Op-amp models in LT Spice
Paul Burridge
Just been flicking through the fairly generous range of models in the
LT op-amp library and have the following questions,,,
Which LT device model is most similar to the LM324?
Which LT device model is most similar to the 741?
When an LT device is described as "35Mhz" or "90Mhz" or whatever, does
that means it's good for anything up to that frequency or only good
for a narrow band around that figure?
Finally, do none of the models 'care' about whether they're single or
dual supply?
Finally, finally, isn't there a model or set of models which you don't
have to connect up to a supply at all? If you're not going to generate
a board from the schematic, it just seems a nuisance to have to 'power
them up.'
Thanks,
p
--
The BBC: Licensed at public expense to spread lies.
Stefan Heinzmann
> Just been flicking through the fairly generous range of models in the
> LT op-amp library and have the following questions,,,
> Which LT device model is most similar to the LM324?
> Which LT device model is most similar to the 741?
There's none. If you want those OpAmps, get one of the models for them
from the manufacturers, i.e. National. They can be used with LtSpice, too.
> When an LT device is described as "35Mhz" or "90Mhz" or whatever, does
> that means it's good for anything up to that frequency or only good
> for a narrow band around that figure?
35MHz what? 90MHz what? GBP, presumably? Well, put simply, this usually
means that the product of gain and maximum signal frequency in your
circuit may not exceed the given number. For the details, consult your
favorite textbook. Didn't you say you had AoE?
> Finally, do none of the models 'care' about whether they're single or
> dual supply?
The OpAmps don't usually care, either. So why should the model?
OpAmps advertised as Single Supply usually have a common mode input
range that extends down to the negative supply. That makes them easy to
deal with when the negative supply is connected to signal ground.
Depending on your operating conditions, a dual supply OpAmp may be
perfectly happy in a single supply circuit (and vice versa). For this
reason, I always thought that Single Supply was just a marketing buzzword.
> Finally, finally, isn't there a model or set of models which you don't
> have to connect up to a supply at all? If you're not going to generate
> a board from the schematic, it just seems a nuisance to have to 'power
> them up.'
Yes. Its symbol is called opamp. Alternatively, you can model an ideal
amplifier using a voltage controlled voltage source.
--
Cheers
Stefan
Winfield Hill
>
> Yes. Its symbol is called opamp. Alternatively, you can model an ideal
> amplifier using a voltage controlled voltage source.
Except you can't model an opamp with a VCVS, instead you'll need a
VCCS followed by a VCVS, and with a capacitor to ground at the node,
to set the gain rolloff to the opamp's f_T. For example if both the
VCCS and VCVS gains are unity, then a C = 1 / 2pi fT = 0.16uF cap
will set the dominant pole for a f_T = 1MHz perfect-opamp bandwidth.
Adding a 100k resistor will set a maximum gain of 100dB at DC. But
real opamps have many other serious defects, like a second pole not
too far above the f_T frequency (which reduces feedback-loop phase
margins), and a non-zero open-loop output impedance. Typically due
to the output-transistor's falling beta, Zout_OL actually goes up
with frequency, although a single say 200-ohm resistor can usually
serve to model this. Despite these efforts, the finished opamp will
still be capable of unrealistically-high slew rates (limit the VCCS
input voltage to fix this), kilovolt output voltages (limit the
VCCS-VCVS node voltages to fix this), and other weaknesses (no input
modeling, etc.). I suppose it can be useful for some quick and dirty
work. Here's the basic idea, which shows it's not that appealing.
| VCVS VCCS VCVS
| ____ R ____ ____ Ro
| --|+ |--/\/--+---|+ |------+---| |--/\/\-- out
| | x1 | | | x1 | | | x1 |
| --|____|--+--- | --|____|--+-- | --|____|--- gnd
| | | | | C
| gnd +--|>|--, gnd +---||---+--- gnd
| +--|<|--+ +--/\/\--'
| '-/\/\--+ | R
| 1000R | +--|>|-- +10V
| gnd '--|<|-- -10V
Thanks,
- Win
whill_at_picovolt-dot-com
Winfield Hill
>
> Except you can't model an opamp with a VCVS, instead you'll need a
> VCCS followed by a VCVS, and with a capacitor to ground at the node,
> VCCS and VCVS gains are unity, then a C = 1 / 2pi fT = 0.16uF cap
> will set the dominant pole for a f_T = 1MHz perfect-opamp bandwidth.
> Adding a 100k resistor will set a maximum gain of 100dB at DC. But
> real opamps have many other serious defects, like a second pole not
> too far above the f_T frequency (which reduces feedback-loop phase
> margins), and a non-zero open-loop output impedance. Typically due
> to the output-transistor's falling beta, Zout_OL actually goes up
> with frequency, although a single say 200-ohm resistor can usually
> serve to model this. Despite these efforts, the finished opamp will
> still be capable of unrealistically-high slew rates (limit the VCCS
> input voltage to fix this), kilovolt output voltages (limit the
> VCCS-VCVS node voltages to fix this), and other weaknesses (no input
> modeling, etc.). I suppose it can be useful for some quick and dirty
> work. Here's the basic idea, which shows it's not that appealing.
| VCVS VCCS VCVS
| ____ R ____ ____ Ro
| --|+ |--/\/--+---|+ |------+---| |--/\/\-- out
| | x1 | | | x1 | | | x1 |
| --|____|--+--- | --|____|--+-- | --|____|--- gnd
| | | | | C
| gnd +--|>|--, gnd +---||---+--- gnd
| +--|<|--+ +--/\/\--'
| +--||---+ C2 | R
| '-/\/\--+ +--|>|-- +10V
| 1000R | '--|<|-- -10V
| gnd
I added the (forgotten) cap required for the opamp's 2nd pole.
Thanks,
- Win
whill_at_picovolt-dot-com
Jim Thompson
<Winfiel...@newsguy.com> wrote:
Both Mike E (of LTSpice fame) and myself have posted behavioral OpAmp
models that have the right phase and gain roll-off, as well as swing
limiting.
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice:(480)460-2350 | |
| E-mail Address at Website Fax:(480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
Leon Heller
Paul Burridge wrote:
> Hi,
>
> Just been flicking through the fairly generous range of models in the
> LT op-amp library and have the following questions,,,
> Which LT device model is most similar to the LM324?
> Which LT device model is most similar to the 741?
LT doesn't make anything remotely like the LM324 and 741. Why not find
LM324 and 741 models and use them in LT Spice, if you must ise obsolete
devices. I've got them with the Pulsonix (SIMetrix) SPICE that I use.
> When an LT device is described as "35Mhz" or "90Mhz" or whatever, does
> that means it's good for anything up to that frequency or only good
> for a narrow band around that figure?
It's the max. frequency. You usually get a graph of voltage gain against
frequency in the data.
> Finally, do none of the models 'care' about whether they're single or
> dual supply?
> Finally, finally, isn't there a model or set of models which you don't
> have to connect up to a supply at all? If you're not going to generate
> a board from the schematic, it just seems a nuisance to have to 'power
> them up.'
SPICE models of active devices usually require supplies. How would the
simulation work, otherwise.
Leon
--
Leon Heller, G1HSM
Email: aqz...@dsl.pipex.com
My low-cost Philips LPC210x ARM development system:
http://www.geocities.com/leon_heller/lpc2104.html
Stefan Heinzmann
> Paul Burridge wrote:
>
>> Hi,
>>
>> Just been flicking through the fairly generous range of models in the
>> LT op-amp library and have the following questions,,,
>> Which LT device model is most similar to the LM324?
>> Which LT device model is most similar to the 741?
>
>
> LT doesn't make anything remotely like the LM324 and 741. Why not find
> LM324 and 741 models and use them in LT Spice, if you must ise obsolete
> devices. I've got them with the Pulsonix (SIMetrix) SPICE that I use.
>
>
>> When an LT device is described as "35Mhz" or "90Mhz" or whatever, does
>> that means it's good for anything up to that frequency or only good
>> for a narrow band around that figure?
>
>
> It's the max. frequency. You usually get a graph of voltage gain against
> frequency in the data.
>
>> Finally, do none of the models 'care' about whether they're single or
>> dual supply?
>> Finally, finally, isn't there a model or set of models which you don't
>> have to connect up to a supply at all? If you're not going to generate
>> a board from the schematic, it just seems a nuisance to have to 'power
>> them up.'
>
>
> SPICE models of active devices usually require supplies. How would the
> simulation work, otherwise.
Which doesn't mean that they accurately model the current drawn through
the supply lines. I came across models that don't seem to take the
current through the output into account. Which leaves you with a mostly
constant quiescent current. Not very useful.
--
Cheers
Stefan
Jim Thompson
wrote:
[snip]
>SPICE models of active devices usually require supplies. How would the
>simulation work, otherwise.
>
>Leon
Since they are subcircuits you just specify a output swing (as a
parameter).
Genome
"Paul Burridge" <p...@osiris1.notthisbit.co.uk> wrote in message
news:j4on10dobsjcha8ab...@4ax.com...
Like you say, you're opinion is worth exactly what I paid for it. When I
reach a pennies worth of download time I'll consider taking you to court.
In the meantime..... you, and others who wibble about with spice would do
well to realise that the level at which you are capable of using it does not
give you the right or ability to start putting 'real' models in. Especially
when you ask the questions you ask.
Keep it simple...... and learn some book stuff..... You know? AoE.
Ignoring your initial dribble you might try a voltage controlled current
source with a parallel RC network on its output followed by a voltage
controlled voltage source with a series resistor.
Ain't I kind?
DNA
Robert C Monsen
"Paul Burridge" <p...@osiris1.notthisbit.co.uk> wrote in message
news:j4on10dobsjcha8ab...@4ax.com...
Here are the models from the freeware version of CircuitMaker (based on
Berkeley SPICE3f5/XSpice):
*Sngl GenPurpose OpAmp pkg:DIP8 3,2,7,4,6
.SUBCKT XUA741 1 2 3 4 5
C1 11 12 4.664E-12
C2 6 7 20E-12
DC 5 53 DX
DE 54 5 DX
DLP 90 91 DX
DLN 92 90 DX
DP 4 3 DX
BGND 99 0 V=V(3)*.5 + V(4)*.5
BB 7 99 I=I(VB)*10.61E6 - I(VC)*10E6 + I(VE)*10E6 +
+ I(VLP)*10E6 - I(VLN)*10E6
GA 6 0 11 12 137.7E-6
GCM 0 6 10 99 2.574E-9
IEE 10 4 DC 10.16E-6
HLIM 90 0 VLIM 1K
Q1 11 2 13 QX
Q2 12 1 14 QX
R2 6 9 100E3
RC1 3 11 7.957E3
RC2 3 12 7.957E3
RE1 13 10 2.74E3
RE2 14 10 2.74E3
REE 10 99 19.69E6
RO1 8 5 150
RO2 7 99 150
RP 3 4 18.11E3
VB 9 0 DC 0
VC 3 53 DC 2.6
VE 54 4 DC 2.6
VLIM 7 8 DC 0
VLP 91 0 DC 25
VLN 0 92 DC 25
.MODEL DX D(IS=800E-18)
.MODEL QX NPN(IS=800E-18 BF=62.5)
.ENDS XUA741
and
*Quad LoPwr SnglSup OpAmp pkg:DIP14
(A:3,2,4,11,1)(B:5,6,4,11,7)(C:10,9,4,11,8)(D:12,13,4,11,14)
.SUBCKT XLM324 1 2 3 4 5
C1 11 12 5.544E-12
C2 6 7 20E-12
DC 5 53 DX
DE 54 5 DX
DLP 90 91 DX
DLN 92 90 DX
DP 4 3 DX
BGND 99 0 V=V(3)*.5 + V(4)*.5
BB 7 99 I=I(VB)*15.91E6 - I(VC)*20E6 + I(VE)*20E6 +
+ I(VLP)*20E6 - I(VLN)*20E6
GA 6 0 11 12 125.7E-6
GCM 0 6 10 99 7.067E-9
IEE 3 10 DC 10.04E-6
HLIM 90 0 VLIM 1K
Q1 11 2 13 QX
Q2 12 1 14 QX
R2 6 9 100E3
RC1 4 11 7.957E3
RC2 4 12 7.957E3
RE1 13 10 2.773E3
RE2 14 10 2.773E3
REE 10 99 19.92E6
RO1 8 5 50
RO2 7 99 50
RP 3 4 30.31E3
VB 9 0 DC 0
VC 3 53 DC 2.1
VE 54 4 DC .6
VLIM 7 8 DC 0
VLP 91 0 DC 40
VLN 0 92 DC 40
.MODEL DX D(IS=800E-18)
.MODEL QX PNP(IS=800E-18 BF=250)
.ENDS XLM324
Jim Thompson
<Winfiel...@newsguy.com> wrote:
>Winfield Hill wrote...
>>
[snip]
>
>| VCVS VCCS VCVS
>| ____ R ____ ____ Ro
>| --|+ |--/\/--+---|+ |------+---| |--/\/\-- out
>| | x1 | | | x1 | | | x1 |
>| --|____|--+--- | --|____|--+-- | --|____|--- gnd
>| | | | | C
>| gnd +--|>|--, gnd +---||---+--- gnd
>| +--|<|--+ +--/\/\--'
>| +--||---+ C2 | R
>| '-/\/\--+ +--|>|-- +10V
>| 1000R | '--|<|-- -10V
>| gnd
>
> I added the (forgotten) cap required for the opamp's 2nd pole.
>
> Thanks,
> - Win
>
> whill_at_picovolt-dot-com
Win, In the real OpAmp where *is* that "2nd pole" ?:-)
(In a real OpAmp *usually* the 2nd pole is *way* out there due to
"pole-splitting" compensation. The excess phase is actually due to
*delay*.)
Mike Engelhardt
> Just been flicking through the fairly generous range of models
> in the LT op-amp library and have the following questions,,,
> Which LT device model is most similar to the LM324?
> Which LT device model is most similar to the 741?
> When an LT device is described as "35Mhz" or "90Mhz" or whatever,
> does that means it's good for anything up to that frequency or
> only good for a narrow band around that figure?
The easiest thing to do is use the symbol "1pole" and set the
open loop voltage gain(Avol) and GBW to that of the target
opamp's DC gain and GBW. There's been threads here about
universal opamp models. You can of course also get a 3rd
party LM324 model and run that in LTspice. See the help
doc's for further info on how to do that.
> Finally, finally, isn't there a model or set of models which
> you don't have to connect up to a supply at all? If you're not
> going to generate a board from the schematic, it just seems a
> nuisance to have to 'power them up.'
There is a symbol "opamp" that requires no supplies, but it isn't
setup to automatically include the subcircuit definition so just
a SPICE directive ".lib opamp.sub" to the schematic.
--Mike
Paul Burridge
wrote:
>Ignoring your initial dribble you might try a voltage controlled current
>source with a parallel RC network on its output followed by a voltage
>controlled voltage source with a series resistor.
>
>Ain't I kind?
More of a "hopeless old soak" I'd have said.
Paul Burridge
<rcsu...@comcast.net> wrote:
>Here are the models from the freeware version of CircuitMaker (based on
>Berkeley SPICE3f5/XSpice
[snipped]
Thanks, Robert.
Winfield Hill
>
> Winfield Hill wrote:
>
>>| VCVS VCCS VCVS
>>| ____ R ____ ____ Ro
>>| --|+ |--/\/--+---|+ |------+---| |--/\/\-- out
>>| | x1 | | | x1 | | | x1 |
>>| --|____|--+--- | --|____|--+-- | --|____|--- gnd
>>| | | | | C
>>| gnd +--|>|--, gnd +---||---+--- gnd
>>| +--|<|--+ +--/\/\--'
>>| +--||---+ C2 | R
>>| '-/\/\--+ +--|>|-- +10V
>>| 1000R | '--|<|-- -10V
>>| gnd
>>
>> I added the (forgotten) cap required for the opamp's 2nd pole.
>
>
> (In a real OpAmp *usually* the 2nd pole is *way* out there due to
> "pole-splitting" compensation. The excess phase is actually due to
> *delay*.)
Poles create delay. As we emphasize in our book to those interested,
serious phase shift starts occurring at frequencies long before the
pole frequency, where it has already reached 45 degrees. So even if
the 2nd pole is as far out as 10x f_T (usually it's much much less)
it had better be included in the model. Furthermore, in most cases
the third pole should be included as well for better accuracy. And
output-stage Zout should be well modeled, to deal with the loads the
opamp drives.
The phase shift I referred to above *is* signal delay. I hear you've
made a special point in the past about delay modeling. I haven't
taken the time to completely understand it (have you explained your
ideas?), but in the end don't we get equivalent results from the time
domain and the frequency domain? At least for oamps with f_T below
30MHz? So it appears to me, both by the usual opamp theories and by
the measurements I've made on real-world parts with modeling to match.
(I've not taken these measurements on the new breeds of 50MHz+ opamps.
In this region pure time-delay elements in the models may make sense.)
Thanks,
- Win
whill_at_picovolt-dot-com
Joe
Hi Paul,
Here's my model for the LM741 in LTSPICE. You should save it as a .mod file.
I used the LT1013 as the .asy file and just renamed it. Helmut described in
great detail how to do this in a thread subject of 'Need help with LTSPICE
library' dated Aug 14, 2003 on this ng. I also have a model of the LM324
opamp below. Its the same, a .MOD file. I got them both from National Semi.
There's a yahoo group for LTSPICE as well. You should check it out.
Have fun with LTSPICE.
Joe
*//////////////////////////////////////////////////////////////////////
* (C) National Semiconductor, Inc.
* Models developed and under copyright by:
* National Semiconductor, Inc.
*/////////////////////////////////////////////////////////////////////
* Legal Notice: This material is intended for free software support.
* The file may be copied, and distributed; however, reselling the
* material is illegal
*////////////////////////////////////////////////////////////////////
* For ordering or technical information on these models, contact:
* National Semiconductor's Customer Response Center
* 7:00 A.M.--7:00 P.M. U.S. Central Time
* (800) 272-9959
* For Applications support, contact the Internet address:
* amps...@galaxy.nsc.com
*//////////////////////////////////////////////////////////
*LM741 OPERATIONAL AMPLIFIER MACRO-MODEL
*//////////////////////////////////////////////////////////
*
* connections: non-inverting input
* | inverting input
* | | positive power supply
* | | | negative power supply
* | | | | output
* | | | | |
* | | | | |
.SUBCKT LM741/NS 1 2 99 50 28
*
*Features:
*Improved performance over industry standards
*Plug-in replacement for LM709,LM201,MC1439,748
*Input and output overload protection
*
****************INPUT STAGE**************
*
IOS 2 1 20N
*^Input offset current
R1 1 3 250K
R2 3 2 250K
I1 4 50 100U
R3 5 99 517
R4 6 99 517
Q1 5 2 4 QX
Q2 6 7 4 QX
*Fp2=2.55 MHz
C4 5 6 60.3614P
*
***********COMMON MODE EFFECT***********
*
I2 99 50 1.6MA
*^Quiescent supply current
EOS 7 1 POLY(1) 16 49 1E-3 1
*Input offset voltage.^
R8 99 49 40K
R9 49 50 40K
*
*********OUTPUT VOLTAGE LIMITING********
V2 99 8 1.63
D1 9 8 DX
D2 10 9 DX
V3 10 50 1.63
*
**************SECOND STAGE**************
*
EH 99 98 99 49 1
G1 98 9 5 6 2.1E-3
*Fp1=5 Hz
R5 98 9 95.493MEG
C3 98 9 333.33P
*
***************POLE STAGE***************
*
*Fp=30 MHz
G3 98 15 9 49 1E-6
R12 98 15 1MEG
C5 98 15 5.3052E-15
*
*********COMMON-MODE ZERO STAGE*********
*
*Fpcm=300 Hz
G4 98 16 3 49 3.1623E-8
L2 98 17 530.5M
R13 17 16 1K
*
**************OUTPUT STAGE**************
*
F6 50 99 POLY(1) V6 450U 1
E1 99 23 99 15 1
R16 24 23 25
D5 26 24 DX
V6 26 22 0.65V
R17 23 25 25
D6 25 27 DX
V7 22 27 0.65V
V5 22 21 0.18V
D4 21 15 DX
V4 20 22 0.18V
D3 15 20 DX
L3 22 28 100P
RL3 22 28 100K
*
***************MODELS USED**************
*
.MODEL DX D(IS=1E-15)
.MODEL QX NPN(BF=625)
*
.ENDS
*$
*//////////////////////////////////////////////////////////////////////
* (C) National Semiconductor, Inc.
* Models developed and under copyright by:
* National Semiconductor, Inc.
*/////////////////////////////////////////////////////////////////////
* Legal Notice: This material is intended for free software support.
* The file may be copied, and distributed; however, reselling the
* material is illegal
*////////////////////////////////////////////////////////////////////
* For ordering or technical information on these models, contact:
* National Semiconductor's Customer Response Center
* 7:00 A.M.--7:00 P.M. U.S. Central Time
* (800) 272-9959
* For Applications support, contact the Internet address:
* amps...@galaxy.nsc.com
*//////////////////////////////////////////////////////////
*LM324 Low Power Quad OPERATIONAL AMPLIFIER MACRO-MODEL
*//////////////////////////////////////////////////////////
*
* connections: non-inverting input
* | inverting input
* | | positive power supply
* | | | negative power supply
* | | | | output
* | | | | |
* | | | | |
.SUBCKT LM324/NS 1 2 99 50 28
*
*Features:
*Eliminates need for dual supplies
*Large DC voltage gain = 100dB
*High bandwidth = 1MHz
*Low input offset voltage = 2mV
*Wide supply range = +-1.5V to +-16V
*
*NOTE: Model is for single device only and simulated
* supply current is 1/4 of total device current.
* Output crossover distortion with dual supplies
* is not modeled.
*
****************INPUT STAGE**************
*
IOS 2 1 5N
*^Input offset current
R1 1 3 500K
R2 3 2 500K
I1 99 4 100U
R3 5 50 517
R4 6 50 517
Q1 5 2 4 QX
Q2 6 7 4 QX
*Fp2=1.2 MHz
C4 5 6 128.27P
*
***********COMMON MODE EFFECT***********
*
I2 99 50 75U
*^Quiescent supply current
EOS 7 1 POLY(1) 16 49 2E-3 1
*Input offset voltage.^
R8 99 49 60K
R9 49 50 60K
*
*********OUTPUT VOLTAGE LIMITING********
V2 99 8 1.63
D1 9 8 DX
D2 10 9 DX
V3 10 50 .635
*
**************SECOND STAGE**************
*
EH 99 98 99 49 1
G1 98 9 POLY(1) 5 6 0 9.8772E-4 0 .3459
*Fp1=7.86 Hz
R5 98 9 101.2433MEG
C3 98 9 200P
*
***************POLE STAGE***************
*
*Fp=2 MHz
G3 98 15 9 49 1E-6
R12 98 15 1MEG
C5 98 15 7.9577E-14
*
*********COMMON-MODE ZERO STAGE*********
*
*Fpcm=10 KHz
G4 98 16 3 49 5.6234E-8
L2 98 17 15.9M
R13 17 16 1K
*
**************OUTPUT STAGE**************
*
F6 50 99 POLY(1) V6 300U 1
E1 99 23 99 15 1
R16 24 23 17.5
D5 26 24 DX
V6 26 22 .63V
R17 23 25 17.5
D6 25 27 DX
V7 22 27 .63V
V5 22 21 0.27V
D4 21 15 DX
V4 20 22 0.27V
D3 15 20 DX
L3 22 28 500P
RL3 22 28 100K
*
***************MODELS USED**************
*
.MODEL DX D(IS=1E-15)
.MODEL QX PNP(BF=1.111E3)
*
.ENDS
*$
Fred Bloggs
The total phase angle is the sum of phase angles so that using ATN(F/Fp)
approximating to F/Fp for F<<Fp, the higher frequency poles do introduce
a phase delay linear with frequency which appears as pure time delay
1/(2.pi.Fp).
Winfield Hill
>
> The total phase angle is the sum of phase angles so that using ATN(F/Fp)
> approximating to F/Fp for F<<Fp, the higher frequency poles do introduce
> a phase delay linear with frequency which appears as pure time delay
> 1/(2.pi.Fp).
I wouldn't call it a pure time delay, in that a frequency rolloff
is included in the bargain. A transmission line would be a "pure"
time delay. Electronic circuits can have more time delay than is
predicted by the observable poles. We'll wait and see what Jim is
referring to, what he has in mind.
Thanks,
- Win
whill_at_picovolt-dot-com
Paul Burridge
<nuisance...@nospamearthlink.net> wrote:
>Hi Paul,
>
>Here's my model for the LM741 in LTSPICE. You should save it as a .mod file.
>I used the LT1013 as the .asy file and just renamed it. Helmut described in
>great detail how to do this in a thread subject of 'Need help with LTSPICE
>library' dated Aug 14, 2003 on this ng. I also have a model of the LM324
>opamp below. Its the same, a .MOD file. I got them both from National Semi.
Thanks, Joe. Those models should be very useful.
>There's a yahoo group for LTSPICE as well. You should check it out.
Will do.
Jim Thompson
<Winfiel...@newsguy.com> wrote:
The problem with pole modeling is that it fails to match the SMOOTH
GAIN curve plus the RAPID PHASE change that most OpAmps exhibit near
0dB.
A pure delay model, such as an ALL-PASS, models this quite accurately.
To get equivalent accuracy you would need a MANY-pole distributed
model.
Jim Thompson
<Winfiel...@newsguy.com> wrote:
Proponents of pole modeling OpAmp performance are invited to try their
hand on a real OpAmp posted at........
Newsgroups: alt.binaries.schematics.electronic
Subject: Modeling OpAmp Gain and Phase - ExampleOpAmpGainPhase.gif
Message-ID: <ov7q109e6fe90hf7g...@4ax.com>
Winfield Hill
>
> Proponents of pole modeling OpAmp performance are invited
> to try their hand on a real OpAmp posted at........
>
> Newsgroups: alt.binaries.schematics.electronic
> Subject: Modeling OpAmp Gain and Phase - ExampleOpAmpGainPhase.gif
> Message-ID: <ov7q109e6fe90hf7g...@4ax.com>
Which real opamp is that?
Thanks,
- Win
whill_at_picovolt-dot-com
Jim Thompson
<Winfiel...@newsguy.com> wrote:
I *guarantee* that it's a real OpAmp that you can purchase. I won't
divulge the part number until you can demonstrate modeling it with
poles. I'll model it my way as well and we'll compare results.
Those of you who recognize the data sheet... ssssshush... this is a
test of modeling concepts... let nature take its course ;-)
Winfield Hill
>
> Winfield Hill wrote:
>
> I won't divulge the part number until you can demonstrate
> modeling it with poles.
Sorry, I won't take the time to do that until I've measured
one and verified the curve is real. It does seem a pretty
gross opamp. What is it, a uA702? Or a micropower opamp?
Thanks,
- Win
whill_at_picovolt-dot-com
Jim Thompson
<Winfiel...@newsguy.com> wrote:
It is a recent vintage OpAmp (LT1880). What you mean to say is you
CAN'T model it with just poles. That's alright Win, I already knew
you couldn't ;-)
Syd Rumpo
<p...@osiris1.notthisbit.co.uk> writes
>On Sat, 31 Jan 2004 19:44:25 -0000, "Genome" <gen...@nothere.com>
>wrote:
>
>>Ignoring your initial dribble you might try a voltage controlled current
>>source with a parallel RC network on its output followed by a voltage
>>controlled voltage source with a series resistor.
>>
>>Ain't I kind?
>
>More of a "hopeless old soak" I'd have said.
Neither hopeless nor old. If you wish to learn, listen to your betters
and accept that they are as they are. That includes Auntie: she is
watching you.
--
Syd
Jim Thompson
Roy McCammon
I haven't been following the discussion closely,
but when looking at the zero crossing delay
of sinewaves from input to output of a low pass
filter and well below the cutoff frequency,
I have observed what looked like a fixed delay.
Paul Burridge
<thegr...@example.com> wrote:
>On Sun, 1 Feb 2004 22:01:37 +0000, Syd Rumpo <sydr...@clarat.co.uk>
>wrote:
>
>>In message <96do10ldm9okk8a43...@4ax.com>, Paul Burridge
>><p...@osiris1.notthisbit.co.uk> writes
>>>On Sat, 31 Jan 2004 19:44:25 -0000, "Genome" <gen...@nothere.com>
>>>wrote:
>>>
>>>>Ignoring your initial dribble you might try a voltage controlled current
>>>>source with a parallel RC network on its output followed by a voltage
>>>>controlled voltage source with a series resistor.
>>>>
>>>>Ain't I kind?
>>>
>>>More of a "hopeless old soak" I'd have said.
>>
>>Neither hopeless nor old.
Obnoxious dypsomaniac, then.
> If you wish to learn, listen to your betters
>>and accept that they are as they are.
And who the hell are *you* to lecture me? I respect people who are
informative and polite. I tolerate people who are informative but
impolite; I'll treat people with neither knowledge nor manners with
the contempt they deserve. That includes you.
>That includes Auntie: she is
>>watching you.
>
>Good one, Syd!
Says someone who doesn't even know who or what "Auntie" is. Best stick
with what you know, Jim.
Jim Thompson
<Winfiel...@newsguy.com> wrote:
See...
Subject: Re: Modeling OpAmp Gain and Phase - ExampleOpAmpGainPhase.gif
- LT1880-SIM.pdf - LT1880-SIM.pdf
Message-ID: <hs1r10p9ono3cho59...@4ax.com>
Jim Thompson
Poor Paul, You still don't understand that Genome has more mental
capacity in his pinky finger than you have if you surround yourself
with all your friends ;-)
Winfield Hill
Absolutely. But it's not the "pure" all-pass time delay Jim has
in mind, since it has no associated single pole.
Thanks,
- Win
whill_at_picovolt-dot-com
Winfield Hill
>
> It is a recent vintage OpAmp (LT1880). What you mean to say is you
> CAN'T model it with just poles. That's alright Win, I already knew
> you couldn't ;-)
Just poles? Who said anything about *just* poles? Detailed high-
frequency modeling of an opamp usually involves not only extra
poles but zeros as well; I certainly have never argued otherwise.
But I thought your point was that the model was simplified or more
accurate with pure time delays instead of pole-zero combinations.
Thanks,
- Win
whill_at_picovolt-dot-com
Syd Rumpo
<p...@osiris1.notthisbit.co.uk> writes
>And who the hell are *you* to lecture me?
I don't have the inclination to engage in a battle of wits with an
unarmed man. This correspondence is now closed.
--
Syd
Bill Sloman
> On Sun, 01 Feb 2004 15:20:16 -0700, Jim Thompson
> <thegr...@example.com> wrote:
>
> >On Sun, 1 Feb 2004 22:01:37 +0000, Syd Rumpo <sydr...@clarat.co.uk>
> >wrote:
> >
> >>In message <96do10ldm9okk8a43...@4ax.com>, Paul Burridge
> >><p...@osiris1.notthisbit.co.uk> writes
> >>>On Sat, 31 Jan 2004 19:44:25 -0000, "Genome" <gen...@nothere.com>
<snip>
> >That includes Auntie: she is
> >>watching you.
> >
> >Good one, Syd!
>
> Says someone who doesn't even know who or what "Auntie" is. Best stick
> with what you know, Jim.
Paul, if you think that "Auntie" is "Licensed at public expense to
spread lies"
you don't know who or what "Auntie" is either. You'd better give up
reading right-wing newspapers for a while ...
-------
Bill Sloman, Nijmegen
Jim Thompson
<Winfiel...@newsguy.com> wrote:
You can't get the rapid phase change without an all-pass-like
structure... of course that means there's a zero in there ;-)
Who was it that said...
"Poles create delay. As we emphasize in our book to those interested,
serious phase shift starts occurring at frequencies long before the
pole frequency, where it has already reached 45 degrees."
Poles not only create delay they also cause roll-off... delay
*doesn't*.
Chris Carlen
> On Sun, 01 Feb 2004 23:41:13 +0000, Paul Burridge
> <p...@osiris1.notthisbit.co.uk> wrote:
>>And who the hell are *you* to lecture me? I respect people who are
>>informative and polite. I tolerate people who are informative but
>>impolite; I'll treat people with neither knowledge nor manners with
>>the contempt they deserve. That includes you.
>>
>>
>>>That includes Auntie: she is
>>>
>>>>watching you.
>>>
>>>Good one, Syd!
>>
>>Says someone who doesn't even know who or what "Auntie" is. Best stick
>>with what you know, Jim.
>
>
> Poor Paul, You still don't understand that Genome has more mental
> capacity in his pinky finger than you have if you surround yourself
> with all your friends ;-)
>
> ...Jim Thompson
I have to agree with Paul on this instance. His point is that he'd
rather not be denigrated whether or not someone else has superior
knowledge. To assume that anyone including Paul *needs* help from
someone who is obnoxious is unfounded. There are plently of
knowledgeable *and* polite folks around to pick up the slack.
In my case, I come here to s.e.d to accelerate the finding of solutions
to problems. That I *need* anyone's help is usually false, as I can
figure things out for myself if I want. I have yet to fail at designing
*anything* I have set out to do, whether or not I sought help and
whether not I listened to help that was offered. Thus, I will promptly
ignore anyone who doesn't treat me with respect. Where I work, I am
respected. I don't have to go around groveling at the feet of giants to
get mentorship. Neither should Paul.
Good day!
--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
crc...@sandia.gov
Jim Thompson
<crc...@BOGUS.sandia.gov> wrote:
>Jim Thompson wrote:
[snip]
>> Poor Paul, You still don't understand that Genome has more mental
>> capacity in his pinky finger than you have if you surround yourself
>> with all your friends ;-)
>>
>> ...Jim Thompson
>
>
>I have to agree with Paul on this instance. His point is that he'd
>rather not be denigrated whether or not someone else has superior
>knowledge. To assume that anyone including Paul *needs* help from
>someone who is obnoxious is unfounded. There are plently of
>knowledgeable *and* polite folks around to pick up the slack.
>
>In my case, I come here to s.e.d to accelerate the finding of solutions
>to problems. That I *need* anyone's help is usually false, as I can
>figure things out for myself if I want. I have yet to fail at designing
>*anything* I have set out to do, whether or not I sought help and
>whether not I listened to help that was offered. Thus, I will promptly
>ignore anyone who doesn't treat me with respect. Where I work, I am
>respected. I don't have to go around groveling at the feet of giants to
>get mentorship. Neither should Paul.
>
>
>Good day!
I pick at Paul because he's also rude. If he wants to be treated with
respect he needs to act accordingly.
Winfield Hill
>
> "Poles create delay. As we emphasize in our book to those interested,
> serious phase shift starts occurring at frequencies long before the
> pole frequency, where it has already reached 45 degrees."
I said that, but as you surely noticed, in the context of
arguing for at least one extra pole in a super-simple opamp
"model," against the suggestion of no poles at all, not even
a dominate pole! ("You can model an ideal amplifier using a
voltage controlled voltage source.") So, yes, I argued for
two poles, which is a big step up in that arena, and actually
quite helpful in the model's quality, despite its simplicity.
I certainly wasn't making any argument for poles only, more
and more poles, pile on the poles, no zeros, or whatever, ...
sheesh! But you knew that. You're just yanking my chain.
:>)
Thanks,
- Win
whill_at_picovolt-dot-com
John Woodgate
<crc...@BOGUS.sandia.gov> wrote (in <bvm5d...@enews2.newsguy.com>)
about 'Op-amp models in LT Spice', on Mon, 2 Feb 2004:
>I have to agree with Paul on this instance. His point is that he'd
>rather not be denigrated whether or not someone else has superior
>knowledge.
of course.
>To assume that anyone including Paul *needs* help from
>someone who is obnoxious is unfounded. There are plently of
>knowledgeable *and* polite folks around to pick up the slack.
Well, not all are of the quality of Fred Bloggs, or Genome when he's on
this planet.
>
>In my case, I come here to s.e.d to accelerate the finding of solutions
>to problems. That I *need* anyone's help is usually false, as I can
>figure things out for myself if I want. I have yet to fail at designing
>*anything* I have set out to do, whether or not I sought help and
>whether not I listened to help that was offered. Thus, I will promptly
>ignore anyone who doesn't treat me with respect.
That may be a counsel of perfection.
> Where I work, I am
>respected. I don't have to go around groveling at the feet of giants to
>get mentorship.
No-one expects you to grovel. But it is not logical to ignore good
advice because it comes with an irrelevant and distasteful burden.
> Neither should Paul.
Ah, well, there are special cases, who tend to get what they deserve.
--
Regards, John Woodgate, OOO - Own Opinions Only.
The good news is that nothing is compulsory.
The bad news is that everything is prohibited.
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk
Mike Engelhardt
> >> I *guarantee* that it's a real OpAmp that you can purchase.
> >> I won't divulge the part number until you can demonstrate
> >> modeling it with poles.
> >
> > Sorry, I won't take the time to do that until I've measured
> > one and verified the curve is real. It does seem a pretty
> > gross opamp. What is it, a uA702? Or a micropower opamp?
>
> It is a recent vintage OpAmp (LT1880). What you mean to say is you
> CAN'T model it with just poles. That's alright Win, I already knew
> you couldn't ;-)
That 1880 is a terrific example of an opamp well
discriped by a dominate pole plus about 50ns delay.
Dominate pole plus a delay is a great way to model
opamps and been around for along time. See e.g.,
page 264 of Modern Filter Design by M.S.Ghausi and
K.R.Laker from over 20 years ago. But to be fair to
the pole modeling crowd, one can model that delay
pretty well with three pi sections for 7 additional
poles, which do simulate quicker than many delay
schemes at least in the time domain. Here's a deck
that uses a look-up table, a dominate pole and a
xmission line, and a dominate pole with three pi
sections. Their outputs are A, B, and C
respectively.
--Mike
*
G1 0 N001 X 0 1
R1 N001 0 3Meg
C1 N001 0 .2u
V1 X 0 ac 1
E2 A 0 FREQ {-V(x)} .01,129,178, .02,129,176, .05,129,169,.1,129,159,
+ .3,126,131,.5,123,118, 1,118,105, 2,112,97,3.5,107,94 10,98,90
+ 10K,40,88, 20K,35,82, 40K,28,81, 100K,18,80, 200K,12,78, 400K,5,71,
+ 1Meg,-4,55, 2Meg,-12,40, 4Meg,-18,16, 10Meg,-28,-80
E1 N002 0 N001 0 1
T1 N002 0 B 0 Td=50n Z0=50
R2 B 0 50
L1 N003 C 1u
C2 C 0 125p
L2 N004 N003 1u
C3 N003 0 250p
L3 N005 N004 1u
C4 N004 0 250p
R3 C 0 63
C5 N005 0 125p
R4 N005 0 63
G2 0 N005 N006 0 0.03125
G3 0 N006 X 0 1
R5 N006 0 3Meg
C6 N006 0 .2u
.ac dec 100 .01 10Meg
.probe
.end
Jim Thompson
<Winfiel...@newsguy.com> wrote:
Now, Win, would I yank your chain ?:-)
Did someone say "You can model an ideal amplifier using a voltage
controlled voltage source"?
I vote for a single dominant pole model plus an all-pass to account
for excess phase.
Maybe an extra pole-zero pair to accommodate weirdies like the LT1880,
though I have suspicions that I can model that with an all-pass that's
slightly asymmetric... that would give a handle on excess phase *and*
that "zero bump" that can be seen in the LT1880 gain-bandwidth curve.
Jim Thompson
wrote:
>Jim,
[snip]
>That 1880 is a terrific example of an opamp well
>discriped by a dominate pole plus about 50ns delay.
>Dominate pole plus a delay is a great way to model
>opamps and been around for along time. See e.g.,
>page 264 of Modern Filter Design by M.S.Ghausi and
>K.R.Laker from over 20 years ago. But to be fair to
>the pole modeling crowd, one can model that delay
>pretty well with three pi sections for 7 additional
>poles, which do simulate quicker than many delay
>schemes at least in the time domain. Here's a deck
>that uses a look-up table, a dominate pole and a
>xmission line, and a dominate pole with three pi
>sections. Their outputs are A, B, and C
>respectively.
>
>--Mike
>
[snip]
That's why I use an all-pass as opposed to a transmission line... to
avoid long simulation times.
Stefan Heinzmann
[...]
> Did someone say "You can model an ideal amplifier using a voltage
> controlled voltage source"?
Yes. I did, lightheadedly, I confess. I have come to regret it.
--
Cheers
Stefan
Paul Burridge
<crc...@BOGUS.sandia.gov> wrote:
>I have to agree with Paul on this instance. His point is that he'd
>rather not be denigrated whether or not someone else has superior
>knowledge. To assume that anyone including Paul *needs* help from
>someone who is obnoxious is unfounded. There are plently of
>knowledgeable *and* polite folks around to pick up the slack.
>
>In my case, I come here to s.e.d to accelerate the finding of solutions
>to problems. That I *need* anyone's help is usually false, as I can
>figure things out for myself if I want. I have yet to fail at designing
>*anything* I have set out to do, whether or not I sought help and
>whether not I listened to help that was offered. Thus, I will promptly
>ignore anyone who doesn't treat me with respect. Where I work, I am
>respected. I don't have to go around groveling at the feet of giants to
>get mentorship. Neither should Paul.
Thanks, Chris. I'm grateful you went to the trouble of explaining the
situation clearly. Given the quality of the "contributors" present, I
simply couldn't be bothered.
Winfield Hill
>
> The LT1880 is a terrific example of an opamp well
> discriped by a dominate pole plus about 50ns delay.
The LT1880 is a rail-rail opamp with a more detailed
output-stage schematic than is usually made available
for this type of opamp. Generally we get a box with
some type of useless label - output-stage driver, etc.
Mike, can you ask someone there and tell us if the
data-sheet drawing is accurate, and does this opamp
design appear in the literature? Also, can you give
us any detailed guidance into the reasons why this
opamp seems to have a "pure" 50ns excess time delay
that extends well beyond the f_T frequency?
Thanks,
- Win
whill_at_picovolt-dot-com
Mike Engelhardt
> > That 1880 is a terrific example of an opamp well
> > discriped by a dominate pole plus about 50ns delay.
> > Dominate pole plus a delay is a great way to model
> > opamps and been around for along time. See e.g.,
> > page 264 of Modern Filter Design by M.S.Ghausi and
> > K.R.Laker from over 20 years ago.
> The LT1880 is a rail-rail opamp with a more detailed
> output-stage schematic than is usually made available
> for this type of opamp. Generally we get a box with
> some type of useless label - output-stage driver, etc.
>
> Mike, can you ask someone there and tell us if the
> data-sheet drawing is accurate, and does this opamp
> design appear in the literature?
The guy who designed the LT1880 tells me that the plots
on page 6 of data datasheet labeled "Gain vs Frequency"
and "Gain and Phase vs Frequency" are indeed accurate
measured data from a device. As to the schematic on
page 11, he says its a very simplied schematic. The
points he's trying to make with it are that it has a lot
of gain in the 1st stage, input bias current cancellation,
and multiple output paths for current capability near the
rails. Myself, I won't try to infer too much about the AC
response of the device from that schematic.
> Also, can you give us any detailed guidance into the
> reasons why this opamp seems to have a "pure" 50ns
> excess time delay that extends well beyond the f_T
> frequency?
Well, it isn't exactly a "pure" delay. You can see
a kink in there. The SPICE deck I had posted glossed
over that. There's really nothing special about this
part that a dominate pole plus delay models it better
than a dominate pole plus a couple lumped constants.
Filter people have known for a long time that's a
better way to model higher order effects over the
dominate pole of a generic opamp. Check out that
reference. Or put a step into a opamp amplifier
with a resistive feedback network. You see what
qualitatively looks like a delay and then the signal.
--Mike
Winfield Hill
>
> Win,
>
> The guy who designed the LT1880 tells me that the plots
> on page 6 of data datasheet labeled "Gain vs Frequency"
> and "Gain and Phase vs Frequency" are indeed accurate
> measured data from a device. As to the schematic on
Thanks very much, Mike.
Thanks,
- Win
whill_at_picovolt-dot-com