Bipolar Model Standardization Mextram 53.2 W.J. Kloosterman J.C.J. Paasschens D.B.M. Klaassen Laboratories, Eindhoven c Electronics N.V. 1999
Outline (1) Introduction Extended Mextram circuit model Results process A Mextram 53.2 Current and charge at onset of quasi-saturation Mextram 54. process C process D process E Remarks
Introduction (2) bipolar vertical NPN/PNP compact transistor model discrete and integrated circuit model analog/digital, RF/power applications low/high voltage, low/high frequency Survey of modelled effects modelling of I c including high injection, bias dependent Early effect modelling of I b including non ideal base current charge storage effects explicit modelling inactive regions (parasitic PNP) extensive modelling of the collector region including quasi-saturation, hot carrier effect, avalanche multiplication temperature effect
Introduction (3) noise modelling geometric scaling rules are not part of the model (in contrast with MOS compact models) public domain since 1993 http://www.semiconductors.philips.com/ Models/ documentation source code new release in june 2
Extended Equivalent Circuit Mextram (4) Substrate R sub Rcc Collector Isub Repi R bc R bv Base I n R e Emitter substrate resistance R sub (depending on circuitry) self-heating is a feature of the in-house circuit simulator T dut = TEMP + R th (I c V ce + I b V be )
Results: Process A (5) process A: Single Poly BiCMOS process emitter size:.6 5.4µm double base contact, R p = 1k maximum cut off frequency f T :1GHzat5VoltV ce (6 GHz at 5 mv V ce )
Process A: Cap Temp=25, 125 Mextram 53.2 (6) 12 Cbe 12 Cbc 5 Ccs 1 1 4 Cbe [ff] 8 6 Cbc [ff] 8 6 Ccs [ff] 3 2 4 4 1 2 4 3 2 1 1 Vbe[V] 2 13 11 9 7 5 3 1 1 Vbc[V] 13 11 9 7 5 3 1 1 Vcs[V] 1 Cbe_ 1 Cbc_ 4 Ccs_ 9 9 38 Cbe_[fF] 8 7 Cbc_[fF] 8 7 Ccs_[fF] 36 34 6 6 32 5 25 5 75 1 125 15 Temp[C] 5 25 5 75 1 125 15 Temp[C] 3 25 5 75 1 125 15 Temp[C]
Process A: fgummel Vbc= Temp=25, 5, 75, 125, 15 Mextram 53.2 (7) 1 1 Ic Ib Is 1 2 Ic [A] 1 6 Ib [A] 1 6 1 7 Is [A] 1 6 1 7 1 7 1 8 1 8 1 8 1 9.2.4.6.8 1. 1.2 Vbe[V] 1 9 1 1.2.4.6.8 1. 1.2 Vbe[V] 1 9 1 1.2.4.6.8 1. 1.2 Vbe[V] Ic/Ib[ ] 2 175 15 125 1 75 5 Ie/Ib 25.2.4.6.8 1. 1.2 Vbe[V] dic/dvbe/ic [1/V] 5 4 3 2 1 (dic/dvbe)/ic.2.4.6.8 1. 1.2 Vbe[V] (dib/dvbe)/ib [1/V] 5 4 3 2 1 (dib/dvbe)/ib.2.4.6.8 1. 1.2 Vbe[V]
Process A: rgummel Vbe= Temp=25, 5, 75, 125, 15 Mextram 53.2 (8) 1 1 Ie Ib Is 1 2 Ie [A] 1 6 Ib [A] 1 6 1 7 Is [A] 1 6 1 7 1 7 1 8 1 8 1 8 1 9.2.4.6.8 1. 1.2 Vbc[V] 1 9 1 1.2.4.6.8 1. 1.2 Vbc[V] 1 9 1 1.2.4.6.8 1. 1.2 Vbc[V] Ie/Ib[ ] Ie/Ib 1..9.8.7.6.5.4.3.2.1..2.4.6.8 1. 1.2 Vbc[V] Ie/(Ib+Is) [ ] Ie/(Ib+Is) 1 9 8 7 6 5 4 3 2 1.2.4.6.8 1. 1.2 Vbc[V] (die/dvbc)/ie [1/V] 5 4 3 2 1 (die/dvbc)/ie.2.4.6.8 1. 1.2 Vbc[V]
Process A: foutput-vbe Vbe=.65 Temp=25, 5, 75, 125, 15 Mextram 53.2 (9) 1 1 Ic Ib Ic [A] 1 2 1 6 2 4 6 8 1 12 Vcb[V] Ib [A] 1 6 1 7 1 8 1 9 2 4 6 8 1 12 Vcb[V] 3 Ic/Ib Ic/(dIc/dVcb) Mult=(Ib_ Ib)/Ic 25 1 Ic/Ib[ ] 2 15 1 Vef[V] 75 5 Mult[ ] 5 25 2 4 6 8 1 12 Vcb[V] 2 4 6 8 1 12 Vcb[V] 2 4 6 8 1 12 Vcb[V]
Process A: foutput-vbc Vbc=.65 Temp=25, 5, 75, 125, 15 Mextram 53.2 (1) 1 1 Ie Ib Is 1 2 Ie [A] Ib [A] 1 6 Is [A] 1 6 1 7 1 8 1 7 1 8 1 6 1 2 3 4 Veb[V] 1 9 1 2 3 4 Veb[V] 1 9 1 2 3 4 Veb[V] 5 Ie/Ib 3 Ie/(Ib+Is) 6 Ie/(dIe/dVeb) 4 25 5 Ie/Ib[ ] 3 2 Ie/(Ib+Is) [ ] 2 15 1 Ver[V] 4 3 2 1 5 1 1 2 3 4 Veb[V] 1 2 3 4 Veb[V] 1 2 3 4 Veb[V]
Process A: foutput-ib Temp=25 Ib=1, 2, 3, 4 ua Mextram 53.2 (11) 3 Ic Is 1.1 Vbe 25 1.5 Ic [ma] 2 15 1 Is [A] 1 6 1 7 1 8 Vbe [V] 1..95.9 5 1 9.85 1 2 3 4 5 6 7 8 Vce[V] 1 1..5 1. 1.5 2. Vce[V].8 1 2 3 4 5 6 7 8 Vce[V] 15 Ic/Ib 1 1 dic/dvce 5 Ic/(dIc/dVce) 125 4 Ic/Ib[ ] 1 75 5 g_out [mho] 1 2 Vef [V] 3 2 25 1 1 2 3 4 5 6 7 8 Vce[V] 1 2 3 4 5 6 7 8 Vce[V] 1 2 3 4 5 6 7 8 Vce [V]
Process A: Temp=25 f=1 GHz, Vce=.2,.5,.8, 2, 5 V Mextram 53.2 (12) 25 Hfe=Ic/Ib 1 R_bb 2 8 15 6 Ic/Ib 1 Rb 4 5 2 1 2 1 1 1 2 1 1 1 1 1 ft 1 11 F_uni 8 1 9 ft[hz] 6 1 9 4 1 9 F_uni[Hz] 1 1 1 9 2 1 9 1 1 2 1 1 1 8 1 2 1 1
Process A: Temp=25 f=1ghz Vce=.2 2,.8, 5 +Mextram 53.2 (13) 1 R(Y11) 1 1 R(Y12) 1 R(Y21) 1 R(Y22) 1 1 1 2 1 1 1 1 R(Y11) [A/V] 1 2 R(Y12) [A/V] R(Y21) [A/V] 1 2 R(Y22) [A/V] 1 2 1 2 1 1 1 2 1 1 1 2 1 1 1 2 1 1 1 1 I(Y11) 1 1 I(Y12) 1 I(Y21) 1 1 I(Y22) I(Y11) [A/V] 1 2 I(Y12) [A/V] 1 2 I(Y21) [A/V] 1 1 1 2 I(Y22) [A/V] 1 2 1 2 1 1 1 2 1 1 1 2 1 1 1 2 1 1
Current and charge at onset of quasi-saturation: Mextram 54 (14) Base widening starts when internal b-c junction becomes forward biased; i.e V B2 C 2 = V dc. We can calculate the current I ck where this happens; I ck = V d c V B2 C 1 SCR Cv Vd c V B2 C 1 + I hc SCR Cv V dc V B2 C 1 + I hc R cv Now this current can be used to determine when Q bc and Q epi should start to increase. This then determines the position of the top of f T. possible solution: Q bc + Q epi = τ xi x i W epi = 1 ( xi W epi ) 2 I c 1 + a xi ln 1 { [ ]} 1 + exp Ic /I ck 1 a xi Xi / Wepi.9.8.7.6.5.4.3.2.1 Mextram 53.2 Alternative Vcb = 1, 3, 5.5.1.15.2 Ic
Process A: Temp=25 f=1 GHz, Vce=.2,.5,.8, 2, 5 V Mextram 54. (15) 25 Hfe=Ic/Ib 1 R_bb 2 8 15 6 Ic/Ib 1 Rb 4 5 2 1 2 1 1 1 2 1 1 1 1 1 ft 1 11 F_uni 8 1 9 ft[hz] 6 1 9 4 1 9 F_uni[Hz] 1 1 1 9 2 1 9 1 1 2 1 1 1 8 1 2 1 1
Process A: Vce=.8 f=1 GHz, Temp=25 2, 5, 75, 125, 15 Mextram 54. (16) 25 Hfe=Ic/Ib 1 R_bb 2 8 15 6 Ic/Ib 1 Rb 4 5 2 1 2 1 1 1 2 1 1 1 1 1 ft 1 11 F_uni 8 1 9 ft[hz] 6 1 9 4 1 9 F_uni[Hz] 1 1 1 9 2 1 9 1 1 2 1 1 1 8 1 2 1 1
Results: Process C (17) process C: NPN on SOI, no parasitic PNP zero substrate current emitter size: 1. 5.µm double base contact, R p = 12k maximum cut off frequency f T :1GHzat6VoltV ce large amount of self-heating: R th = 25 o C/W
Process C: fgummel Vbc= Temp=27, 75, 125 Mextram 53.2 (18) 1 1 Ic Ib 1 2 Ic [A] 1 6 Ib [A] 1 6 1 7 1 7 1 8 1 8 1 9.2.4.6.8 1. 1.2 Vbe[V] 1 9 1 1.2.4.6.8 1. 1.2 Vbe[V] Ic/Ib[ ] 2 175 15 125 1 75 5 Ie/Ib 25.2.4.6.8 1. 1.2 Vbe[V] (dic/dvbe)/ic [1/V] 5 4 3 2 1 (dic/dvbe)/ic.2.4.6.8 1. 1.2 Vbe[V] (dib/dvbe)/ib [1/V] 5 4 3 2 1 (dib/dvbe)/ib.2.4.6.8 1. 1.2 Vbe[V]
Process C: foutput-ib Temp=27 Ib=1, 3, 5, 7, 9 ua Mextram 53.2 (19) 8 Ic 1.1 Vbe 6 1.5 1. Ic [ma] 4 2 Vbe [V].95.9.85 2 4 6 8 1 Vce[V].8 2 4 6 8 1 Vce[V] Ic/Ib[ ] 2 175 15 125 1 75 5 Ic/Ib 25 2 4 6 8 1 Vce[V] g_out [A/V] dic/dvce 2 4 6 8 1 Vce[V] Vef [V] 1 8 6 4 2 Ic/(dIc/dVce) 2 4 6 8 1 Vce[V]
Process C: Temp=27 f=993mhz Vce=.5 2, 1., 2., 6. Volt Mextram 53.2 (2) 5 R_bb 1 1 1 ft 4 8 1 9 R_bb[Ohm] 3 2 ft[hz] 6 1 9 4 1 9 1 2 1 9 1 2 1 1 2 Ic [A]
Results: Process D (21) process D: NPN of a double poly BiCMOS process with selective implanted collector (SIC) emitter size:.55 3.45µm single base contact, R p = 9k f T :25GHzat1VoltV ce F max : 25 GHz at.9 Volt V ce
Process D: fgummel+rgummel Temp=-5, 25, 62.5, 137.5, 2 Mextram 53.2 (22) Ic [A] 1 1 1 2 1 6 1 7 1 8 Ic: Vbc=.36 Volt 1 9 1 1.3..3.6.9 Vce[V] Ib [A] 1 6 1 7 1 8 1 9 1 1 Ib: Vbc=.36 Volt 1 11 1 12.3..3.6.9 Vce[V] Is [A] 1 1 1 2 1 6 1 7 1 8 Is: Vbc=.36 Volt 1 9 1 1.3..3.6.9 Vce[V] Ie: Vbe=.36 V Ib: Vbe=.36 V Is: Vbe=.36 V Ie [A] 1 6 1 7 Ib [A] 1 6 1 7 Is [A] 1 6 1 7 1 8 1 8 1 8 1 9 1 9 1 9 1 1 1 11.6.3..3.6.9 Vec[V] 1 1 1 11.6.3..3.6.9 Vec[V] 1 1 1 11.6.3..3.6.9 Vec[V]
Process D: Temp=25, freq=1.8ghz, Vce=.2,.4,.9, 1.5 Volt Mextram 53.2 (23) 1 1 R(Y11) 1 1 R(Y12) 1 1 R(Y21) 1 1 R(Y22) 1 2 1 2 1 2 1 2 R(Y11) [A/V] R(Y12) [A/V] R(Y21) [A/V] R(Y22) [A/V] 1 7 1 6 1 7 1 6 1 7 1 6 1 7 1 6 1 1 I(Y11) 1 1 I(Y12) 1 1 I(Y21) 1 1 I(Y22) 1 2 1 2 1 2 1 2 I(Y11) [A/V] I(Y12) [A/V] I(Y21) [A/V] I(Y22) [A/V] 1 7 1 6 1 7 1 6 1 7 1 6 1 7 1 6
Process D: Temp=25, freq=1.8ghz, Vce=.2,.4,.9, 1.5 Volt Mextram 53.2 (24) 14 hfe 5 R_bb 12 4 Ic/Ib [ ] 1 8 6 R_bb [Ohm] 3 2 4 2 1 1 6 1 6 3 ft 1 11 F_uni 25 ft [GHz] 2 15 1 F_uni [Hz] 1 1 1 9 5 1 6 1 8 1 6
Results: Process E (25) process E: NPN of a BiCMOS process emitter size:.6 4.8µm Sheet resestance base, R p 1k
Process E: Mextram 53.2 (26) 15 Ic/Ib: Vbc=,1, 2 Volt 1. Ie/Ib: Vbe=. V 5 Ie/(Ib+Is): Vbe=. V 125.8 4 Ic/Ib [ ] 1 75 5 Ie/Ib [ ].6.4 Ie/(Ib+Is) [ ] 3 2 25.2 1 1 1 1 9 1 8 1 7 1 6 Ic [A]. 1 1 1 9 1 8 1 7 1 6 Ie [A] 1 1 1 9 1 8 1 7 1 6 Ie [A] 4 Ic/(dIc/dVce): Ib=4,12,2 ua 5 Ie/(dIe/dVec): Ib=4,12,2 ua 2 ft: Vbc=, 1, 2, 4 Volt 3 4 15 Vef [V] 2 Ver [V] 3 2 ft [GHz] 1 1 1 5 1 2 3 4 5 Vce [V]..4.8 1.2 1.6 2. Vec [V] 1 6 1 2 1 1 Ic [A]
Process E: Vcb=.V, Vbe=.7,.8,.9, 1. Volt Temp=21.4 Mextram 53.2 (27) 1 1 R(Y11) R(Y12) 1 1 R(Y21) R(Y22) 1 2 1 2 R(Y11) [A/V] R(Y12) [A/V] 1 6 R(Y21) [A/V] R(Y22) [A/V] 1 7 1 6 1 7 1 8 1 9 1 1 Freq[Hz] 1 8 1 7 1 8 1 9 1 1 Freq [Hz] 1 6 1 7 1 8 1 9 1 1 Freq [Hz] 1 6 1 7 1 8 1 9 1 1 Freq [Hz] 1 1 I(Y11) 1 1 I(Y12) 1 1 I(Y21) 1 1 I(Y22) 1 2 1 2 1 2 1 2 I(Y11) [A/V] I(Y12) [A/V] I(Y21) [A/V] I(Y22) [A/V] 1 6 1 7 1 8 1 9 1 1 Freq [Hz] 1 6 1 7 1 8 1 9 1 1 Freq [Hz] 1 6 1 7 1 8 1 9 1 1 Freq [Hz] 1 6 1 7 1 8 1 9 1 1 Freq [Hz]
Remarks (28) We have characterized 5 of the 6 transistors Mextram 53.2 parameter set + substrate resistance + self-heating parameters simulated data for all measurements are provided Current data of process B (SiGe) can not be used for benchmarking DC and AC data are measured on different wafers in time. At least 2 parameter sets are needed Request to place simulated data + parameters of VIBIC, Hicum and Mextram on CMC web site in the same format and filenames as the measured data. Benchmarking of Mextram 54 will be done for the same data sets.