On Aug 18, 7:28=A0pm, John Larkin <jjlar...@highNOTlandTHIStechnologyPART.com> wrote:> On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: > > > > > > > > > > >Phil Hobbs a =E9crit : > >> dagmargoodb...@yahoo.com wrote: > >>> On Aug 15, 4:34 pm, Phil Hobbs > >>> <pcdhSpamMeSensel...@electrooptical.net> wrote: > > >>> [...] > > >>>> The NPN/PNP series-shunt pair (the one that folks with nothing bette=r to> >>>> do are are slagging each other off about at the moment) runs the dri=ver> >>>> transistor at constant I_C, and lets you use a much faster NPN to > >>>> stiffen the PNP at high frequency. =A0Try simulating it--it's really=good> >>>> medicine. > > >>>> (I spent most of July up to my ears in high performance laser noise > >>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pre=tty> >>>> well up on all this at the moment.) > >>> I like the pseudo-Darlington, I'm just leery because I had those > >>> scream wickedly once using fast transistors. =A0I don't think I ever > >>> fixed it. > > >>> It has a bunch of advantages here, including reducing the base curren=t> >>> i(b) error. > > >>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: > > >>> =A0 =A0 Vcc =A0 =A0 =A0 =A0 =A0 Vcc > >>> =A0 =A0 -+- =A0 =A0 =A0 =A0 =A0 -+- > >>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0 | > >>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0.-. R1 > >>> =A0 =A0 =A0| U1 =A0 =A0 =A0 =A0 | | 120 > >>> =A0 =A0 =A0| LM385 =A0 =A0 =A0'-' > >>> =A0 =A0 =A0| =A0-adj =A0 =A0 =A0 | > >>> =A0 =A0.---' =A0 =A0 =A0 =A0 =A0 | > >>> =A0 =A0 =A0A<-----R2-----+---------. > >>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0 =A0 | > >>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 |<' Q1 =A0 =A0 =A0| > >>> =A0 =A0 =A0+-----+-----| =A0BFT92 =A0 =A0| > >>> =A0 =A0 =A0| =A0 =A0 | =A0 =A0 |\ =A0 =A0 =A0 =A0 =A0| > >>> =A0R4 .-. =A0 --- C1 =A0 | =A0 =A0 =A0 |/ =A0Q2 > >>> =A01K | | =A0 --- =A0 =A0 =A0+--f.b.-| =A0BFP640 > >>> =A0 =A0 '-' =A0 =A0| =A0 =A0 =A0 | =A0 =A0 =A0 |>. > >>> =A0 =A0 =A0| =A0 =A0=3D=3D=3D =A0 =A0 .-. =A0 =A0 =A0 =A0| > >>> =A0 =A0 =3D=3D=3D =A0 =A0 =A0 =A0 =A0 | | R3 =A0 =A0 | > >>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 '-' 600 =A0 =A0| > >>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 =A0 =A0 =A0 | > >>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0'---------+----> > > >>> R2's a pathetic provision for isolating U1 from the nS stuff. =A0The > >>> whole thing needs stabilizing against UHF oscillation, these are just > >>> the bones. > > >>> -- > >>> Cheers, > >>> James Arthur > > >> It wouldn't need much more than that, I don't think. =A0Series-shunt p=airs> >> have poor stability when the transistors are about the same speed, due > >> to having two lags in the loop, but the BFP640 is ~10x faster than the > >> BFT92, which helps a lot. =A0(Reducing the BFT92's collector current m=akes> >> the difference wider and helps with stability.) > > >> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet > >> charge-sensitive preamp (noise floor ~10 electrons per inverse > >> bandwidth), and all it needed was an 0402 bead in the base--no worries > >> at all. > > >> Cheers > > >> Phil Hobbs > > >I had a try at John's pb... > > >Here it is, using a current source bootstrapped bootstrapped current > >source (yes I really meant to say that :-). > > >I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. > >Some temperature dependent biasing is also provided but I relied on the > >8V supply stability over temp which may be wrong and you may have to > >adapt a bit. > > >Some simulations including parasitics give pretty interesting results. > >Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have > >any to measure it) > > >Despite this this circuit shows excellent results (in simulation, so > >apply the usual grain of salt). > > >Still: > >* better than 0.1% linearity with this low VAF figure which can be > >further improved > >* start up transients are under 1ns and the integrated error is > >negligible... > >* 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I > >relied on your 8V supply stability, so you may have to adapt a bit) > >* low output impedance, so no need for buffering if you just drive one > >input (modeled here as the 2pF capacitance) and can stand the temp > >dependent offset at start (or you can still add a NPN buffer to > >compensate for VBE) > > >Here's a LTspice netlist. It's just for the drawing as I don't use it > >for sim and have almost no BJT models and even less with parasitics... > > >If you want to see the results I'll make a pdf. > > >******************************************** > > >Version 4 > >SHEET 1 948 1004 > >WIRE 16 32 -128 32 > >WIRE 96 32 16 32 > >WIRE 208 32 96 32 > >WIRE 416 32 208 32 > >WIRE 528 32 416 32 > >WIRE -128 48 -128 32 > >WIRE 16 48 16 32 > >WIRE 96 48 96 32 > >WIRE 208 48 208 32 > >WIRE -128 144 -128 128 > >WIRE -96 144 -128 144 > >WIRE 16 144 16 128 > >WIRE 16 144 -16 144 > >WIRE 16 176 16 144 > >WIRE 96 176 96 112 > >WIRE 96 176 16 176 > >WIRE 144 176 96 176 > >WIRE -128 224 -128 144 > >WIRE -48 224 -128 224 > >WIRE -128 256 -128 224 > >WIRE 880 304 112 304 > >WIRE 944 304 880 304 > >WIRE 16 336 16 272 > >WIRE 208 384 208 224 > >WIRE 240 384 208 384 > >WIRE 368 384 320 384 > >WIRE 784 384 368 384 > >WIRE 880 384 784 384 > >WIRE 944 384 880 384 > >WIRE 784 400 784 384 > >WIRE 368 416 368 384 > >WIRE 784 480 784 464 > >WIRE 208 496 208 384 > >WIRE 368 528 368 496 > >WIRE 640 528 368 528 > >WIRE 704 528 640 528 > >WIRE 368 560 368 528 > >WIRE 640 576 640 528 > >WIRE 704 576 704 528 > >WIRE 112 608 112 304 > >WIRE 112 608 -128 608 > >WIRE 208 608 208 576 > >WIRE 208 608 112 608 > >WIRE 304 608 208 608 > >WIRE -128 640 -128 608 > >WIRE 208 656 208 608 > >WIRE -176 704 -272 704 > >WIRE 368 704 368 656 > >WIRE 480 752 432 752 > >WIRE 528 752 480 752 > >WIRE 640 752 640 640 > >WIRE 640 752 608 752 > >WIRE 704 752 704 656 > >WIRE 704 752 640 752 > >WIRE -128 768 -128 736 > >WIRE 208 768 208 720 > >WIRE 480 784 480 752 > >WIRE 704 800 704 752 > >WIRE 480 880 480 848 > >WIRE 224 960 128 960 > >WIRE 368 960 368 800 > >WIRE 368 960 224 960 > >WIRE 704 960 704 880 > >WIRE 704 960 368 960 > >WIRE 848 960 704 960 > >FLAG 16 336 0 > >FLAG -128 336 0 > >FLAG 208 768 0 > >FLAG 784 480 0 > >FLAG 416 32 +8V_rail > >FLAG 224 960 -5V_rail > >FLAG 480 880 0 > >FLAG -128 768 0 > >FLAG 880 304 Out > >FLAG 880 384 LZ_Out > >SYMBOL pnp 144 224 M180 > >SYMATTR InstName Q1 > >SYMATTR Value BFT92 > >SYMBOL res 192 32 R0 > >SYMATTR InstName R1 > >SYMATTR Value 68R > >SYMBOL res 0 32 R0 > >SYMATTR InstName R2 > >SYMATTR Value 1.8K > >SYMBOL res 0 128 R90 > >WINDOW 0 0 56 VBottom 2 > >WINDOW 3 32 56 VTop 2 > >SYMATTR InstName R3 > >SYMATTR Value 1K > >SYMBOL res -112 144 R180 > >WINDOW 0 36 76 Left 2 > >WINDOW 3 36 40 Left 2 > >SYMATTR InstName R4 > >SYMATTR Value 1K > >SYMBOL res -112 352 R180 > >WINDOW 0 36 76 Left 2 > >WINDOW 3 36 40 Left 2 > >SYMATTR InstName R5 > >SYMATTR Value 1.5K > >SYMBOL pnp -48 272 M180 > >SYMATTR InstName Q2 > >SYMATTR Value MMBT3906 > >SYMBOL cap 80 48 R0 > >SYMATTR InstName C1 > >SYMATTR Value 100n > >SYMBOL cap 192 656 R0 > >SYMATTR InstName C2 > >SYMATTR Value 10p > >SYMBOL pnp 304 656 M180 > >WINDOW 0 -19 136 Left 2 > >WINDOW 3 -20 111 Left 2 > >SYMATTR InstName Q3 > >SYMATTR Value BFT92 > >SYMBOL pnp 432 800 R180 > >WINDOW 0 117 63 Left 2 > >WINDOW 3 73 37 Left 2 > >SYMATTR InstName Q4 > >SYMATTR Value BFT92 > >SYMBOL res 224 480 M0 > >SYMATTR InstName R6 > >SYMATTR Value 82R > >SYMBOL res 336 368 R90 > >WINDOW 0 0 56 VBottom 2 > >WINDOW 3 32 56 VTop 2 > >SYMATTR InstName R7 > >SYMATTR Value 10R > >SYMBOL ind 352 400 R0 > >SYMATTR InstName L1 > >SYMATTR Value 6.8nH > >SYMBOL cap 768 400 R0 > >SYMATTR InstName C3 > >SYMATTR Value 2p > >SYMBOL res 624 736 R90 > >WINDOW 0 0 56 VBottom 2 > >WINDOW 3 32 56 VTop 2 > >SYMATTR InstName R8 > >SYMATTR Value 100R > >SYMBOL res 688 560 R0 > >SYMATTR InstName R9 > >SYMATTR Value 10K > >SYMBOL res 688 784 R0 > >SYMATTR InstName R10 > >SYMATTR Value 6.2K > >SYMBOL cap 656 576 M0 > >SYMATTR InstName C4 > >SYMATTR Value 10n > >SYMBOL cap 464 784 R0 > >SYMATTR InstName C5 > >SYMATTR Value 0.2p > >SYMBOL njf -176 640 R0 > >SYMATTR InstName J1 > >SYMATTR Value NE3509 > > >******************************************** > > Hi, Fred, > > I tried it but it won't run without the BFT92 model. Do you have that?NXP does * Filename: BFT92_SPICE.PRM * BFT92 SPICE MODEL * PHILIPS SEMICONDUCTORS * Date : September 1995 * * PACKAGE : SOT23 DIE MODEL : BFT92 * 1: COLLECTOR; 2: BASE; 3: EMITTER; .SUBCKT BFT92 1 2 3 Q1 6 5 7 7 BFT92 * SOT23 parasitic model Lb 4 5 .4n Le 7 8 .83n L1 2 4 .35n L2 1 6 .17n L3 3 8 .35n Ccb 4 6 71f Cbe 4 8 2f Cce 6 8 71f * PHILIPS SEMICONDUCTORS Version: 2.0 * Filename: BFT92.PRM Date: July 1992 * .MODEL BFT92 PNP + IS =3D 4.37563E-016 + BF =3D 3.35815E+001 + NF =3D 1.00972E+000 + VAF =3D 2.33946E+001 + IKF =3D 9.95381E-002 + ISE =3D 8.70539E-014 + NE =3D 1.94395E+000 + BR =3D 4.94721E+000 + NR =3D 1.00254E+000 + VAR =3D 3.90385E+000 + IKR =3D 5.28157E-003 + ISC =3D 3.58864E-014 + NC =3D 1.39333E+000 + RB =3D 5.00000E+000 + IRB =3D 1.00000E-006 + RBM =3D 5.00000E+000 + RE =3D 1.00000E+000 + RC =3D 1.00000E+001 + EG =3D 1.11000E+000 + XTI =3D 3.00000E+000 + CJE =3D 7.46659E-013 + VJE =3D 6.00000E-001 + MJE =3D 3.56829E-001 + TF =3D 1.74921E-011 + XTF =3D 1.35455E+000 + VTF =3D 1.55654E-001 + ITF =3D 1.00000E-003 + PTF =3D 4.50000E+001 + CJC =3D 9.37103E-013 + VJC =3D 3.96455E-001 + MJC =3D 1.99949E-001 .ENDS Persuading it to fit into LTSpice is tedious, but it can be done. -- Bill Sloman, Nijmegen
fast ramp follies
Started by ●August 14, 2012
Reply by ●August 18, 20122012-08-18
Reply by ●August 18, 20122012-08-18
On Sat, 18 Aug 2012 20:03:59 +0200, Fred Bartoli <" "> wrote:>John Larkin a �crit : >> On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: >> >>> Phil Hobbs a �crit : >>>> dagmargoodboat@yahoo.com wrote: >>>>> On Aug 15, 4:34 pm, Phil Hobbs >>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: >>>>> >>>>> [...] >>>>> >>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing better to >>>>>> do are are slagging each other off about at the moment) runs the driver >>>>>> transistor at constant I_C, and lets you use a much faster NPN to >>>>>> stiffen the PNP at high frequency. Try simulating it--it's really good >>>>>> medicine. >>>>>> >>>>>> (I spent most of July up to my ears in high performance laser noise >>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pretty >>>>>> well up on all this at the moment.) >>>>> I like the pseudo-Darlington, I'm just leery because I had those >>>>> scream wickedly once using fast transistors. I don't think I ever >>>>> fixed it. >>>>> >>>>> It has a bunch of advantages here, including reducing the base current >>>>> i(b) error. >>>>> >>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: >>>>> >>>>> Vcc Vcc >>>>> -+- -+- >>>>> | | >>>>> | .-. R1 >>>>> | U1 | | 120 >>>>> | LM385 '-' >>>>> | -adj | >>>>> .---' | >>>>> A<-----R2-----+---------. >>>>> | | | >>>>> | |<' Q1 | >>>>> +-----+-----| BFT92 | >>>>> | | |\ | >>>>> R4 .-. --- C1 | |/ Q2 >>>>> 1K | | --- +--f.b.-| BFP640 >>>>> '-' | | |>. >>>>> | === .-. | >>>>> === | | R3 | >>>>> '-' 600 | >>>>> | | >>>>> '---------+----> >>>>> >>>>> R2's a pathetic provision for isolating U1 from the nS stuff. The >>>>> whole thing needs stabilizing against UHF oscillation, these are just >>>>> the bones. >>>>> >>>>> -- >>>>> Cheers, >>>>> James Arthur >>>> It wouldn't need much more than that, I don't think. Series-shunt pairs >>>> have poor stability when the transistors are about the same speed, due >>>> to having two lags in the loop, but the BFP640 is ~10x faster than the >>>> BFT92, which helps a lot. (Reducing the BFT92's collector current makes >>>> the difference wider and helps with stability.) >>>> >>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet >>>> charge-sensitive preamp (noise floor ~10 electrons per inverse >>>> bandwidth), and all it needed was an 0402 bead in the base--no worries >>>> at all. >>>> >>>> Cheers >>>> >>>> Phil Hobbs >>>> >>> >>> I had a try at John's pb... >>> >>> Here it is, using a current source bootstrapped bootstrapped current >>> source (yes I really meant to say that :-). >>> >>> I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. >>> Some temperature dependent biasing is also provided but I relied on the >>> 8V supply stability over temp which may be wrong and you may have to >>> adapt a bit. >>> >>> Some simulations including parasitics give pretty interesting results. >>> Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have >>> any to measure it) >>> >>> Despite this this circuit shows excellent results (in simulation, so >>> apply the usual grain of salt). >>> >>> Still: >>> * better than 0.1% linearity with this low VAF figure which can be >>> further improved >>> * start up transients are under 1ns and the integrated error is >>> negligible... >>> * 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I >>> relied on your 8V supply stability, so you may have to adapt a bit) >>> * low output impedance, so no need for buffering if you just drive one >>> input (modeled here as the 2pF capacitance) and can stand the temp >>> dependent offset at start (or you can still add a NPN buffer to >>> compensate for VBE) >>> >>> >>> Here's a LTspice netlist. It's just for the drawing as I don't use it >>> for sim and have almost no BJT models and even less with parasitics... >>> >>> If you want to see the results I'll make a pdf. >>> >>> >>> ******************************************** >>> >>> Version 4 >>> SHEET 1 948 1004 >>> WIRE 16 32 -128 32 ><< ... >> >>> SYMBOL njf -176 640 R0 >>> SYMATTR InstName J1 >>> SYMATTR Value NE3509 >>> >>> >>> >>> ******************************************** >> >> >> Hi, Fred, >> >> I tried it but it won't run without the BFT92 model. Do you have that? >> > > >As I said I don't have them for LTspice, so it's why this netlist don't > run flat. But they are available from the Infineon datasheet, with all >the parasitics... (I don't know how to do build a new model for ltspice) > >You'd have to build a subcircuit for the SOT23 parasitics and include >the die model. Here is the one from the spice database I used. > > >.SUBCKT BFT92s 200 100 300 >** ^ ^ ^ >** C B E >** >*SOT23 >Q1 2 1 3 BFT92Xm >LBI 1 10 0.85NH >LEI 3 30 0.69NH >CCB 10 2 84FF >CCE 2 30 165FF >CBE 10 30 73FF >LBO 10 100 0.51NH >LCO 2 200 0.49NH >LEO 30 300 0.61NH > >.MODEL BFT92Xm PNP( >+ IS = 4.5354E-15 BF = 98.533 NF = 0.90551 >+ VAF = 10.983 IKF = 0.016123 ISE = 1.2196E-14 >+ NE = 1.1172 BR = 10.297 NR = 1.2703 >+ VAR = 47.577 IKR = 0.019729 ISC = 2.4709E-17 >+ NC = 1.206 RB = 7.9562 IRB = 0.00079584 >+ RBM = 1.5939 RE = 1.5119 RC = 0.66749 >+ CJE = 1.7785E-15 VJE = 0.79082 MJE = 0.32167 >+ TF = 3.2171E-11 XTF = 0.30227 VTF = 0.21451 >+ ITF = 1.3277E-05 PTF = 0 CJC = 9.2207E-13 >+ VJC = 1.2 MJC = 0.3 XCJC = 0.3 >+ TR = 2.0779E-09 CJS= 0 VJS = 0.75 >+ MJS = 0 XTB = 0 EG = 1.11 >+ XTI = 3 FC = 0.75167) > >.ENDS > > >Note that on my database model CBE is specified but is missing from the >datasheet that you can find here > >http://www.farnell.com/datasheets/62457.pdf (probably a datasheet error) > > >That 73fF value is coherent with what can be expected from a SOT23 >package so it probably is right. >You can try with and without it, but in this circuit it probably don't >matter much.Thanks, Fred! I should have thought to look in my PSpice libraries, found it in Philips_RF.lib... minor differences in parasitics... * SOT23 parasitic model Lb 4 5 .4n Le 7 8 .83n L1 2 4 .35n L2 1 6 .17n L3 3 8 .35n Ccb 4 6 71f Cbe 4 8 71f ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
Reply by ●August 18, 20122012-08-18
Bill Sloman a �crit :> On Aug 18, 7:28 pm, John Larkin > <jjlar...@highNOTlandTHIStechnologyPART.com> wrote: >> On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: >> >> >> >> >> >> >> >> >> >>> Phil Hobbs a �crit : >>>> dagmargoodb...@yahoo.com wrote: >>>>> On Aug 15, 4:34 pm, Phil Hobbs >>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: >>>>> [...] >>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing better to >>>>>> do are are slagging each other off about at the moment) runs the driver >>>>>> transistor at constant I_C, and lets you use a much faster NPN to >>>>>> stiffen the PNP at high frequency. Try simulating it--it's really good >>>>>> medicine. >>>>>> (I spent most of July up to my ears in high performance laser noise >>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pretty >>>>>> well up on all this at the moment.) >>>>> I like the pseudo-Darlington, I'm just leery because I had those >>>>> scream wickedly once using fast transistors. I don't think I ever >>>>> fixed it. >>>>> It has a bunch of advantages here, including reducing the base current >>>>> i(b) error. >>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: >>>>> Vcc Vcc >>>>> -+- -+- >>>>> | | >>>>> | .-. R1 >>>>> | U1 | | 120 >>>>> | LM385 '-' >>>>> | -adj | >>>>> .---' | >>>>> A<-----R2-----+---------. >>>>> | | | >>>>> | |<' Q1 | >>>>> +-----+-----| BFT92 | >>>>> | | |\ | >>>>> R4 .-. --- C1 | |/ Q2 >>>>> 1K | | --- +--f.b.-| BFP640 >>>>> '-' | | |>. >>>>> | === .-. | >>>>> === | | R3 | >>>>> '-' 600 | >>>>> | | >>>>> '---------+----> >>>>> R2's a pathetic provision for isolating U1 from the nS stuff. The >>>>> whole thing needs stabilizing against UHF oscillation, these are just >>>>> the bones. >>>>> -- >>>>> Cheers, >>>>> James Arthur >>>> It wouldn't need much more than that, I don't think. Series-shunt pairs >>>> have poor stability when the transistors are about the same speed, due >>>> to having two lags in the loop, but the BFP640 is ~10x faster than the >>>> BFT92, which helps a lot. (Reducing the BFT92's collector current makes >>>> the difference wider and helps with stability.) >>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet >>>> charge-sensitive preamp (noise floor ~10 electrons per inverse >>>> bandwidth), and all it needed was an 0402 bead in the base--no worries >>>> at all. >>>> Cheers >>>> Phil Hobbs >>> I had a try at John's pb... >>> Here it is, using a current source bootstrapped bootstrapped current >>> source (yes I really meant to say that :-). >>> I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. >>> Some temperature dependent biasing is also provided but I relied on the >>> 8V supply stability over temp which may be wrong and you may have to >>> adapt a bit. >>> Some simulations including parasitics give pretty interesting results. >>> Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have >>> any to measure it) >>> Despite this this circuit shows excellent results (in simulation, so >>> apply the usual grain of salt). >>> Still: >>> * better than 0.1% linearity with this low VAF figure which can be >>> further improved >>> * start up transients are under 1ns and the integrated error is >>> negligible... >>> * 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I >>> relied on your 8V supply stability, so you may have to adapt a bit) >>> * low output impedance, so no need for buffering if you just drive one >>> input (modeled here as the 2pF capacitance) and can stand the temp >>> dependent offset at start (or you can still add a NPN buffer to >>> compensate for VBE) >>> Here's a LTspice netlist. It's just for the drawing as I don't use it >>> for sim and have almost no BJT models and even less with parasitics... >>> If you want to see the results I'll make a pdf. >>> ******************************************** >>> Version 4 >>> SHEET 1 948 1004 >>> WIRE 16 32 -128 32 >>> WIRE 96 32 16 32.....>>> SYMBOL cap 464 784 R0 >>> SYMATTR InstName C5 >>> SYMATTR Value 0.2p >>> SYMBOL njf -176 640 R0 >>> SYMATTR InstName J1 >>> SYMATTR Value NE3509 >>> ******************************************** >> Hi, Fred, >> >> I tried it but it won't run without the BFT92 model. Do you have that? > > NXP does > > * Filename: BFT92_SPICE.PRM > * BFT92 SPICE MODEL > * PHILIPS SEMICONDUCTORS > * Date : September 1995 > * > * PACKAGE : SOT23 DIE MODEL : BFT92 > * 1: COLLECTOR; 2: BASE; 3: EMITTER; > .SUBCKT BFT92 1 2 3 > Q1 6 5 7 7 BFT92 > * SOT23 parasitic model > Lb 4 5 .4n > Le 7 8 .83n > L1 2 4 .35n > L2 1 6 .17n > L3 3 8 .35n > Ccb 4 6 71f > Cbe 4 8 2f > Cce 6 8 71f > * PHILIPS SEMICONDUCTORS > Version: 2.0 > * Filename: BFT92.PRM Date: > July 1992 > * > .MODEL BFT92 PNP > + IS = 4.37563E-016 > + BF = 3.35815E+001 > + NF = 1.00972E+000 > + VAF = 2.33946E+001 > + IKF = 9.95381E-002 > + ISE = 8.70539E-014 > + NE = 1.94395E+000 > + BR = 4.94721E+000 > + NR = 1.00254E+000 > + VAR = 3.90385E+000 > + IKR = 5.28157E-003 > + ISC = 3.58864E-014 > + NC = 1.39333E+000 > + RB = 5.00000E+000 > + IRB = 1.00000E-006 > + RBM = 5.00000E+000 > + RE = 1.00000E+000 > + RC = 1.00000E+001 > + EG = 1.11000E+000 > + XTI = 3.00000E+000 > + CJE = 7.46659E-013 > + VJE = 6.00000E-001 > + MJE = 3.56829E-001 > + TF = 1.74921E-011 > + XTF = 1.35455E+000 > + VTF = 1.55654E-001 > + ITF = 1.00000E-003 > + PTF = 4.50000E+001 > + CJC = 9.37103E-013 > + VJC = 3.96455E-001 > + MJC = 1.99949E-001 > .ENDS > > Persuading it to fit into LTSpice is tedious, but it can be done. > > -- > Bill Sloman, NijmegenFor you posting this I had a closer look and it happens that I also have the NXP model in my database, so a sanity check simulation was done. Apart from the slope difference coming from a huge difference in Q3 VBE (110 to 115mV) the transient responses are *very* similar, despite the different ft (5.8G for the Q3 NXP against 4.8G for the infineon model) which is a good sign :-) BTW, at more than 800mV VBE for the NXP model I tend to not trust it in that respect... -- Thanks, Fred.
Reply by ●August 18, 20122012-08-18
Jim Thompson a �crit :> On Sat, 18 Aug 2012 20:03:59 +0200, Fred Bartoli <" "> wrote: > >> John Larkin a �crit : >>> On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: >>> >>>> Phil Hobbs a �crit : >>>>> dagmargoodboat@yahoo.com wrote: >>>>>> On Aug 15, 4:34 pm, Phil Hobbs >>>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: >>>>>> >>>>>> [...] >>>>>> >>>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing better to >>>>>>> do are are slagging each other off about at the moment) runs the driver >>>>>>> transistor at constant I_C, and lets you use a much faster NPN to >>>>>>> stiffen the PNP at high frequency. Try simulating it--it's really good >>>>>>> medicine. >>>>>>> >>>>>>> (I spent most of July up to my ears in high performance laser noise >>>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pretty >>>>>>> well up on all this at the moment.) >>>>>> I like the pseudo-Darlington, I'm just leery because I had those >>>>>> scream wickedly once using fast transistors. I don't think I ever >>>>>> fixed it. >>>>>> >>>>>> It has a bunch of advantages here, including reducing the base current >>>>>> i(b) error. >>>>>> >>>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: >>>>>> >>>>>> Vcc Vcc >>>>>> -+- -+- >>>>>> | | >>>>>> | .-. R1 >>>>>> | U1 | | 120 >>>>>> | LM385 '-' >>>>>> | -adj | >>>>>> .---' | >>>>>> A<-----R2-----+---------. >>>>>> | | | >>>>>> | |<' Q1 | >>>>>> +-----+-----| BFT92 | >>>>>> | | |\ | >>>>>> R4 .-. --- C1 | |/ Q2 >>>>>> 1K | | --- +--f.b.-| BFP640 >>>>>> '-' | | |>. >>>>>> | === .-. | >>>>>> === | | R3 | >>>>>> '-' 600 | >>>>>> | | >>>>>> '---------+----> >>>>>> >>>>>> R2's a pathetic provision for isolating U1 from the nS stuff. The >>>>>> whole thing needs stabilizing against UHF oscillation, these are just >>>>>> the bones. >>>>>> >>>>>> -- >>>>>> Cheers, >>>>>> James Arthur >>>>> It wouldn't need much more than that, I don't think. Series-shunt pairs >>>>> have poor stability when the transistors are about the same speed, due >>>>> to having two lags in the loop, but the BFP640 is ~10x faster than the >>>>> BFT92, which helps a lot. (Reducing the BFT92's collector current makes >>>>> the difference wider and helps with stability.) >>>>> >>>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet >>>>> charge-sensitive preamp (noise floor ~10 electrons per inverse >>>>> bandwidth), and all it needed was an 0402 bead in the base--no worries >>>>> at all. >>>>> >>>>> Cheers >>>>> >>>>> Phil Hobbs >>>>> >>>> I had a try at John's pb... >>>> >>>> Here it is, using a current source bootstrapped bootstrapped current >>>> source (yes I really meant to say that :-). >>>> >>>> I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. >>>> Some temperature dependent biasing is also provided but I relied on the >>>> 8V supply stability over temp which may be wrong and you may have to >>>> adapt a bit. >>>> >>>> Some simulations including parasitics give pretty interesting results. >>>> Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have >>>> any to measure it) >>>> >>>> Despite this this circuit shows excellent results (in simulation, so >>>> apply the usual grain of salt). >>>> >>>> Still: >>>> * better than 0.1% linearity with this low VAF figure which can be >>>> further improved >>>> * start up transients are under 1ns and the integrated error is >>>> negligible... >>>> * 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I >>>> relied on your 8V supply stability, so you may have to adapt a bit) >>>> * low output impedance, so no need for buffering if you just drive one >>>> input (modeled here as the 2pF capacitance) and can stand the temp >>>> dependent offset at start (or you can still add a NPN buffer to >>>> compensate for VBE) >>>> >>>> >>>> Here's a LTspice netlist. It's just for the drawing as I don't use it >>>> for sim and have almost no BJT models and even less with parasitics... >>>> >>>> If you want to see the results I'll make a pdf. >>>> >>>> >>>> ******************************************** >>>> >>>> Version 4 >>>> SHEET 1 948 1004 >>>> WIRE 16 32 -128 32 >> << ... >> >>>> SYMBOL njf -176 640 R0 >>>> SYMATTR InstName J1 >>>> SYMATTR Value NE3509 >>>> >>>> >>>> >>>> ******************************************** >>> >>> Hi, Fred, >>> >>> I tried it but it won't run without the BFT92 model. Do you have that? >>> >> >> As I said I don't have them for LTspice, so it's why this netlist don't >> run flat. But they are available from the Infineon datasheet, with all >> the parasitics... (I don't know how to do build a new model for ltspice) >> >> You'd have to build a subcircuit for the SOT23 parasitics and include >> the die model. Here is the one from the spice database I used. >> >> >> .SUBCKT BFT92s 200 100 300 >> ** ^ ^ ^ >> ** C B E >> ** >> *SOT23 >> Q1 2 1 3 BFT92Xm >> LBI 1 10 0.85NH >> LEI 3 30 0.69NH >> CCB 10 2 84FF >> CCE 2 30 165FF >> CBE 10 30 73FF >> LBO 10 100 0.51NH >> LCO 2 200 0.49NH >> LEO 30 300 0.61NH >> >> .MODEL BFT92Xm PNP( >> + IS = 4.5354E-15 BF = 98.533 NF = 0.90551 >> + VAF = 10.983 IKF = 0.016123 ISE = 1.2196E-14 >> + NE = 1.1172 BR = 10.297 NR = 1.2703 >> + VAR = 47.577 IKR = 0.019729 ISC = 2.4709E-17 >> + NC = 1.206 RB = 7.9562 IRB = 0.00079584 >> + RBM = 1.5939 RE = 1.5119 RC = 0.66749 >> + CJE = 1.7785E-15 VJE = 0.79082 MJE = 0.32167 >> + TF = 3.2171E-11 XTF = 0.30227 VTF = 0.21451 >> + ITF = 1.3277E-05 PTF = 0 CJC = 9.2207E-13 >> + VJC = 1.2 MJC = 0.3 XCJC = 0.3 >> + TR = 2.0779E-09 CJS= 0 VJS = 0.75 >> + MJS = 0 XTB = 0 EG = 1.11 >> + XTI = 3 FC = 0.75167) >> >> .ENDS >> >> >> Note that on my database model CBE is specified but is missing from the >> datasheet that you can find here >> >> http://www.farnell.com/datasheets/62457.pdf (probably a datasheet error) >> >> >> That 73fF value is coherent with what can be expected from a SOT23 >> package so it probably is right. >> You can try with and without it, but in this circuit it probably don't >> matter much. > > Thanks, Fred! I should have thought to look in my PSpice libraries, > found it in Philips_RF.lib... minor differences in parasitics... > > * SOT23 parasitic model > Lb 4 5 .4n > Le 7 8 .83n > L1 2 4 .35n > L2 1 6 .17n > L3 3 8 .35n > Ccb 4 6 71f > Cbe 4 8 71f > > ...Jim ThompsonYep, and in the die parameters too. Not surprising, but the bias point discrepancies make this NXP model suspicious. First the 'measured' beta at 10mA is way in the low range for NXP at 25 (for 20 min/50 typ in NXP datasheet). So much for centering... Also the FTs are 4.8 for Infineon against 5.8 for NXP... Not far, but the datasheets say 5GHz at 15mA/ and 10V VCE bias where it's used here with only 1.8V I guess I trust the Infineon model more... -- Thanks, Fred.
Reply by ●August 18, 20122012-08-18
On Sat, 18 Aug 2012 21:21:32 +0200, Fred Bartoli <" "> wrote:>Jim Thompson a �crit : >> On Sat, 18 Aug 2012 20:03:59 +0200, Fred Bartoli <" "> wrote: >> >>> John Larkin a �crit : >>>> On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: >>>> >>>>> Phil Hobbs a �crit : >>>>>> dagmargoodboat@yahoo.com wrote: >>>>>>> On Aug 15, 4:34 pm, Phil Hobbs >>>>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: >>>>>>> >>>>>>> [...] >>>>>>> >>>>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing better to >>>>>>>> do are are slagging each other off about at the moment) runs the driver >>>>>>>> transistor at constant I_C, and lets you use a much faster NPN to >>>>>>>> stiffen the PNP at high frequency. Try simulating it--it's really good >>>>>>>> medicine. >>>>>>>> >>>>>>>> (I spent most of July up to my ears in high performance laser noise >>>>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pretty >>>>>>>> well up on all this at the moment.) >>>>>>> I like the pseudo-Darlington, I'm just leery because I had those >>>>>>> scream wickedly once using fast transistors. I don't think I ever >>>>>>> fixed it. >>>>>>> >>>>>>> It has a bunch of advantages here, including reducing the base current >>>>>>> i(b) error. >>>>>>> >>>>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: >>>>>>> >>>>>>> Vcc Vcc >>>>>>> -+- -+- >>>>>>> | | >>>>>>> | .-. R1 >>>>>>> | U1 | | 120 >>>>>>> | LM385 '-' >>>>>>> | -adj | >>>>>>> .---' | >>>>>>> A<-----R2-----+---------. >>>>>>> | | | >>>>>>> | |<' Q1 | >>>>>>> +-----+-----| BFT92 | >>>>>>> | | |\ | >>>>>>> R4 .-. --- C1 | |/ Q2 >>>>>>> 1K | | --- +--f.b.-| BFP640 >>>>>>> '-' | | |>. >>>>>>> | === .-. | >>>>>>> === | | R3 | >>>>>>> '-' 600 | >>>>>>> | | >>>>>>> '---------+----> >>>>>>> >>>>>>> R2's a pathetic provision for isolating U1 from the nS stuff. The >>>>>>> whole thing needs stabilizing against UHF oscillation, these are just >>>>>>> the bones. >>>>>>> >>>>>>> -- >>>>>>> Cheers, >>>>>>> James Arthur >>>>>> It wouldn't need much more than that, I don't think. Series-shunt pairs >>>>>> have poor stability when the transistors are about the same speed, due >>>>>> to having two lags in the loop, but the BFP640 is ~10x faster than the >>>>>> BFT92, which helps a lot. (Reducing the BFT92's collector current makes >>>>>> the difference wider and helps with stability.) >>>>>> >>>>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet >>>>>> charge-sensitive preamp (noise floor ~10 electrons per inverse >>>>>> bandwidth), and all it needed was an 0402 bead in the base--no worries >>>>>> at all. >>>>>> >>>>>> Cheers >>>>>> >>>>>> Phil Hobbs >>>>>> >>>>> I had a try at John's pb... >>>>> >>>>> Here it is, using a current source bootstrapped bootstrapped current >>>>> source (yes I really meant to say that :-). >>>>> >>>>> I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. >>>>> Some temperature dependent biasing is also provided but I relied on the >>>>> 8V supply stability over temp which may be wrong and you may have to >>>>> adapt a bit. >>>>> >>>>> Some simulations including parasitics give pretty interesting results. >>>>> Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have >>>>> any to measure it) >>>>> >>>>> Despite this this circuit shows excellent results (in simulation, so >>>>> apply the usual grain of salt). >>>>> >>>>> Still: >>>>> * better than 0.1% linearity with this low VAF figure which can be >>>>> further improved >>>>> * start up transients are under 1ns and the integrated error is >>>>> negligible... >>>>> * 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I >>>>> relied on your 8V supply stability, so you may have to adapt a bit) >>>>> * low output impedance, so no need for buffering if you just drive one >>>>> input (modeled here as the 2pF capacitance) and can stand the temp >>>>> dependent offset at start (or you can still add a NPN buffer to >>>>> compensate for VBE) >>>>> >>>>> >>>>> Here's a LTspice netlist. It's just for the drawing as I don't use it >>>>> for sim and have almost no BJT models and even less with parasitics... >>>>> >>>>> If you want to see the results I'll make a pdf. >>>>> >>>>> >>>>> ******************************************** >>>>> >>>>> Version 4 >>>>> SHEET 1 948 1004 >>>>> WIRE 16 32 -128 32 >>> << ... >> >>>>> SYMBOL njf -176 640 R0 >>>>> SYMATTR InstName J1 >>>>> SYMATTR Value NE3509 >>>>> >>>>> >>>>> >>>>> ******************************************** >>>> >>>> Hi, Fred, >>>> >>>> I tried it but it won't run without the BFT92 model. Do you have that? >>>> >>> >>> As I said I don't have them for LTspice, so it's why this netlist don't >>> run flat. But they are available from the Infineon datasheet, with all >>> the parasitics... (I don't know how to do build a new model for ltspice) >>> >>> You'd have to build a subcircuit for the SOT23 parasitics and include >>> the die model. Here is the one from the spice database I used. >>> >>> >>> .SUBCKT BFT92s 200 100 300 >>> ** ^ ^ ^ >>> ** C B E >>> ** >>> *SOT23 >>> Q1 2 1 3 BFT92Xm >>> LBI 1 10 0.85NH >>> LEI 3 30 0.69NH >>> CCB 10 2 84FF >>> CCE 2 30 165FF >>> CBE 10 30 73FF >>> LBO 10 100 0.51NH >>> LCO 2 200 0.49NH >>> LEO 30 300 0.61NH >>> >>> .MODEL BFT92Xm PNP( >>> + IS = 4.5354E-15 BF = 98.533 NF = 0.90551 >>> + VAF = 10.983 IKF = 0.016123 ISE = 1.2196E-14 >>> + NE = 1.1172 BR = 10.297 NR = 1.2703 >>> + VAR = 47.577 IKR = 0.019729 ISC = 2.4709E-17 >>> + NC = 1.206 RB = 7.9562 IRB = 0.00079584 >>> + RBM = 1.5939 RE = 1.5119 RC = 0.66749 >>> + CJE = 1.7785E-15 VJE = 0.79082 MJE = 0.32167 >>> + TF = 3.2171E-11 XTF = 0.30227 VTF = 0.21451 >>> + ITF = 1.3277E-05 PTF = 0 CJC = 9.2207E-13 >>> + VJC = 1.2 MJC = 0.3 XCJC = 0.3 >>> + TR = 2.0779E-09 CJS= 0 VJS = 0.75 >>> + MJS = 0 XTB = 0 EG = 1.11 >>> + XTI = 3 FC = 0.75167) >>> >>> .ENDS >>> >>> >>> Note that on my database model CBE is specified but is missing from the >>> datasheet that you can find here >>> >>> http://www.farnell.com/datasheets/62457.pdf (probably a datasheet error) >>> >>> >>> That 73fF value is coherent with what can be expected from a SOT23 >>> package so it probably is right. >>> You can try with and without it, but in this circuit it probably don't >>> matter much. >> >> Thanks, Fred! I should have thought to look in my PSpice libraries, >> found it in Philips_RF.lib... minor differences in parasitics... >> >> * SOT23 parasitic model >> Lb 4 5 .4n >> Le 7 8 .83n >> L1 2 4 .35n >> L2 1 6 .17n >> L3 3 8 .35n >> Ccb 4 6 71f >> Cbe 4 8 71f >> >> ...Jim Thompson > >Yep, and in the die parameters too. >Not surprising, but the bias point discrepancies make this NXP model >suspicious. > >First the 'measured' beta at 10mA is way in the low range for NXP at 25 >(for 20 min/50 typ in NXP datasheet). So much for centering... > >Also the FTs are 4.8 for Infineon against 5.8 for NXP... >Not far, but the datasheets say 5GHz at 15mA/ and 10V VCE bias where >it's used here with only 1.8V > >I guess I trust the Infineon model more...I would. For another project I just downloaded the Infineon CoolMOS libraries. I guess I should go back and get the BJT's as well. ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
Reply by ●August 18, 20122012-08-18
On Sat, 18 Aug 2012 11:01:01 -0700, Jim Thompson <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:>On Sat, 18 Aug 2012 10:27:34 -0700, John Larkin ><jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: > >>On Sat, 18 Aug 2012 09:46:55 -0700, Jim Thompson >><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >> >>>On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: >>> >>>>Phil Hobbs a �crit : >>>>> dagmargoodboat@yahoo.com wrote: >>>>>> On Aug 15, 4:34 pm, Phil Hobbs >>>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: >>>>>> >>>>>> [...] >>>>>> >>>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing better to >>>>>>> do are are slagging each other off about at the moment) runs the driver >>>>>>> transistor at constant I_C, and lets you use a much faster NPN to >>>>>>> stiffen the PNP at high frequency. Try simulating it--it's really good >>>>>>> medicine. >>>>>>> >>>>>>> (I spent most of July up to my ears in high performance laser noise >>>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pretty >>>>>>> well up on all this at the moment.) >>>>>> I like the pseudo-Darlington, I'm just leery because I had those >>>>>> scream wickedly once using fast transistors. I don't think I ever >>>>>> fixed it. >>>>>> >>>>>> It has a bunch of advantages here, including reducing the base current >>>>>> i(b) error. >>>>>> >>>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: >>>>>> >>>>>> Vcc Vcc >>>>>> -+- -+- >>>>>> | | >>>>>> | .-. R1 >>>>>> | U1 | | 120 >>>>>> | LM385 '-' >>>>>> | -adj | >>>>>> .---' | >>>>>> A<-----R2-----+---------. >>>>>> | | | >>>>>> | |<' Q1 | >>>>>> +-----+-----| BFT92 | >>>>>> | | |\ | >>>>>> R4 .-. --- C1 | |/ Q2 >>>>>> 1K | | --- +--f.b.-| BFP640 >>>>>> '-' | | |>. >>>>>> | === .-. | >>>>>> === | | R3 | >>>>>> '-' 600 | >>>>>> | | >>>>>> '---------+----> >>>>>> >>>>>> R2's a pathetic provision for isolating U1 from the nS stuff. The >>>>>> whole thing needs stabilizing against UHF oscillation, these are just >>>>>> the bones. >>>>>> >>>>>> -- >>>>>> Cheers, >>>>>> James Arthur >>>>> >>>>> It wouldn't need much more than that, I don't think. Series-shunt pairs >>>>> have poor stability when the transistors are about the same speed, due >>>>> to having two lags in the loop, but the BFP640 is ~10x faster than the >>>>> BFT92, which helps a lot. (Reducing the BFT92's collector current makes >>>>> the difference wider and helps with stability.) >>>>> >>>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet >>>>> charge-sensitive preamp (noise floor ~10 electrons per inverse >>>>> bandwidth), and all it needed was an 0402 bead in the base--no worries >>>>> at all. >>>>> >>>>> Cheers >>>>> >>>>> Phil Hobbs >>>>> >>>> >>>> >>>>I had a try at John's pb... >>>> >>>>Here it is, using a current source bootstrapped bootstrapped current >>>>source (yes I really meant to say that :-). >>>> >>>>I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. >>>>Some temperature dependent biasing is also provided but I relied on the >>>>8V supply stability over temp which may be wrong and you may have to >>>>adapt a bit. >>>> >>>>Some simulations including parasitics give pretty interesting results. >>>>Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have >>>>any to measure it) >>>> >>>>Despite this this circuit shows excellent results (in simulation, so >>>>apply the usual grain of salt). >>>> >>>>Still: >>>>* better than 0.1% linearity with this low VAF figure which can be >>>>further improved >>>>* start up transients are under 1ns and the integrated error is >>>>negligible... >>>>* 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I >>>>relied on your 8V supply stability, so you may have to adapt a bit) >>>>* low output impedance, so no need for buffering if you just drive one >>>>input (modeled here as the 2pF capacitance) and can stand the temp >>>>dependent offset at start (or you can still add a NPN buffer to >>>>compensate for VBE) >>>> >>>> >>>>Here's a LTspice netlist. It's just for the drawing as I don't use it >>>>for sim and have almost no BJT models and even less with parasitics... >>>> >>>>If you want to see the results I'll make a pdf. >>>> >>>> >>>>******************************************** >>>> >>>>Version 4 >>>>SHEET 1 948 1004 >>>[snip .ASC, retrieve at Message-ID: >>><502fc06b$0$6124$426a74cc@news.free.fr>] >>>>SYMATTR InstName J1 >>>>SYMATTR Value NE3509 >>>> >>>> >>>> >>>>******************************************** >>> >>>Finally! An inductor used properly. I can envision other such >>>structures that are possibly simpler, but I ain'ta-gonna give Larkin >>>no free ride ;-) >> >>In other words, as usual, you contribute nothing on topic. All you >>remember how to do is boast and cluck. > >Just as long as I keep you frothing at the mouth ;-)You delude yourself, old man.> >> >>I don't think you could play the circuit design game if you wanted to. >>You are hiding your senility behind "not helping Larkin." You're >>useless off-chip anyhow. >> >>> >>>Good job, Fred! >>> >> >>So, you approve of Fred's design? > >I couldn't run it, but it certainly is heading in a better direction >than all those insane IB correctors.I did Ib correction with one resistor and half of a dual opamp. Microwave PNPs have insane betas, like 20. You propose to ignore base current over temperature? Yeah, right, you have a better way to do this, but you won't reveal it. Pitiful bluff.> >What don't _you_ post your solution as an LTspice ASC file so that >everyone can verify your brilliance ?:-)I wouldn't Spice my fast current source designs. The fast stuff would be inaccurate because the part models aren't good enough. The slow loops, like the Ib correction, are way too simple to need Spicing.> >What have you designed lately? Yesterday I designed a magnetically >controlled sign board :-)EUV light source controller. EUV energy/waveform digitizer. Particle discriminator. Four different laser drivers. LVDT/synchro simulator. USB picosecond pulse generator. Ethernet interfaced picosecond-resolution 5-channel time stamper. NMR gradient driver and temperature controller combo. Bunch of fiberoptic stuff. A couple of fast high-voltage pulsers. The thing I like about designing this stuff is that I release the drawings to production and they crank out copies, and revenue, for a decade or so afterwards. I never sell IP or engineering. -- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom timing and laser controllers Photonics and fiberoptic TTL data links VME analog, thermocouple, LVDT, synchro, tachometer Multichannel arbitrary waveform generators
Reply by ●August 18, 20122012-08-18
On Sat, 18 Aug 2012 13:24:08 -0700, John Larkin <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote:>On Sat, 18 Aug 2012 11:01:01 -0700, Jim Thompson ><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: > >>On Sat, 18 Aug 2012 10:27:34 -0700, John Larkin >><jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: >> >>>On Sat, 18 Aug 2012 09:46:55 -0700, Jim Thompson >>><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>> >>>>On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: >>>> >>>>>Phil Hobbs a �crit : >>>>>> dagmargoodboat@yahoo.com wrote: >>>>>>> On Aug 15, 4:34 pm, Phil Hobbs >>>>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: >>>>>>> >>>>>>> [...] >>>>>>> >>>>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing better to >>>>>>>> do are are slagging each other off about at the moment) runs the driver >>>>>>>> transistor at constant I_C, and lets you use a much faster NPN to >>>>>>>> stiffen the PNP at high frequency. Try simulating it--it's really good >>>>>>>> medicine. >>>>>>>> >>>>>>>> (I spent most of July up to my ears in high performance laser noise >>>>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pretty >>>>>>>> well up on all this at the moment.) >>>>>>> I like the pseudo-Darlington, I'm just leery because I had those >>>>>>> scream wickedly once using fast transistors. I don't think I ever >>>>>>> fixed it. >>>>>>> >>>>>>> It has a bunch of advantages here, including reducing the base current >>>>>>> i(b) error. >>>>>>> >>>>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: >>>>>>> >>>>>>> Vcc Vcc >>>>>>> -+- -+- >>>>>>> | | >>>>>>> | .-. R1 >>>>>>> | U1 | | 120 >>>>>>> | LM385 '-' >>>>>>> | -adj | >>>>>>> .---' | >>>>>>> A<-----R2-----+---------. >>>>>>> | | | >>>>>>> | |<' Q1 | >>>>>>> +-----+-----| BFT92 | >>>>>>> | | |\ | >>>>>>> R4 .-. --- C1 | |/ Q2 >>>>>>> 1K | | --- +--f.b.-| BFP640 >>>>>>> '-' | | |>. >>>>>>> | === .-. | >>>>>>> === | | R3 | >>>>>>> '-' 600 | >>>>>>> | | >>>>>>> '---------+----> >>>>>>> >>>>>>> R2's a pathetic provision for isolating U1 from the nS stuff. The >>>>>>> whole thing needs stabilizing against UHF oscillation, these are just >>>>>>> the bones. >>>>>>> >>>>>>> -- >>>>>>> Cheers, >>>>>>> James Arthur >>>>>> >>>>>> It wouldn't need much more than that, I don't think. Series-shunt pairs >>>>>> have poor stability when the transistors are about the same speed, due >>>>>> to having two lags in the loop, but the BFP640 is ~10x faster than the >>>>>> BFT92, which helps a lot. (Reducing the BFT92's collector current makes >>>>>> the difference wider and helps with stability.) >>>>>> >>>>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet >>>>>> charge-sensitive preamp (noise floor ~10 electrons per inverse >>>>>> bandwidth), and all it needed was an 0402 bead in the base--no worries >>>>>> at all. >>>>>> >>>>>> Cheers >>>>>> >>>>>> Phil Hobbs >>>>>> >>>>> >>>>> >>>>>I had a try at John's pb... >>>>> >>>>>Here it is, using a current source bootstrapped bootstrapped current >>>>>source (yes I really meant to say that :-). >>>>> >>>>>I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. >>>>>Some temperature dependent biasing is also provided but I relied on the >>>>>8V supply stability over temp which may be wrong and you may have to >>>>>adapt a bit. >>>>> >>>>>Some simulations including parasitics give pretty interesting results. >>>>>Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have >>>>>any to measure it) >>>>> >>>>>Despite this this circuit shows excellent results (in simulation, so >>>>>apply the usual grain of salt). >>>>> >>>>>Still: >>>>>* better than 0.1% linearity with this low VAF figure which can be >>>>>further improved >>>>>* start up transients are under 1ns and the integrated error is >>>>>negligible... >>>>>* 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I >>>>>relied on your 8V supply stability, so you may have to adapt a bit) >>>>>* low output impedance, so no need for buffering if you just drive one >>>>>input (modeled here as the 2pF capacitance) and can stand the temp >>>>>dependent offset at start (or you can still add a NPN buffer to >>>>>compensate for VBE) >>>>> >>>>> >>>>>Here's a LTspice netlist. It's just for the drawing as I don't use it >>>>>for sim and have almost no BJT models and even less with parasitics... >>>>> >>>>>If you want to see the results I'll make a pdf. >>>>> >>>>> >>>>>******************************************** >>>>> >>>>>Version 4 >>>>>SHEET 1 948 1004 >>>>[snip .ASC, retrieve at Message-ID: >>>><502fc06b$0$6124$426a74cc@news.free.fr>] >>>>>SYMATTR InstName J1 >>>>>SYMATTR Value NE3509 >>>>> >>>>> >>>>> >>>>>******************************************** >>>> >>>>Finally! An inductor used properly. I can envision other such >>>>structures that are possibly simpler, but I ain'ta-gonna give Larkin >>>>no free ride ;-) >>> >>>In other words, as usual, you contribute nothing on topic. All you >>>remember how to do is boast and cluck. >> >>Just as long as I keep you frothing at the mouth ;-) > >You delude yourself, old man. > > >> >>> >>>I don't think you could play the circuit design game if you wanted to. >>>You are hiding your senility behind "not helping Larkin." You're >>>useless off-chip anyhow. >>> >>>> >>>>Good job, Fred! >>>> >>> >>>So, you approve of Fred's design? >> >>I couldn't run it, but it certainly is heading in a better direction >>than all those insane IB correctors. > >I did Ib correction with one resistor and half of a dual opamp. >Microwave PNPs have insane betas, like 20. You propose to ignore base >current over temperature? Yeah, right, you have a better way to do >this, but you won't reveal it. Pitiful bluff.Crap! I have ignition systems (still) running in cars where the control device had a nominal beta of THREE with IB correction. Read Frederiksen's tome... I invented such schemes.> > >> >>What don't _you_ post your solution as an LTspice ASC file so that >>everyone can verify your brilliance ?:-) > >I wouldn't Spice my fast current source designs. The fast stuff would >be inaccurate because the part models aren't good enough. The slow >loops, like the Ib correction, are way too simple to need Spicing. > >> >>What have you designed lately? Yesterday I designed a magnetically >>controlled sign board :-) > >EUV light source controller. EUV energy/waveform digitizer. Particle >discriminator. Four different laser drivers. LVDT/synchro simulator. >USB picosecond pulse generator. Ethernet interfaced >picosecond-resolution 5-channel time stamper. NMR gradient driver and >temperature controller combo. Bunch of fiberoptic stuff. A couple of >fast high-voltage pulsers. > >The thing I like about designing this stuff is that I release the >drawings to production and they crank out copies, and revenue, for a >decade or so afterwards. I never sell IP or engineering.You don't have any to sell. I have 18 patents. How many do you have? ...Jim Thompson -- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | I love to cook with wine. Sometimes I even put it in the food.
Reply by ●August 19, 20122012-08-19
On Sat, 18 Aug 2012 18:17:36 -0700, Jim Thompson <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:>On Sat, 18 Aug 2012 13:24:08 -0700, John Larkin ><jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: > >>On Sat, 18 Aug 2012 11:01:01 -0700, Jim Thompson >><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >> >>>On Sat, 18 Aug 2012 10:27:34 -0700, John Larkin >>><jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: >>> >>>>On Sat, 18 Aug 2012 09:46:55 -0700, Jim Thompson >>>><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>>> >>>>>On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: >>>>> >>>>>>Phil Hobbs a �crit : >>>>>>> dagmargoodboat@yahoo.com wrote: >>>>>>>> On Aug 15, 4:34 pm, Phil Hobbs >>>>>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: >>>>>>>> >>>>>>>> [...] >>>>>>>> >>>>>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing better to >>>>>>>>> do are are slagging each other off about at the moment) runs the driver >>>>>>>>> transistor at constant I_C, and lets you use a much faster NPN to >>>>>>>>> stiffen the PNP at high frequency. Try simulating it--it's really good >>>>>>>>> medicine. >>>>>>>>> >>>>>>>>> (I spent most of July up to my ears in high performance laser noise >>>>>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'm pretty >>>>>>>>> well up on all this at the moment.) >>>>>>>> I like the pseudo-Darlington, I'm just leery because I had those >>>>>>>> scream wickedly once using fast transistors. I don't think I ever >>>>>>>> fixed it. >>>>>>>> >>>>>>>> It has a bunch of advantages here, including reducing the base current >>>>>>>> i(b) error. >>>>>>>> >>>>>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: >>>>>>>> >>>>>>>> Vcc Vcc >>>>>>>> -+- -+- >>>>>>>> | | >>>>>>>> | .-. R1 >>>>>>>> | U1 | | 120 >>>>>>>> | LM385 '-' >>>>>>>> | -adj | >>>>>>>> .---' | >>>>>>>> A<-----R2-----+---------. >>>>>>>> | | | >>>>>>>> | |<' Q1 | >>>>>>>> +-----+-----| BFT92 | >>>>>>>> | | |\ | >>>>>>>> R4 .-. --- C1 | |/ Q2 >>>>>>>> 1K | | --- +--f.b.-| BFP640 >>>>>>>> '-' | | |>. >>>>>>>> | === .-. | >>>>>>>> === | | R3 | >>>>>>>> '-' 600 | >>>>>>>> | | >>>>>>>> '---------+----> >>>>>>>> >>>>>>>> R2's a pathetic provision for isolating U1 from the nS stuff. The >>>>>>>> whole thing needs stabilizing against UHF oscillation, these are just >>>>>>>> the bones. >>>>>>>> >>>>>>>> -- >>>>>>>> Cheers, >>>>>>>> James Arthur >>>>>>> >>>>>>> It wouldn't need much more than that, I don't think. Series-shunt pairs >>>>>>> have poor stability when the transistors are about the same speed, due >>>>>>> to having two lags in the loop, but the BFP640 is ~10x faster than the >>>>>>> BFT92, which helps a lot. (Reducing the BFT92's collector current makes >>>>>>> the difference wider and helps with stability.) >>>>>>> >>>>>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very quiet >>>>>>> charge-sensitive preamp (noise floor ~10 electrons per inverse >>>>>>> bandwidth), and all it needed was an 0402 bead in the base--no worries >>>>>>> at all. >>>>>>> >>>>>>> Cheers >>>>>>> >>>>>>> Phil Hobbs >>>>>>> >>>>>> >>>>>> >>>>>>I had a try at John's pb... >>>>>> >>>>>>Here it is, using a current source bootstrapped bootstrapped current >>>>>>source (yes I really meant to say that :-). >>>>>> >>>>>>I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT92. >>>>>>Some temperature dependent biasing is also provided but I relied on the >>>>>>8V supply stability over temp which may be wrong and you may have to >>>>>>adapt a bit. >>>>>> >>>>>>Some simulations including parasitics give pretty interesting results. >>>>>>Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't have >>>>>>any to measure it) >>>>>> >>>>>>Despite this this circuit shows excellent results (in simulation, so >>>>>>apply the usual grain of salt). >>>>>> >>>>>>Still: >>>>>>* better than 0.1% linearity with this low VAF figure which can be >>>>>>further improved >>>>>>* start up transients are under 1ns and the integrated error is >>>>>>negligible... >>>>>>* 0.25% dispersion at 3.9V final value over temperature (-50/+75) (I >>>>>>relied on your 8V supply stability, so you may have to adapt a bit) >>>>>>* low output impedance, so no need for buffering if you just drive one >>>>>>input (modeled here as the 2pF capacitance) and can stand the temp >>>>>>dependent offset at start (or you can still add a NPN buffer to >>>>>>compensate for VBE) >>>>>> >>>>>> >>>>>>Here's a LTspice netlist. It's just for the drawing as I don't use it >>>>>>for sim and have almost no BJT models and even less with parasitics... >>>>>> >>>>>>If you want to see the results I'll make a pdf. >>>>>> >>>>>> >>>>>>******************************************** >>>>>> >>>>>>Version 4 >>>>>>SHEET 1 948 1004 >>>>>[snip .ASC, retrieve at Message-ID: >>>>><502fc06b$0$6124$426a74cc@news.free.fr>] >>>>>>SYMATTR InstName J1 >>>>>>SYMATTR Value NE3509 >>>>>> >>>>>> >>>>>> >>>>>>******************************************** >>>>> >>>>>Finally! An inductor used properly. I can envision other such >>>>>structures that are possibly simpler, but I ain'ta-gonna give Larkin >>>>>no free ride ;-) >>>> >>>>In other words, as usual, you contribute nothing on topic. All you >>>>remember how to do is boast and cluck. >>> >>>Just as long as I keep you frothing at the mouth ;-) >> >>You delude yourself, old man. >> >> >>> >>>> >>>>I don't think you could play the circuit design game if you wanted to. >>>>You are hiding your senility behind "not helping Larkin." You're >>>>useless off-chip anyhow. >>>> >>>>> >>>>>Good job, Fred! >>>>> >>>> >>>>So, you approve of Fred's design? >>> >>>I couldn't run it, but it certainly is heading in a better direction >>>than all those insane IB correctors. >> >>I did Ib correction with one resistor and half of a dual opamp. >>Microwave PNPs have insane betas, like 20. You propose to ignore base >>current over temperature? Yeah, right, you have a better way to do >>this, but you won't reveal it. Pitiful bluff. > >Crap! I have ignition systems (still) running in cars where the >control device had a nominal beta of THREE with IB correction. Read >Frederiksen's tome... I invented such schemes. > >> >> >>> >>>What don't _you_ post your solution as an LTspice ASC file so that >>>everyone can verify your brilliance ?:-) >> >>I wouldn't Spice my fast current source designs. The fast stuff would >>be inaccurate because the part models aren't good enough. The slow >>loops, like the Ib correction, are way too simple to need Spicing. >> >>> >>>What have you designed lately? Yesterday I designed a magnetically >>>controlled sign board :-) >> >>EUV light source controller. EUV energy/waveform digitizer. Particle >>discriminator. Four different laser drivers. LVDT/synchro simulator. >>USB picosecond pulse generator. Ethernet interfaced >>picosecond-resolution 5-channel time stamper. NMR gradient driver and >>temperature controller combo. Bunch of fiberoptic stuff. A couple of >>fast high-voltage pulsers. >> >>The thing I like about designing this stuff is that I release the >>drawings to production and they crank out copies, and revenue, for a >>decade or so afterwards. I never sell IP or engineering. > >You don't have any to sell.Of course I have IP. But I use it in my own products and don't sell it for a one-time consulting fee. In a few cases, I license it, which just means that someone else does the manufacturing to make my ongoing revenue stream. But they don't own it. I have 18 patents. How many do you have?> > ...Jim ThompsonOne, but I have rights to use it. How many of your 18 are assigned to someone else? How much of the IP in your active patents can you sell? Patents are a suckers's game if they are assigned to someone else. You get to brag, but they get the money. -- John Larkin Highland Technology Inc www.highlandtechnology.com jlarkin at highlandtechnology dot com Precision electronic instrumentation Picosecond-resolution Digital Delay and Pulse generators Custom timing and laser controllers Photonics and fiberoptic TTL data links VME analog, thermocouple, LVDT, synchro, tachometer Multichannel arbitrary waveform generators
Reply by ●August 19, 20122012-08-19
On Aug 19, 6:28=A0am, John Larkin <jjlar...@highNOTlandTHIStechnologyPART.com> wrote:> On Sat, 18 Aug 2012 18:17:36 -0700, Jim Thompson > > > > > > > > > > <To-Email-Use-The-Envelope-I...@On-My-Web-Site.com> wrote: > >On Sat, 18 Aug 2012 13:24:08 -0700, John Larkin > ><jjlar...@highNOTlandTHIStechnologyPART.com> wrote: > > >>On Sat, 18 Aug 2012 11:01:01 -0700, Jim Thompson > >><To-Email-Use-The-Envelope-I...@On-My-Web-Site.com> wrote: > > >>>On Sat, 18 Aug 2012 10:27:34 -0700, John Larkin > >>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote: > > >>>>On Sat, 18 Aug 2012 09:46:55 -0700, Jim Thompson > >>>><To-Email-Use-The-Envelope-I...@On-My-Web-Site.com> wrote: > > >>>>>On Sat, 18 Aug 2012 18:18:51 +0200, Fred Bartoli <" "> wrote: > > >>>>>>Phil Hobbs a =E9crit : > >>>>>>> dagmargoodb...@yahoo.com wrote: > >>>>>>>> On Aug 15, 4:34 pm, Phil Hobbs > >>>>>>>> <pcdhSpamMeSensel...@electrooptical.net> wrote: > > >>>>>>>> [...] > > >>>>>>>>> The NPN/PNP series-shunt pair (the one that folks with nothing =better to> >>>>>>>>> do are are slagging each other off about at the moment) runs th=e driver> >>>>>>>>> transistor at constant I_C, and lets you use a much faster NPN =to> >>>>>>>>> stiffen the PNP at high frequency. =A0Try simulating it--it's r=eally good> >>>>>>>>> medicine. > > >>>>>>>>> (I spent most of July up to my ears in high performance laser n=oise> >>>>>>>>> cancellers, which rely on a lot of those sorts of tricks, so I'=m pretty> >>>>>>>>> well up on all this at the moment.) > >>>>>>>> I like the pseudo-Darlington, I'm just leery because I had those > >>>>>>>> scream wickedly once using fast transistors. =A0I don't think I =ever> >>>>>>>> fixed it. > > >>>>>>>> It has a bunch of advantages here, including reducing the base c=urrent> >>>>>>>> i(b) error. > > >>>>>>>> Here's a skeleton idea for a Bloggs-Hobbs hybrid: > > >>>>>>>> =A0 =A0 Vcc =A0 =A0 =A0 =A0 =A0 Vcc > >>>>>>>> =A0 =A0 -+- =A0 =A0 =A0 =A0 =A0 -+- > >>>>>>>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0 | > >>>>>>>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0.-. R1 > >>>>>>>> =A0 =A0 =A0| U1 =A0 =A0 =A0 =A0 | | 120 > >>>>>>>> =A0 =A0 =A0| LM385 =A0 =A0 =A0'-' > >>>>>>>> =A0 =A0 =A0| =A0-adj =A0 =A0 =A0 | > >>>>>>>> =A0 =A0.---' =A0 =A0 =A0 =A0 =A0 | > >>>>>>>> =A0 =A0 =A0A<-----R2-----+---------. > >>>>>>>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 =A0 | =A0 =A0 =A0 =A0 | > >>>>>>>> =A0 =A0 =A0| =A0 =A0 =A0 =A0 =A0 |<' Q1 =A0 =A0 =A0| > >>>>>>>> =A0 =A0 =A0+-----+-----| =A0BFT92 =A0 =A0| > >>>>>>>> =A0 =A0 =A0| =A0 =A0 | =A0 =A0 |\ =A0 =A0 =A0 =A0 =A0| > >>>>>>>> =A0R4 .-. =A0 --- C1 =A0 | =A0 =A0 =A0 |/ =A0Q2 > >>>>>>>> =A01K | | =A0 --- =A0 =A0 =A0+--f.b.-| =A0BFP640 > >>>>>>>> =A0 =A0 '-' =A0 =A0| =A0 =A0 =A0 | =A0 =A0 =A0 |>. > >>>>>>>> =A0 =A0 =A0| =A0 =A0=3D=3D=3D =A0 =A0 .-. =A0 =A0 =A0 =A0| > >>>>>>>> =A0 =A0 =3D=3D=3D =A0 =A0 =A0 =A0 =A0 | | R3 =A0 =A0 | > >>>>>>>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 '-' 600 =A0 =A0| > >>>>>>>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0| =A0 =A0 =A0 =A0 | > >>>>>>>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0'---------+----> > > >>>>>>>> R2's a pathetic provision for isolating U1 from the nS stuff. ==A0The> >>>>>>>> whole thing needs stabilizing against UHF oscillation, these are=just> >>>>>>>> the bones. > > >>>>>>>> -- > >>>>>>>> Cheers, > >>>>>>>> James Arthur > > >>>>>>> It wouldn't need much more than that, I don't think. =A0Series-sh=unt pairs> >>>>>>> have poor stability when the transistors are about the same speed=, due> >>>>>>> to having two lags in the loop, but the BFP640 is ~10x faster tha=n the> >>>>>>> BFT92, which helps a lot. =A0(Reducing the BFT92's collector curr=ent makes> >>>>>>> the difference wider and helps with stability.) > > >>>>>>> I used the BFP640 as a cascode for an ATF38143 pHEMT in a very qu=iet> >>>>>>> charge-sensitive preamp (noise floor ~10 electrons per inverse > >>>>>>> bandwidth), and all it needed was an 0402 bead in the base--no wo=rries> >>>>>>> at all. > > >>>>>>> Cheers > > >>>>>>> Phil Hobbs > > >>>>>>I had a try at John's pb... > > >>>>>>Here it is, using a current source bootstrapped bootstrapped curren=t> >>>>>>source (yes I really meant to say that :-). > > >>>>>>I use the +8V/-5V supplies that you mentioned IIRC and only 3 x BFT=92.> >>>>>>Some temperature dependent biasing is also provided but I relied on=the> >>>>>>8V supply stability over temp which may be wrong and you may have t=o> >>>>>>adapt a bit. > > >>>>>>Some simulations including parasitics give pretty interesting resul=ts.> >>>>>>Note that my BFT92 model VAF is 11V ! Is this low real ? (I don't h=ave> >>>>>>any to measure it) > > >>>>>>Despite this this circuit shows excellent results (in simulation, s=o> >>>>>>apply the usual grain of salt). > > >>>>>>Still: > >>>>>>* better than 0.1% linearity with this low VAF figure which can be > >>>>>>further improved > >>>>>>* start up transients are under 1ns and the integrated error is > >>>>>>negligible... > >>>>>>* 0.25% dispersion at 3.9V final value over temperature (-50/+75) (=I> >>>>>>relied on your 8V supply stability, so you may have to adapt a bit) > >>>>>>* low output impedance, so no need for buffering if you just drive =one> >>>>>>input (modeled here as the 2pF capacitance) and can stand the temp > >>>>>>dependent offset at start (or you can still add a NPN buffer to > >>>>>>compensate for VBE) > > >>>>>>Here's a LTspice netlist. It's just for the drawing as I don't use =it> >>>>>>for sim and have almost no BJT models and even less with parasitics=...> > >>>>>>If you want to see the results I'll make a pdf. > > >>>>>>******************************************** > > >>>>>>Version 4 > >>>>>>SHEET 1 948 1004 > >>>>>[snip .ASC, retrieve at Message-ID: > >>>>><502fc06b$0$6124$426a7...@news.free.fr>] > >>>>>>SYMATTR InstName J1 > >>>>>>SYMATTR Value NE3509 > > >>>>>>******************************************** > > >>>>>Finally! =A0An inductor used properly. =A0I can envision other such > >>>>>structures that are possibly simpler, but I ain'ta-gonna give Larkin > >>>>>no free ride ;-) > > >>>>In other words, as usual, you contribute nothing on topic. All you > >>>>remember how to do is boast and cluck. > > >>>Just as long as I keep you frothing at the mouth ;-) > > >>You delude yourself, old man. > > >>>>I don't think you could play the circuit design game if you wanted to=.> >>>>You are hiding your senility behind "not helping Larkin." You're > >>>>useless off-chip anyhow. > > >>>>>Good job, Fred! > > >>>>So, you approve of Fred's design? > > >>>I couldn't run it, but it certainly is heading in a better direction > >>>than all those insane IB correctors. > > >>I did Ib correction with one resistor and half of a dual opamp. > >>Microwave PNPs have insane betas, like 20. You propose to ignore base > >>current over temperature? Yeah, right, you have a better way to do > >>this, but you won't reveal it. Pitiful bluff. > > >Crap! =A0I have ignition systems (still) running in cars where the > >control device had a nominal beta of THREE with IB correction. =A0Read > >Frederiksen's tome... I invented such schemes. > > >>>What don't _you_ post your solution as an LTspice ASC file so that > >>>everyone can verify your brilliance ?:-) > > >>I wouldn't Spice my fast current source designs. The fast stuff would > >>be inaccurate because the part models aren't good enough. The slow > >>loops, like the Ib correction, are way too simple to need Spicing. > > >>>What have you designed lately? =A0Yesterday I designed a magnetically > >>>controlled sign board :-) > > >>EUV light source controller. EUV energy/waveform digitizer. Particle > >>discriminator. Four different laser drivers. LVDT/synchro simulator. > >>USB picosecond pulse generator. Ethernet interfaced > >>picosecond-resolution 5-channel time stamper. NMR gradient driver and > >>temperature controller combo. Bunch of fiberoptic stuff. A couple of > >>fast high-voltage pulsers. > > >>The thing I like about designing this stuff is that I release the > >>drawings to production and they crank out copies, and revenue, for a > >>decade or so afterwards. I never sell IP or engineering. > > >You don't have any to sell. > > Of course I have IP. But I use it in my own products and don't sell it > for a one-time consulting fee. In a few cases, I license it, which > just means that someone else does the manufacturing to make my ongoing > revenue stream. But they don't own it. > > =A0I have 18 patents. =A0How many do you have? > > > > > =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0==A0 =A0...Jim Thompson> > One, but I have rights to use it. How many of your 18 are assigned to > someone else? > > How much of the IP in your active patents can you sell? Patents are a > suckers's game if they are assigned to someone else. You get to brag, > but they get the money.Patents aren't any kind of sucker's game, but it doesn't make sense to own one if you haven't got the resources to defend it in court, which can take years and cost millions of dollars. http://en.wikipedia.org/wiki/Edwin_Howard_Armstrong Those of us who worked for other people rather liked being named as the inventor on a patent. It looks good on your CV, and doesn't cost you anything. Admittedly some employers are more enthusiastic about patents than others. Two of my three patents come from my three years at EMI Central Research, and the other one from my nine years with Cambridge Instruments (in Cambridge UK) where I was no less productive but where the company didn't have a full-time patent department. -- Bill Sloman, Nijmegen
Reply by ●August 19, 20122012-08-19
On Aug 16, 10:01=A0pm, Phil Hobbs <pcdhSpamMeSensel...@electrooptical.net> wrote:> John Larkin wrote: > > > On Thu, 16 Aug 2012 17:28:16 -0400, Phil Hobbs > > <pcdhSpamMeSensel...@electrooptical.net> wrote: > > > >On 08/16/2012 05:19 PM, John Larkin wrote: > > >> On Thu, 16 Aug 2012 12:20:49 -0700 (PDT), dagmargoodb...@yahoo.com > > >> wrote: > > > >>> On Aug 16, 11:29 am, Jim Thompson<To-Email-Use-The-Envelope-I...@On=-> > >>> My-Web-Site.com> =A0wrote: > > >>>> On Thu, 16 Aug 2012 07:39:04 -0700 (PDT), dagmargoodb...@yahoo.com > > >>>> wrote: > > > >>>>> On Aug 15, 10:43 pm, John Larkin > > >>>>> <jjlar...@highNOTlandTHIStechnologyPART.com> =A0wrote: > > > >>>>>> Hey, how about this? > > > >>>>>>https://dl.dropbox.com/u/53724080/Circuits/Isrc_simpler.JPG > > > >>>>> It's kind of a variation on Phil's active inductor. =A0That defea=ts all> > >>>>> the oscillation modes I can think of so far. (c-b, e-b, c-e)(Whic=h> > >>>>> means little, since circuits are so ingenious inventing new ones.=)> > > >>>>>> Ludicrously simple. No Miller nonsense. Might not oscillate. MMB=TH81> > >>>>>> would have about 2 pF out, or BFT92 would have about 0.75 but lo=wer> > >>>>>> beta. > > > >>>>> You could use the beta-cancellation trick. =A0But, stray and para=sitic> > >>>>> capacitance will be big factors compared to a 10pF integration ca=p.> > > >>>>>> If this works, I should have done it years ago. > > > >>>> Yup. =A0But I'd still add a base resistor, for the very reasons Ge=rhard> > >>>> Hoffmann brought up... the output Z of the OpAmp is a big unknown =(*).> > > >>>> (*) Which brings to mind... maybe add some DC load to the OpAmp to > > >>>> keep it out of the class-B region. > > > >>> I agree with both those points, and meant to mention them. =A0I don='t> > >>> trust the op-amp to behave nicely to impulses, or its output to be > > >>> stiff. =A0I'd expect the output to be hi-z on this timescale. > > > >>> If anything I'd be tempted to bypass the bjt's base to the supply > > >>> rail. The BJT then handles the fast stuff. =A0Isolate that capaciti=ve> > >>> load from the op amp's output with a series resistor, and stabilize > > >>> the op amp loop itself with a feed-forward cap. =A0Standard stuff. > > > >>> Does it all matter? =A0Unclamping the integration cap will feed-thr=u a> > >>> small impulse through to the bjt / CCS. =A0After that, it's pretty > > >>> clean. =A0So, the op-amp's output impedance might not matter, but I > > >>> suspect it will. > > > >>> I invented loading LM324 outputs all by myself waayyyy back as a pu=nk> > >>> kid, to stiffen them up and kill the crossover THD. =A0Of course > > >>> everyone else on the planet thought of it too. :-) > > > >> Me too! And I also invented loading the charge pump phase detector i=n> > >> the 4046. > > > >> I also invented the successive-detection log video detector, and the > > >> dual-slope ADC. I did those at times when I could have patented them=.> > >> I'd be smoking seegars on the Riviera. > > > >DLVAs have been around since at least the war: e.g. US Patent 2577506 =to> > >Belleville, filed 1945. > > > >Cheers > > > >Phil Hobbs > > > Oh well. But I did invent the dual-slope some years before Fairchild > > (?) patented it. > > I used to be a big fan of the Intersil ICL7109 back in the day. =A0Dual > slope was magic. > > CheersDon't forget the ICL7135 ... this datasheet was updated in 2007. http://www.intersil.com/content/dam/Intersil/documents/fn30/fn3093.pdf I lusted for that part in a design ages ago, but it was much too expensive, ~$15 IIRC (roughly $40 today). I rolled my own instead: 82c53, 74HCxx state machine controller, etc. It was a big upgrade over our existing products, all 12-bit. I got a bucket of galvanically isolated 4-1/2 digit ICL7135 isolated LED digital panel meters at a flea market not too long ago, for less than the 1pc chip price back in the day. -- Cheers, James Arthur
