄接於一第三節點Vint-,與該第二CMOS反 相器Inv2之輸出端耦接於一第四節點Vmt+。該第三CMOS 反相器Inv3之輸入端耦接該第三節點Vmt_,輸出端與該第 四CMOS反相器Inv4之輸入與輸出端耦接於該第四節點 Vmt+。該第六CMOS反相器Inv6之輸入端耦接該第四節點 Vint+,輸出端與該第五CMOS反相器inV5之輸入與輸出端 耦接於該第三節點Vint_。該共模回授電路502乃用來控制 該弟四郎點Villt+與该弟二郎點Vin“之共模(c〇mmon-mode) 電壓準位。為簡化起見先假設上述CMOS反相器Inv3、 I11V4、I11V5以及Inv6可具有匹配的電路特性。 假設該等CMOS反相器Inv3、Inv4、jnV5以及inV6所 使用的NMOS以及PMOS電晶體全部操作於強反轉(strong inversion)及飽和(saturation)區,則每一個上述CM〇s反相 器Inv3、I11V4、Invs以及Inv6之輸出電流與其輸入電 壓(Vin)的關係為:Iout=gm.(Vin-Vc),其中,一個單獨的反相 器之轉導(transconductance)gm與電壓%如下:Neutralization circuit. The fully differential amplifier 500 further includes a common mode feedback circuit 502 including a third CMOS inverter Inv3, a fourth CMOS inverter I11V4, a fifth CMOS inverter Inv5, and a sixth CMOS counter. Phaser Inv6. The common mode feedback circuit 502 and the output end of the first CMOS inverter Iirv are connected to a third node Vint-, and the output end of the second CMOS inverter Inv2 is coupled to a fourth node Vmt+ . The input terminal of the third CMOS inverter Inv3 is coupled to the third node Vmt_, and the output terminal and the input and output terminals of the fourth CMOS inverter Inv4 are coupled to the fourth node Vmt+. The input end of the sixth CMOS inverter Inv6 is coupled to the fourth node Vint+, and the output end and the input and output terminals of the fifth CMOS inverter inV5 are coupled to the third node Vint_. The common mode feedback circuit 502 is used to control the common mode (c〇mmon-mode) voltage level of the younger brother Villt+ and the brother Erlang point Vin. For the sake of simplicity, the above CMOS inverter Inv3 is assumed. I11V4, I11V5, and Inv6 can have matched circuit characteristics. It is assumed that the NMOS and PMOS transistors used in these CMOS inverters Inv3, Inv4, jnV5, and inV6 are all operated in strong inversion and saturation regions. The relationship between the output current of each of the above-mentioned CM〇s inverters Inv3, I11V4, Invs, and Inv6 and its input voltage (Vin) is: Iout=gm.(Vin-Vc), where a single inverter is used. Transconductance gm and voltage % are as follows:
+ L,Vtll 與 Vtp 分 gm 二 /5inv.(Vdd+Vtp-Vtn),Vc〉 別為NMOS電晶體與PMOS電晶體之臨界電壓“泣以!^^ voltage)。其中’若每一單獨反相器之札,則 0955-A21636TWF(N2);glorious_fien 200822531 V —匕〆"Ά 〇 C 2 上述CMOS反相器Inv3、Inv4、Ιην5以及ιην6之輸出 電流與其輸入電壓的關係分別為: I〇3 = gm3.(Vmt,VC)、 I〇4 - gm4’(Vint+ - Vc)、 I〇5 = gm5.(Vint,Vc)、 I〇6 = gm6,(Vint+ - Vc) 0 上述第二以及第四節點 Vint_與 Vint+可表示為 Vint-—以及 Vint+—Vintc+Vintd/2 ’ 其中,内部 才莫 (internal common-mode voltage)電壓為 Vintc、内部差模電舞 (internal differential-mode voltage)為 Vintd。因此,可得以 下關係式: I〇3 + I〇4 = (gm3 + gm4)( Vintc - Vc) + (gm4 - gm3) ·%福/2、 I〇5 + 1。6 二(gm5 + gm6)( Vmtc - Vc) — (gm5 - gm6) ·Υιηί(1/2 〇 如弟5圖所不’上述CMOS反相器I11V4以及Inv5如同電阻 般跨接在該第四節點vmt+、該第三節點Vlnt_、以及該共模 電壓Vc之間,其電阻值分別為gm4以及丨/ gm5。上述 CMOS反相器I11V3以及inV6分別將電流gm3.dVmt )以及 gmdVc-VmnO注入上述電阻中。對共模輸出信號而言,該第 二卽點Vint_上有一虛負載(vjrtuai i〇a(j),其值為 l/(gm5 + gm6) ·’該第四節點Vmt+上之虛負載為l/(gnl3+gm4)。 對差動輸出號而言,若gm3 = gm4並且=gm6,則上述 第三與第四節點(vmt-與Vmt+)為無負載。 在本發明的一實施例中,上述CMOS反相器InV3、 0955-A21636TWF(N2);gl〇rious_tien 14 200822531+ L, Vtll and Vtp are divided into gm two/5inv. (Vdd+Vtp-Vtn), Vc> is the threshold voltage of the NMOS transistor and the PMOS transistor. "When! ^^ voltage". For the phase device, 0955-A21636TWF(N2); glorious_fien 200822531 V —匕〆"Ά 〇C 2 The relationship between the output currents of the above CMOS inverters Inv3, Inv4, Ιην5 and ιην6 and their input voltages are: I〇 3 = gm3.(Vmt,VC), I〇4 - gm4'(Vint+ - Vc), I〇5 = gm5.(Vint,Vc), I〇6 = gm6,(Vint+ - Vc) 0 The fourth nodes Vint_ and Vint+ can be expressed as Vint-- and Vint+-Vintc+Vintd/2', wherein the internal common-mode voltage is Vinct, internal differential-mode voltage It is Vindd. Therefore, the following relationship can be obtained: I〇3 + I〇4 = (gm3 + gm4)( Vintc - Vc) + (gm4 - gm3) ·%福/2, I〇5 + 1. 6 II ( Gm5 + gm6)( Vmtc - Vc) — (gm5 - gm6) ·Υιηί(1/2 〇如弟5图不的'The above CMOS inverters I11V4 and Inv5 are connected as resistors at the fourth node vmt+, third Between the point Vlnt_ and the common mode voltage Vc, the resistance values are gm4 and 丨/gm5, respectively. The CMOS inverters I11V3 and inV6 respectively inject currents gm3.dVmt) and gmdVc-VmnO into the above-mentioned resistors. In terms of the output signal, there is a virtual load on the second point Vint_ (vjrtuai i〇a(j), whose value is l/(gm5 + gm6) · 'The virtual load on the fourth node Vmt+ is l/( Gnl3+gm4) For the differential output number, if gm3 = gm4 and = gm6, the third and fourth nodes (vmt- and Vmt+) are unloaded. In an embodiment of the invention, the above CMOS Inverter InV3, 0955-A21636TWF(N2); gl〇rious_tien 14 200822531
Inv4、hiv5以及Inv6之電路特性完全匹配(perfectly matched) 並且電壓源皆為vdd,因此具有相同的轉導(gm3=gm4=gm5 =gm6 )。對共模輸出信號而言,該共模迴授電路502提供 了一個低歐姆負載,對差模輸出信號而言,該共模迴授電 路502提供了 一個近似開路的負載。因此靜態共模電壓The circuit characteristics of Inv4, hiv5, and Inv6 are perfectly matched and the voltage sources are both vdd, thus having the same transduction (gm3 = gm4 = gm5 = gm6). The common mode feedback circuit 502 provides a low ohmic load for the common mode output signal, and the common mode feedback circuit 502 provides an approximately open load for the differential mode output signal. Static common mode voltage
(quiescent common-mode voltage)會是 Vc。若上述 CMOS(quiescent common-mode voltage) will be Vc. If the above CMOS
反相器In”、In%、Invs以及InV6(/3n=/5p=/5mv)的假設不成立, 但電路特性仍然完全匹配,則對共模信號而言,負載電阻 是非線性(nonlinear):對差動信號而言,所有偶數(even) 及奇數(odd)非線性項都會被銷除,詳細内容詳述於此篇論 文中· IEEE J· Solid-State Circuits, νο\· 27,ρρ Ά2The assumptions of inverters In", In%, Invs, and InV6 (/3n=/5p=/5mv) are not true, but the circuit characteristics are still perfectly matched. For common mode signals, the load resistance is nonlinear: For the differential signal, all even (even) and odd (odd) nonlinear terms are eliminated. Details are detailed in this paper. IEEE J· Solid-State Circuits, νο\· 27, ρρ Ά2
February 1992 0 ’ 如第5圖所示,該全差動式放大器的輪出緩衝電 路乃由-第-源極隨耦器Buffi以及一第二源極隨耦哭 耐f2所組成。該第三節點V』接該第一源極隨輕哭财; 之輸入端。該第-源極隨耗器減丨之輸出端•馬接:入1 至動式放大器500之負輸出端u第四節 二 該第二源極隨輕器响之輸入端。該第二源極If Buff2之輸接至該全差動式放大器5⑽之 V:_。^由於土述源極隨麵器喊與β吨具有低:二 抗,會導致該全差動式放大器' 5⑻具有較大的頻寬^有助 ,„之超高速低功率低保存基底全差動式互補式 至氧半取樣保存電路之取樣速度。 式 線性 弟6圖為本案之線性化關的—種實施方式 0955-A21636TWF(N2);g|〇rj〇us_tien 15 200822531 化開關600包括—第三 電晶體Μ 4、一箆$二 S電晶體Μη3、〜第四NM〇s 电日日肢Mn4 弟五NMOS電$ _ M 、一淫 晶體Μ 6、一筮冬ΧΤΛ 电日日肢、一弟六NMOS電 :116弟七NM〇S電晶體Μη7、-第三PMOS中曰 脰mp3、一源極隨耦器6〇2。該曰 书日日 源極與汲極分別耦接哕 Λ 一 $晶體的 V。,,。該第四難〇=:開關的輸入…輪出端 第三NM0S恭曰雕A n4之源極接地、汲極耦接該February 1992 0 ′ As shown in Figure 5, the full-amplifier amplifier's turn-out buffer circuit consists of a -first-source follower Buffi and a second source-coupled crying-resistant f2. The third node V′′ is connected to the first source and is crying; The output of the first-source follower is reduced. • The input is connected to the negative output of the dynamic amplifier 500. The fourth section of the second source is connected to the input of the lighter. The second source If Buff2 is connected to V:_ of the fully differential amplifier 5 (10). ^Because the source of the face-to-face screamer has a low beta: beta, it will cause the fully differential amplifier '5(8) to have a larger bandwidth^Help, „the ultra-high speed, low power, low storage base, all the difference The sampling speed of the dynamic complementary to oxygen half-sampling preserving circuit. The linear linear 6 diagram is the linearization of the case - an embodiment of the method 0955-A21636TWF (N2); g|〇rj〇us_tien 15 200822531 The switch 600 includes - The third transistor Μ 4, one 箆 $ two S transistor Μ 3 3, ~ the fourth NM 〇 s electric Japanese and Japanese limbs Mn4 弟 five NMOS electric $ _ M, a kinky crystal Μ 6, a 筮 winter ΧΤΛ electric day limbs, One brother six NMOS: 116 brother seven NM 〇 S transistor Μ η7, - third PMOS 曰脰 mp3, a source follower 6 〇 2. The source of the day and the pole are coupled respectively 哕Λ A $ crystal V.,.. The fourth difficulty =: switch input... wheel output third NM0S Gong Yan carving A n4 source ground, the pole is coupled to the
甩日日版n3之閘極。該第四NM〇s電 閘極的電壓由一時脈_轳电日日版Mn4 m於 1°號的反相信號卩)控制。該源極隨耦 二602之輸入端與輪出端分別耦接該第三顧OS電㈣ I之源極與閘極,以令該第三NMOS電晶體Mn3之閘^ 與源極的電壓差為定值。嗜箓 與該第三P_s電晶/M iNM〇S電晶體Mn5之沒極 _之輸出端〃二;五L之源;!咖線_^^ h卜一 MOS電晶體Mn5之源極與該 弟三PM0S電晶體Mp3之没麵皆雛該第六麵s電晶體The Japanese version of the n3 gate. The voltage of the fourth NM 〇s gate is controlled by a clock signal 反相 卩 轳 Mn Mn Mn Mn 。 。 。 。 。 。 。 。 。 。). The input end and the output end of the source with the coupling 602 are respectively coupled to the source and the gate of the third (4) I, so that the voltage difference between the gate and the source of the third NMOS transistor Mn3 Is a fixed value. The eosinophilic and the third P_s electro-crystal/M iNM〇S transistor Mn5 is the output terminal ;2; the source of the five L; the coffee line _^^ h _ MOS transistor Mn5 source and the The third PM0S transistor Mp3 has no face, the sixth surface s transistor
Mn6之錄。該第六NM0S電晶體I之源極接地。該第 五NM0 S電晶體Μn5之閘極電壓由上述時脈信號之反相信 號⑷控制。該第六NM0S電晶體I與該第三pM〇s電 晶體My之閘極電塵皆由上述時脈信號控制。該第七 NM0S電晶體Mn7之源極耗接電壓源Vdd,汲極辆接該線 性化開關600之輸出端ν_,閘極輕接該第六NM〇s電晶 體Μηό之〉及極端。该弟七NMOS電晶體]yjn7乃用來消除該 第二NMOS電晶體Μα之閘汲極重疊電容(gate-drain overlap capacitance)所產生的信號相關滲透效應 (single-dependent feedthrough) 〇 0955-A21636TWF(N2);glorious_tien 200822531 弟7、8、9、10以及11圖為該超咼速低功率低保存基 底全差動式互補式金氧半取樣保存電路200採用該全差動 式放大斋500以及該線性化開關600時的模擬結果,其模 擬軟體為Synopsys Hspice,模擬的製程為TSMC 〇.35/xm MIXED MODE (2P4M,3.3V/5V) POLYCIDE。如第 7 圖所Record of Mn6. The source of the sixth NMOS transistor I is grounded. The gate voltage of the fifth NM0 S transistor Μn5 is controlled by the inverse signal (4) of the above clock signal. The gate galvanic dust of the sixth NMOS transistor I and the third pM 〇s transistor My is controlled by the clock signal. The source of the seventh NMOS transistor Mn7 consumes the voltage source Vdd, the drain is connected to the output terminal ν_ of the linearization switch 600, and the gate is lightly connected to the sixth NM 〇s transistor 及 ό and the extreme. The seven-NMOS transistor yjn7 is used to eliminate the signal-dependent feedthrough effect of the gate-drain overlap capacitance of the second NMOS transistor 〇α 〇0955-A21636TWF ( N2);glorious_tien 200822531 brothers 7, 8, 9, 10 and 11 are the super-idle low-power low-storage substrate full differential complementary metal-oxygen half-sampling preservation circuit 200 using the fully differential amplifier 500 and the The simulation result of the linearization switch 600 is Synopsys Hspice, and the simulation process is TSMC 〇.35/xm MIXED MODE (2P4M, 3.3V/5V) POLYCIDE. As shown in Figure 7
示,該取樣保存電路之輸入差動正弦波(sinus〇idal)信號7〇2 之峰對峰(peak-to-peak)擺幅為1.2V,頻率為 8^24MHZ(f^=80.24MHz)。該取樣保存電路所採用的時脈 信號706之頻率為330MHz,該時脈信號7〇6的轉態時間 (transition time,ts)為(Uns。該取樣保存電路所產生的輸 出差動信號為信號704。 弟8圖為本模敎贿基底賴擬結果,橫軸為輸入 信號正、負端(Vin+,Vm_)之差動電壓差,縱軸為輸出信號 正、負端(vout+、v〇ut0所量測到的差動保存基底㈣d Ped_⑽變化。很明顯地,在轉態時間是ο -時,在差 =輸入電壓邊6V的範圍内’該差動保存基底受差動輸入 =.影響不大’故該取樣保存電路的__灿差動輸入範 圍(伽她al lnput咖㈣是。在轉態時間是Oh ns時’該差動保存基底皆小於或等於0篇。鱼傳 =二式取樣保存電路相比’本發明之取樣保存電路的 且Γ ’本發明之取樣保存電路可以實現超 回速取松亚且具有低保存基底。 地=二為在不同端點所測量到的保存基底。报明顯 虎正、負端(V_+、V-)所量測到的差動保存基 0955-A21636TWF(N2);g|〇rj〇us__tjen 17 200822531 * 底902較單端輪出信號端(ν_+或ν__)所量測到的保存基 底小。 第10圖為該輸出差動信號(vQUt+,v。^)進行快速傅立 葉轉換(Fast Fourier Transfom,FFT)後所產生的結果。該輸 入差動信號702的頻率(fm=80.24MHz,1.2Vpp)造成標示 1002之柱狀頻譜。一第三階諧波(|;hird_〇rder harmonic)如標 示1104之柱狀頻議。如第1〇圖所示,本發明之全差動式 架構大幅降低偶數階(even-order harmonics)的譜波,並且總 括諧波失真量(Total Harmonic Distortion, THD)為-68.3dB。 第11圖為該輸入差動信號702之頻率由ΙΟΜΉζ至 120MHz時(fln=10〜120MHz,L2Vpp),對該輸出差動信號 進行快速傅立葉轉換的結果。在輸入差動信號702之頻率 為90MHz時,該總括諧波失真量為-67.98dB,該取樣保存 笔路之線性度相當於11有效位元數(eIYective nuni|3er 0f bits)。在更低的輪入頻率範圍中,該取樣保存電路之線性 度甚至更好。 • 為了比較本發明與其他傳統技術的效能,本發明以一 Figure of Ment (F.aM)數值作為比較的依據。其中,該 F.O.M的計算方法為: F.O.M二一^^,The peak-to-peak swing of the input differential sine wave (sinus〇idal) signal 7〇2 of the sampling and holding circuit is 1.2V, and the frequency is 8^24MHZ (f^=80.24MHz). . The frequency of the clock signal 706 used by the sample preserving circuit is 330 MHz, and the transition time (ts) of the clock signal 7〇6 is (Uns. The output differential signal generated by the sample preserving circuit is a signal. 704. The younger brother is the result of the model, the horizontal axis is the differential voltage difference between the positive and negative ends of the input signal (Vin+, Vm_), and the vertical axis is the positive and negative ends of the output signal (vout+, v〇ut0). The measured differential storage base (4) d Ped_(10) changes. Obviously, when the transition time is ο -, in the range of difference = input voltage side 6V 'the differential preservation substrate is differential input =. The big _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ The sample-storage circuit is compared to the 'sample-storage circuit of the present invention' and the sample-save circuit of the present invention can achieve super-return speed and has a low-storage substrate. Ground = two is a storage substrate measured at different end points Reporting the differential measured by the positive and negative ends (V_+, V-) Preservation base 0955-A21636TWF(N2);g|〇rj〇us__tjen 17 200822531 * The base 902 is smaller than the single-end round-out signal terminal (ν_+ or ν__). Figure 10 shows the output differential. The result of the fast Fourier Transfom (FFT) signal (vQUt+, v.^). The frequency of the input differential signal 702 (fm = 80.24 MHz, 1.2 Vpp) results in a columnar spectrum labeled 1002. A third-order harmonic (|;hird_〇rder harmonic) is labeled as a columnar frequency of 1104. As shown in Figure 1, the fully differential architecture of the present invention substantially reduces even-order harmonics. The spectral wave, and the Total Harmonic Distortion (THD) is -68.3 dB. Figure 11 shows the frequency of the input differential signal 702 from ΙΟΜΉζ to 120 MHz (fln=10~120MHz, L2Vpp), The output differential signal is subjected to fast Fourier transform. When the frequency of the input differential signal 702 is 90 MHz, the total harmonic distortion amount is -67.98 dB, and the linearity of the sample holding stroke is equivalent to 11 effective bits. (eIYective nuni|3er 0f bits). At lower round-in frequency range , The linearity of the sample and hold circuits even better. • In order to compare the present invention with the performance of other traditional techniques, the present invention in a Figure of Ment (F.aM) value as a basis for comparison. Among them, the calculation method of F.O.M is: F.O.M 二一^^,
2 麵 X 八中ENOB為有效值元數(effect|ve number 〇f biis),忍為 取樣頻率,P為取樣保存電路的功率消耗。圖表i所提及 的參考文獻包括: 麥考文獻 1 ’ A. B〇m,A Pierazzi,and c M〇㈣di,“A ⑺一匕 0955-A21636TWF(N2);glorious_t}en 200822531 • 185MS/s track-and-hold in 0.35- //m CMOS/5 IEEE J. Solid-State Circuits, vol. 36? pp. 195-203, Feb. 2001 ; 參考文獻 2,A· Baschirotto, “A low-voltage sample-and_hold circuit in standard CMOS technology operating at 40 Ms/s,’’ IEEE Trans, on Circuits and Systems-II, vol. 48, pp. 394-399, Apr. 2001 ;以及 參考文獻 3,Τ· S· Lee and C. C· Lu,“A 1.5V 50MHz Pscudo~Diffcrcxiti3.1 CIViOS S3,rtiplc~3,rici-PJ〇J(J Circuit \vith 鲁 Low Hold Pedestal, IEEE Transactions on Circuits and /,vol· 52, no· 9, pp· 1752-1757, 2005。 由圖表1可知,本發明的取樣保存電路的效能遠遠超 越傳統取樣保存電路。在此模擬中,在操作電壓源電壓為 3V下,本發明的取樣保存電路的取樣頻率高達33〇MHz, full-scale差動輸入範圍高達丨.2 Vpp,在輪入差動信號的頻 率80.24MHz時,總括諧波失真量為_68 3dB,並且有效位 元數朗11位元以上。本發明之差動輪出保存基底在 • 〇.8mV 以下。本發明之 F.aM 為 3.9G6x1(r2(MW/MHz)。2 face X 八 ENOB is the effective value element (effect|ve number 〇f biis), forbearing the sampling frequency, P is the power consumption of the sample preservation circuit. The references mentioned in Figure i include: Macquarie Document 1 'A. B〇m, A Pierazzi, and c M〇 (4) di, “A (7) 匕 0955-A21636TWF (N2); glorious_t}en 200822531 • 185MS/s Track-and-hold in 0.35- //m CMOS/5 IEEE J. Solid-State Circuits, vol. 36? pp. 195-203, Feb. 2001 ; Reference 2, A· Baschirotto, “A low-voltage sample -and_hold circuit in standard CMOS technology operating at 40 Ms/s,'' IEEE Trans, on Circuits and Systems-II, vol. 48, pp. 394-399, Apr. 2001; and reference 3, Τ·S· Lee And C. C· Lu, “A 1.5V 50MHz Pscudo~Diffcrcxiti3.1 CIViOS S3, rtiplc~3, rici-PJ〇J (J Circuit \vith Lu Low Hold Pedestal, IEEE Transactions on Circuits and /, vol· 52, No. 9, pp·1752-1757, 2005. As can be seen from the graph 1, the performance of the sampling and holding circuit of the present invention far exceeds the conventional sampling and holding circuit. In this simulation, the operating voltage source voltage is 3V, the present invention The sampling and saving circuit has a sampling frequency of up to 33〇MHz, and the full-scale differential input range is up to 22Vpp. When the frequency of the number is 80.24MHz, the total harmonic distortion amount is _68 3dB, and the effective bit number is more than 11 bits. The differential wheel storage substrate of the present invention is below 〇.8mV. The F.aM of the present invention is 3.9G6x1 (r2 (MW/MHz).
參考嫌1 參考德2 製程 0.35-/xm CMOS 〇.5-μπι CMOS 取樣頻率 185MHz 40MHz 0955-A21636TWF(N2);g|〇ri〇us_tien ^ 200822531 foll-scale差動輸 入範圍 ι.〇νρρ 〇.6Fpp 0.8 Vpp 12VPP 保存基底 ^5.0mV ^0.8mV ^0.8mV 下降率 (droop rate) ^1.0mV/us —1.2mV/us ^0.8mV/us 總括諧波失真量 45MHz 時 2MHz 時 2.5MHz 時 80.24MHz 時 -63dB -50dB -58.2dB -68.3dB 電壓源 3.3V 1.2V 1.5 V 3V 功率消耗 70mW 1.2mW 2.6mW 26.4mW ENOB(bit) 10 8 9 11 FOM (μΨ/ΜΗζ) 3.695X10'1 1.171X10'1 1.015xl0_1 3.906xl0"2 圖表1 本發明雖以較佳實施例揭露如上,然其並非用以限定 本發明的範圍,任何熟習此項技藝者,在不脫離本發明之 精神和範圍内,當可做些許的更動與潤飾,因此本發明之 保護範圍當視後附之申請專利範圍所界定者為準。 0955-A21636TWF(N2);glorious_tien 20 200822531 【圖式簡單說明】 # / 一種傳統的米勒電容取樣保存電路; 補式金氧半轉=^速低功率健縣底全差動式互 ^ 杈保存電路的一實施例; ί ::t:互補式金氧半(CM0S)反相器; ^ 發明之放大電路與電容中和電路& ^^ 施方式; 七丁不冤路的一種實Reference 1 Reference De 2 Process 0.35-/xm CMOS 〇.5-μπι CMOS sampling frequency 185MHz 40MHz 0955-A21636TWF(N2);g|〇ri〇us_tien ^ 200822531 foll-scale differential input range ι.〇νρρ 〇. 6Fpp 0.8 Vpp 12VPP Preservation substrate ^5.0mV ^0.8mV ^0.8mV drow rate ^1.0mV/us —1.2mV/us ^0.8mV/us Total harmonic distortion 45MHz 2MHz at 2.5MHz 80.24MHz -63dB -50dB -58.2dB -68.3dB Voltage Source 3.3V 1.2V 1.5 V 3V Power Consumption 70mW 1.2mW 2.6mW 26.4mW ENOB(bit) 10 8 9 11 FOM (μΨ/ΜΗζ) 3.695X10'1 1.171X10' 1 1.015xl0_1 3.906xl0"2 Figure 1 The present invention is disclosed in the above preferred embodiments, and is not intended to limit the scope of the present invention, and those skilled in the art, without departing from the spirit and scope of the present invention, A number of changes and modifications may be made, and the scope of the present invention is defined by the scope of the appended claims. 0955-A21636TWF(N2);glorious_tien 20 200822531 [Simplified illustration] # / A traditional Miller capacitor sampling and saving circuit; Complementary gold and oxygen half-turn = ^ speed low power Jianxian bottom full differential mutual 杈 save An embodiment of the circuit; ί:t: complementary CMOS (CMOS) inverter; ^ amplifying circuit and capacitor neutralization circuit of the invention & ^^ implementation mode;
^ 5圖為本發明之全差動式放大器的 =為本發明之線性化開關的一種實施二f 圖為本發明的-實施例之模擬結果。 【主要元件符號說明】 100〜傳統米勒電容取樣保存電路; 102、HM〜MOS 開關; 106、1〇8〜保存電容,· 110〜增益為A的單端放大器; 112〜尚輸入阻抗單一增益緩衝器; 114、116〜保存電容1〇6與1〇8之寄生電容; 200〜本發明之超高速低功率低保存基底全差動式互補 式金氧半取樣保存電路; 202〜全差動式放大器; 204、206〜緩衝器; 300〜CMOS反相器; 500〜全差動式放大器; 0955-A21636TWF(N2);gl〇rious_tien 21 200822531 - 502〜共模迴授電路; 600〜線性化開關; 602〜源極隨耦器; 702〜輸入差動信號; 704〜輸出差動信號; 706〜時脈信號; 902〜輸出信號正、負端(ν_+、Vout_)所量測到的差動保 存基底; • 1002〜該輸入差動信號702的頻率(fm=80.24MHz)對第 10圖之頻譜所造成的影響; 1004〜第三階諧波造對第10圖之頻譜所造成的影響; Buffi、Buff]〜源極隨輕器;Figure 5 is a full differential amplifier of the present invention = an implementation of the linearized switch of the present invention. Figure 2 is a simulation result of the embodiment of the present invention. [Main component symbol description] 100~ traditional Miller capacitor sampling and saving circuit; 102, HM~MOS switch; 106, 1〇8~ save capacitor, · 110~ gain single-ended amplifier of A; 112~ input impedance single gain Buffer; 114, 116~ save the parasitic capacitance of capacitors 1〇6 and 1〇8; 200~ ultra-high speed low power low-storage substrate full differential complementary gold-oxygen half-sampling save circuit of the invention; 202~ full differential Amplifier; 204, 206~ buffer; 300~CMOS inverter; 500~ fully differential amplifier; 0955-A21636TWF(N2); gl〇rious_tien 21 200822531 - 502~ common mode feedback circuit; 600~ linearization Switch; 602 ~ source follower; 702 ~ input differential signal; 704 ~ output differential signal; 706 ~ clock signal; 902 ~ output signal positive and negative (ν_+, Vout_) measured difference Dynamically save the substrate; • 1002~ The effect of the frequency of the input differential signal 702 (fm=80.24MHz) on the spectrum of Figure 10; The effect of the 1004~3rd harmonic generation on the spectrum of Figure 10 Buffi, Buff] ~ source with light device;
Ci、c2、c3、c4〜電容;Ci, c2, c3, c4~ capacitor;
CiB、C2B、〇3Β 以及 C4B〜電容 Cl、C2、C3、C4 之寄生 底板電容; ck〜時脈信號; 馨 Invrlnv6〜CMOS反相器;CiB, C2B, 〇3Β and C4B~capacitors Cl, C2, C3, C4 parasitic backplane capacitance; ck~clock signal; Xin Invrlnv6~CMOS inverter;
Mn、Mnl-Mn7〜NMOS 電晶體;Mn, Mnl-Mn7~NMOS transistor;
Mp、Mpl-Mp3〜PMOS 電晶體; S/l、S/2〜線性化開關; 、si2〜互補式傳輸開關; va_m+〜正輸入端;Mp, Mpl-Mp3~PMOS transistor; S/l, S/2~ linearization switch; si2~complementary transfer switch; va_m+~ positive input terminal;
Va_in_〜負輸入端; va。+〜正輸出端; 0955-A21636TWF(N2);glonous_tien 22 200822531 . Vai〜負輸出端;Va_in_~ negative input; va. +~ positive output; 0955-A21636TWF(N2);glonous_tien 22 200822531 . Vai~ negative output;
Vin+〜輸入信號正端; Vin_〜輸入信號負端; Vmt+〜第四節點;Vin+~ input signal positive terminal; Vin_~ input signal negative terminal; Vmt+~ fourth node;
Vlllt_〜第三節點; V〇ut +〜輸出信號正端; V〇ut_〜輸出信號負端; vx+〜第一節點; ⑩ vx_〜第二節點。 0955-A21636TWF(N2);glonous_t!enVlllt_~ third node; V〇ut+~ output signal positive terminal; V〇ut_~ output signal negative terminal; vx+~ first node; 10 vx_~ second node. 0955-A21636TWF(N2);glonous_t!en
Claims (1)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW95140562A TWI330931B (en) | 2006-11-02 | 2006-11-02 | Very-high-speed low-power low-hold pedestal fully differential cmos sample-and-hold circuit |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW95140562A TWI330931B (en) | 2006-11-02 | 2006-11-02 | Very-high-speed low-power low-hold pedestal fully differential cmos sample-and-hold circuit |
Publications (2)
Publication Number | Publication Date |
---|---|
TW200822531A true TW200822531A (en) | 2008-05-16 |
TWI330931B TWI330931B (en) | 2010-09-21 |
Family
ID=44770800
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW95140562A TWI330931B (en) | 2006-11-02 | 2006-11-02 | Very-high-speed low-power low-hold pedestal fully differential cmos sample-and-hold circuit |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI330931B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113328746A (en) * | 2020-02-28 | 2021-08-31 | 瑞昱半导体股份有限公司 | Sample-and-hold amplifying circuit |
TWI748352B (en) * | 2020-02-25 | 2021-12-01 | 瑞昱半導體股份有限公司 | Sample and hold amplifier circuit |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI473102B (en) * | 2011-09-22 | 2015-02-11 | Univ Nat Taipei Technology | High speed, high linear BiCMOS sample and hold circuit |
EP3716487A1 (en) | 2019-03-27 | 2020-09-30 | Nxp B.V. | A capacitive sampling circuit |
-
2006
- 2006-11-02 TW TW95140562A patent/TWI330931B/en not_active IP Right Cessation
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI748352B (en) * | 2020-02-25 | 2021-12-01 | 瑞昱半導體股份有限公司 | Sample and hold amplifier circuit |
US11581858B2 (en) | 2020-02-25 | 2023-02-14 | Realtek Semiconductor Corporation | Sample and hold amplifier circuit |
CN113328746A (en) * | 2020-02-28 | 2021-08-31 | 瑞昱半导体股份有限公司 | Sample-and-hold amplifying circuit |
CN113328746B (en) * | 2020-02-28 | 2024-05-14 | 瑞昱半导体股份有限公司 | Sample-hold amplifying circuit |
Also Published As
Publication number | Publication date |
---|---|
TWI330931B (en) | 2010-09-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ramirez-Angulo et al. | Very low-voltage analog signal processing based on quasi-floating gate transistors | |
Khateb | Bulk-driven floating-gate and bulk-driven quasi-floating-gate techniques for low-voltage low-power analog circuits design | |
US20090185406A1 (en) | Switched-Capacitor Circuit Having Two Feedback Capacitors | |
JP3824989B2 (en) | Integrated analog multiplexer | |
TW200822531A (en) | Very-high-speed low-power low-hold pedestal fully differential CMOS sample-and-hold circuit | |
US6608503B2 (en) | Hybrid comparator and method | |
Lee et al. | A 1.5-V 50-MHz pseudodifferential CMOS sample-and-hold circuit with low hold pedestal | |
CN106208982B (en) | A kind of trsanscondutance amplifier imbalance elimination structure based on output electric current storage | |
Wu et al. | A CMOS transistor-only 8-b 4.5-Ms/s pipelined analog-to-digital converter using fully-differential current-mode circuit techniques | |
JPH02131005A (en) | Control circuit for operational amplifier | |
Bult | The effect of technology scaling on power dissipation in analog circuits | |
Zhang et al. | An input buffer for 4 GS/s 14-b time-interleaved ADC | |
Lee et al. | A very-high-speed low-power low-voltage fully differential CMOS sample-and-hold circuit with low hold pedestal | |
Lee et al. | A 330MHz 11 bit 26.4 mW CMOS low-hold-pedestal fully differential track-and-hold circuit | |
Lee et al. | A 250MHz 11Bit 20mW CMOS low-hold-pedestal fully differential track-and-hold circuit | |
Bernal et al. | Performance enhancement in the design of amplifier and amplifier-less circuits in modern CMOS technologies | |
Carvajal et al. | New low-power low-voltage differential class-AB OTA for SC circuits | |
Ramirez-Angulo et al. | Low-voltage closed-loop amplifier circuits based on quasi-floating gate transistors | |
Atzeni et al. | An energy-efficient low-noise complementary parametric amplifier achieving 0.89 NEF | |
Lee et al. | A 200 MHz 4.8 mW 3 V fully differential CMOS sample-and-hold circuit with low hold pedestal | |
Lee et al. | Design techniques for low-voltage fully differential CMOS switched-capacitor amplifiers | |
Lee et al. | A 250MHz 11bit 20mW low-hold-pedestal CMOS fully differential track-and-hold circuit | |
Lee et al. | A 330áMHz 11ábit 26.4 ámW CMOS Low-Hold-Pedestal Fully Differential Sample-and-Hold Circuit | |
Lee et al. | A 250 MHz 11 bit 22 mW CMOS low-hold-pedestal fully differential sample-and-hold circuit | |
Campos et al. | New 12-bit source-follower track-and-hold circuit suitable for high-speed applications |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
MM4A | Annulment or lapse of patent due to non-payment of fees |