[問題] 差動放大器之增益計算

看板Electronics作者xx5236294roy (roy)時間8年前 (2015/09/27 18:54)推噓4(4推 0噓 5→)

留言9則, 5人參與討論串1/1

大家好，我最近在計算一題差動放大器的增益時發現用pspice及手算的結果有所差異不知是小弟在手算部份疏忽了什麼或是觀念錯誤還懇請各位大大指教 m(_ _)m [題目] 圖中為一差動放大器，V_DD = 3.3V、V_CM = V_DD/2 = 1.65V Find Vo / Vid with its DC gain and pole locations. + - ( Vid = Vin - Vin ) [圖] https://drive.google.com/open?id=0ByQPX3nLOyxfb3hWa0ZNdlJjdDA [其餘參數] (W/L)M1,M2 = 25um / 0.5um (W/L)M3,M4 = 1.25um / 0.5um (W/L)M5 = 10um / 0.5um (W/L)M6 = 4um / 0.5um NMOS PMOS Units LEVEL 1 1 TOX 9.50e-09 9.50e-09 m UO 460 115 cm^2/Vs LAMBDA 0.1 0.2 V^-1 GAMMA 0.5 0.45 V^(1/2) VTO 0.7 0.75 V PHI 0.8 0.75 V LD 8.00e-08 9.00e-08 m JS 1.00e-08 5.00e-09 A/m^2 CJ 5.70e-04 9.30e-04 F/m^2 MJ 0.5 0.5 CJSW 1.20e-10 1.70e-10 MJSW 0.4 0.35 PB 0.9 0.9 V CGBO 3.80e-10 3.80e-10 F/m CGDO 4.00e-10 3.50e-10 F/m CGSO 4.00e-10 3.50e-10 F/m 我的算法如下 I1 = 1/2 * uC * (W/L) * (V_S6 - V_G6 - |Vtp|)^2 = 1/2 * uC * (4/0.5) * (3.3 - V_G - 0.8)^2 --(1) 其中 uC = UO(cm^2/Vs) * (εox/ TOX) = (115*10^-4) * (3.45*10^-11 / 9.50*10^-9) = 42u -- (2) 又V_G6 = 100k * I1 -- (3) 將(2)(3)代入(1)得 I1 = 1/2 * 42u * (4/0.5) * (3.3 - 100k * I1 - 0.8)^2 => I1 = 21.42uA ∵V_G6 = V_G5, V_S6 = V_S5 ∴I2 = [(W/L)M5 / (W/L)M6 ] * I1 = 10/4 * I1 = 53.55uA g_m2 = sqrt( 2 * (42u * 25/0.5) * (53.55/2) ) = 335uA/V r_o2 = 1 / ( λ* (I1/2) ) = 1 / ( 0.2 * 26.775u) = 187k r_o4 = 1 / ( λ* (I1/2) ) = 1 / ( 0.1 * 26.775u) = 373k ∴ Vo / Vid = g_m2 (r_o2 // r_o4) = 335u * (187k // 373k) = 335u * (125k) = 41.875 = 32.44 (in dB) 但是我用pspice摹擬出來的增益約為37 dB [pspice之schematic] https://drive.google.com/open?id=0ByQPX3nLOyxfRW0xVDFVaUllb28 [pspice摹擬結果] https://drive.google.com/open?id=0ByQPX3nLOyxfbS10VjlkdjBjaVk 圖為 DB(V(Vout) / (V(Vin+) - V(Vin-))) by AC Sweep/Noise 在摹擬的時候AC Sweep Type我選擇Logarithmic Start Frequency: 1 End Frequency: 10e12 Points/Decade : 10000 以下為我pspice摹擬的檔案 (跑摹擬前需Edit Simulation Profile) https://drive.google.com/open?id=0ByQPX3nLOyxfS3VGRmU5V0NqeFk 謝謝大家了 m(_ _)m -- ※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 118.160.23.170 ※ 文章網址: https://www.ptt.cc/bbs/Electronics/M.1443351256.A.A22.html

推

W28327688

09/27 20:24, , 1^F

09/27 20:24, 1^F

→

W28327688

09/27 20:27, , 2^F

09/27 20:27, 2^F

Thx! 我驗證一下 Vg = Vd = 100k * 21.42u = 2.142 V Vsg = 3.3 - 2.142 = 1.158 V > |Vth| = 0.8 (on) Vsd = 1.158 > Vsg - |Vth| = 0.358 (sat) 所以好像應該有平方>< channel length modulation等會我算算先謝謝了！

→

xx5236294roy

09/27 20:41, , 3^F

09/27 20:41, 3^F

→

xx5236294roy

09/27 20:41, , 4^F

09/27 20:41, 4^F

→

xx5236294roy

09/27 20:41, , 5^F

09/27 20:41, 5^F