Reply by Kevin Aylward●February 20, 20182018-02-20
"Jim Thompson" wrote in message
news:1utj8d1qh9ebbt2a47me85vvebthmnipch@4ax.com...
On Sun, 18 Feb 2018 15:53:08 -0600, "Tim Williams"
<tiwill@seventransistorlabs.com> wrote:
>"Gerhard Hoffmann" <gerhard@hoffmann-hochfrequenz.de> wrote in message
>news:feu1tpFoar8U1@mid.individual.net...
>> Using pspice and then complaining that LTspice does not conform
>> to the "standards"!
>>
>> May I remind you that we had that problem with Pspice models
>> that could not run on anything else for 20 years or so?
>>
>
>>Recently poked at a TI SMPS controller model that's a complete clusterfuck
>>of discontinuous IF's, behavioral G sources and as far as I can tell, very
>>little if any modeling of actual pin characteristics.
>
>>It's ostensibly a PSPICE model, but I have a sneaking suspicion even
>>PSPICE
>>must have a hard time running it.
>
>>I fixed a bunch of those statements with continuous, 3f5 compatible
>>models,
>>but naturally it still doesn't work. Hard to say if it's because I goofed
>>a
>>substitution, overlooked more problem statements, or it's just altogether
>>wrong. All seem equally likely.
>
>Tim
>Bad use of "IF" statements isn't a peculiarity of PSpice, it's what
>happens when you allow PhD's just out of school to write models >:-}
On Sun, 18 Feb 2018 15:53:08 -0600, "Tim Williams"
<tiwill@seventransistorlabs.com> wrote:
>"Gerhard Hoffmann" <gerhard@hoffmann-hochfrequenz.de> wrote in message
>news:feu1tpFoar8U1@mid.individual.net...
>> Using pspice and then complaining that LTspice does not conform
>> to the "standards"!
>>
>> May I remind you that we had that problem with Pspice models
>> that could not run on anything else for 20 years or so?
>>
>
>Recently poked at a TI SMPS controller model that's a complete clusterfuck
>of discontinuous IF's, behavioral G sources and as far as I can tell, very
>little if any modeling of actual pin characteristics.
>
>It's ostensibly a PSPICE model, but I have a sneaking suspicion even PSPICE
>must have a hard time running it.
>
>I fixed a bunch of those statements with continuous, 3f5 compatible models,
>but naturally it still doesn't work. Hard to say if it's because I goofed a
>substitution, overlooked more problem statements, or it's just altogether
>wrong. All seem equally likely.
>
>Tim
Bad use of "IF" statements isn't a peculiarity of PSpice, it's what
happens when you allow PhD's just out of school to write models >:-}
...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| STV, Queen Creek, AZ 85142 Skype: skypeanalog | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
It's what you learn, after you know it all, that counts.
Reply by Jim Thompson●February 18, 20182018-02-18
On Sun, 18 Feb 2018 15:53:08 -0600, "Tim Williams"
<tiwill@seventransistorlabs.com> wrote:
>"Gerhard Hoffmann" <gerhard@hoffmann-hochfrequenz.de> wrote in message
>news:feu1tpFoar8U1@mid.individual.net...
>> Using pspice and then complaining that LTspice does not conform
>> to the "standards"!
>>
>> May I remind you that we had that problem with Pspice models
>> that could not run on anything else for 20 years or so?
>>
>
>Recently poked at a TI SMPS controller model that's a complete clusterfuck
>of discontinuous IF's, behavioral G sources and as far as I can tell, very
>little if any modeling of actual pin characteristics.
>
>It's ostensibly a PSPICE model, but I have a sneaking suspicion even PSPICE
>must have a hard time running it.
>
>I fixed a bunch of those statements with continuous, 3f5 compatible models,
>but naturally it still doesn't work. Hard to say if it's because I goofed a
>substitution, overlooked more problem statements, or it's just altogether
>wrong. All seem equally likely.
>
>Tim
Confucius say, "He who uses IF statements within a simulation shall
die by IF statements" >:-}
(*) Except for setting up preconditions that don't change during a
simulation run.
...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| STV, Queen Creek, AZ 85142 Skype: skypeanalog | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
It's what you learn, after you know it all, that counts.
Reply by Jim Thompson●February 18, 20182018-02-18
On Sun, 18 Feb 2018 20:19:20 +0100, Gerhard Hoffmann
<gerhard@hoffmann-hochfrequenz.de> wrote:
>Am 18.02.2018 um 20:01 schrieb Jim Thompson:
>
>>> Is it really so hard to write the word "noiseless" after the resistance
>>> value?
>>>
>>> <
>>> https://www.flickr.com/photos/137684711@N07/38533153940/in/dateposted-public/
>>> >
>>>
>>>
>>> Gerhard
>>
>> Try that in any other Spice. That's my beef. LTspice doesn't follow
>> convention. Thus its models won't play elsewhere.
>
>Using pspice and then complaining that LTspice does not conform
>to the "standards"!
>
>May I remind you that we had that problem with Pspice models
>that could not run on anything else for 20 years or so?
That would be the digital models... but no one's digital models are
cross compatible... including Cadence's, PSpice, LTspice, HSpice....
The digital models I've written use standard Spice (Berkeley) language
and work anywhere.
Analog models: I've been a user of PSpice user day one... I don't know
of an analog model that isn't cross-compatible.
...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| STV, Queen Creek, AZ 85142 Skype: skypeanalog | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
It's what you learn, after you know it all, that counts.
Reply by Jeroen Belleman●February 18, 20182018-02-18
On 18/02/18 19:45, Gerhard Hoffmann wrote:
> Am 17.02.2018 um 18:01 schrieb Jim Thompson:
>
>>> Version 4
>>> SHEET 1 880 680
>>> WIRE 128 -96 -64 -96
>>> WIRE 176 -96 128 -96
>> ]snip]
>>> TEXT 288 -144 Left 2 !;ac dec 100 1k 100meg
>>> TEXT 288 -88 Left 2 !.noise v(out) v3 dec 100 1k 100meg
>>> TEXT 248 -32 Left 2 ;G1 behaves like a noiseless 1k resistor
>>
>> In _many_ instances where a G-source is used as a resistor in LTspice,
>> LTspice can't find the operating point because Mikey turns off all
>> current sources during the .OP calculation.
>>
>> I've taken to modeling noiseless resistors with an E-source to
>> accommodate LTspice users... not as precise as a G-source :-(
>>
>> ...Jim Thompson
>>
> Is it really so hard to write the word "noiseless" after the resistance
> value?
>
> <
> https://www.flickr.com/photos/137684711@N07/38533153940/in/dateposted-public/
> >
Well, if I ever...!
Thanks!
Jeroen Belleman
Reply by Tim Williams●February 18, 20182018-02-18
"Gerhard Hoffmann" <gerhard@hoffmann-hochfrequenz.de> wrote in message
news:feu1tpFoar8U1@mid.individual.net...
> Using pspice and then complaining that LTspice does not conform
> to the "standards"!
>
> May I remind you that we had that problem with Pspice models
> that could not run on anything else for 20 years or so?
>
Recently poked at a TI SMPS controller model that's a complete clusterfuck
of discontinuous IF's, behavioral G sources and as far as I can tell, very
little if any modeling of actual pin characteristics.
It's ostensibly a PSPICE model, but I have a sneaking suspicion even PSPICE
must have a hard time running it.
I fixed a bunch of those statements with continuous, 3f5 compatible models,
but naturally it still doesn't work. Hard to say if it's because I goofed a
substitution, overlooked more problem statements, or it's just altogether
wrong. All seem equally likely.
Tim
--
Seven Transistor Labs, LLC
Electrical Engineering Consultation and Contract Design
Website: https://www.seventransistorlabs.com/
Reply by Steve Wilson●February 18, 20182018-02-18
Gerhard Hoffmann <gerhard@hoffmann-hochfrequenz.de> wrote:
> Am 17.02.2018 um 18:01 schrieb Jim Thompson:
>>> Version 4
>>> SHEET 1 880 680
>>> WIRE 128 -96 -64 -96
>>> WIRE 176 -96 128 -96
>> ]snip]
>>> TEXT 288 -144 Left 2 !;ac dec 100 1k 100meg
>>> TEXT 288 -88 Left 2 !.noise v(out) v3 dec 100 1k 100meg
>>> TEXT 248 -32 Left 2 ;G1 behaves like a noiseless 1k resistor
>> In _many_ instances where a G-source is used as a resistor in LTspice,
>> LTspice can't find the operating point because Mikey turns off all
>> current sources during the .OP calculation.
>>
>> I've taken to modeling noiseless resistors with an E-source to
>> accommodate LTspice users... not as precise as a G-source :-(
>> ...Jim Thompson
> Is it really so hard to write the word "noiseless" after the resistance
> value?
I have extended Jeroen's analysis to employ noiseless resistors or
conventional resistors in the emitter or collector. The noise analysis runs
fine with either choice.
This gives an excellent display of how the noise from the base appears in
the collector but not in the emitter.
Watch the wrap at the end.
Version 4
SHEET 1 880 680
WIRE 176 -48 -64 -48
WIRE 368 -48 176 -48
WIRE -64 -32 -64 -48
WIRE 176 0 176 -48
WIRE 176 0 128 0
WIRE 176 16 176 0
WIRE 128 32 128 0
WIRE -64 64 -64 48
WIRE 128 112 128 80
WIRE 176 112 176 96
WIRE 176 112 128 112
WIRE 288 112 176 112
WIRE 368 112 368 80
WIRE 368 112 288 112
WIRE 176 144 176 112
WIRE 48 192 -64 192
WIRE 112 192 48 192
WIRE -64 208 -64 192
WIRE 176 272 176 240
WIRE 240 272 176 272
WIRE 368 272 240 272
WIRE 448 272 368 272
WIRE 448 288 448 272
WIRE -64 304 -64 288
WIRE 176 320 176 272
WIRE 176 320 128 320
WIRE 368 320 368 272
WIRE 176 336 176 320
WIRE 128 352 128 320
WIRE 448 384 448 352
WIRE 128 432 128 400
WIRE 176 432 176 416
WIRE 176 432 128 432
WIRE 272 432 176 432
WIRE 368 432 368 400
WIRE 272 448 272 432
WIRE 272 544 272 528
FLAG 448 384 0
FLAG -64 64 0
FLAG -64 304 0
FLAG 288 112 out
FLAG 48 192 in
FLAG 272 544 0
FLAG 240 272 Q1E
SYMBOL npn 112 144 R0
SYMATTR InstName Q1
SYMATTR Value 2N3904
SYMBOL cap 432 288 R0
SYMATTR InstName C1
SYMATTR Value 10p
SYMBOL voltage -64 -48 R0
SYMATTR InstName V1
SYMATTR Value 10V
SYMBOL voltage 272 544 R180
WINDOW 0 36 55 Left 2
WINDOW 3 24 16 Left 2
SYMATTR InstName V2
SYMATTR Value 5V
SYMBOL voltage -64 192 R0
WINDOW 123 24 124 Left 2
WINDOW 39 0 0 Left 2
SYMATTR Value2 AC 1
SYMATTR InstName V3
SYMATTR Value SINE(0 10m 1meg)
SYMBOL g2 176 0 R0
WINDOW 3 34 56 Left 2
SYMATTR Value 1m
SYMATTR InstName G1
SYMBOL g2 176 320 R0
WINDOW 3 36 57 Left 2
SYMATTR Value 1m
SYMATTR InstName G2
SYMBOL res 352 -16 R0
SYMATTR InstName R1
SYMATTR Value 1k
SYMBOL res 352 304 R0
SYMATTR InstName R2
SYMATTR Value 1k
TEXT -56 -104 Left 2 !.noise v(out) v3 dec 1000 1 10meg
TEXT -56 -136 Left 2 ;'G1 behaves like a noiseless 1k resistor
TEXT 480 -48 Left 2 ;Data is in nV/rt(Hz)\n2SC4102 2.05nV\n2N2222A 2.83nV
\n2N3904 2.21nV
Reply by Gerhard Hoffmann●February 18, 20182018-02-18
Am 18.02.2018 um 20:01 schrieb Jim Thompson:
>> Is it really so hard to write the word "noiseless" after the resistance
>> value?
>>
>> <
>> https://www.flickr.com/photos/137684711@N07/38533153940/in/dateposted-public/
>> >
>>
>>
>> Gerhard
>
> Try that in any other Spice. That's my beef. LTspice doesn't follow
> convention. Thus its models won't play elsewhere.
Using pspice and then complaining that LTspice does not conform
to the "standards"!
May I remind you that we had that problem with Pspice models
that could not run on anything else for 20 years or so?
Reply by Jim Thompson●February 18, 20182018-02-18
On Sun, 18 Feb 2018 19:45:11 +0100, Gerhard Hoffmann
<gerhard@hoffmann-hochfrequenz.de> wrote:
>Am 17.02.2018 um 18:01 schrieb Jim Thompson:
>
>>> Version 4
>>> SHEET 1 880 680
>>> WIRE 128 -96 -64 -96
>>> WIRE 176 -96 128 -96
>> ]snip]
>>> TEXT 288 -144 Left 2 !;ac dec 100 1k 100meg
>>> TEXT 288 -88 Left 2 !.noise v(out) v3 dec 100 1k 100meg
>>> TEXT 248 -32 Left 2 ;G1 behaves like a noiseless 1k resistor
>>
>> In _many_ instances where a G-source is used as a resistor in LTspice,
>> LTspice can't find the operating point because Mikey turns off all
>> current sources during the .OP calculation.
>>
>> I've taken to modeling noiseless resistors with an E-source to
>> accommodate LTspice users... not as precise as a G-source :-(
>>
>> ...Jim Thompson
>>
>Is it really so hard to write the word "noiseless" after the resistance
>value?
>
><
>https://www.flickr.com/photos/137684711@N07/38533153940/in/dateposted-public/
> >
>
>
>Gerhard
Try that in any other Spice. That's my beef. LTspice doesn't follow
convention. Thus its models won't play elsewhere.
...Jim Thompson
--
| James E.Thompson | mens |
| Analog Innovations | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| STV, Queen Creek, AZ 85142 Skype: skypeanalog | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
It's what you learn, after you know it all, that counts.
Reply by Gerhard Hoffmann●February 18, 20182018-02-18
Am 17.02.2018 um 18:01 schrieb Jim Thompson:
>> Version 4
>> SHEET 1 880 680
>> WIRE 128 -96 -64 -96
>> WIRE 176 -96 128 -96
> ]snip]
>> TEXT 288 -144 Left 2 !;ac dec 100 1k 100meg
>> TEXT 288 -88 Left 2 !.noise v(out) v3 dec 100 1k 100meg
>> TEXT 248 -32 Left 2 ;G1 behaves like a noiseless 1k resistor
>
> In _many_ instances where a G-source is used as a resistor in LTspice,
> LTspice can't find the operating point because Mikey turns off all
> current sources during the .OP calculation.
>
> I've taken to modeling noiseless resistors with an E-source to
> accommodate LTspice users... not as precise as a G-source :-(
>
> ...Jim Thompson
>