Re: LTspice models do not work in Qspice
|
LTspice models do not work in Qspice
Hi,
I have models (.suckt) that are working perfectly in LTspice, when I import them in Qspice they dont work
Does Qspice use a different syntax?
Regards,
Suded
|
Re: Does anyone have the model file for FinFET that will be compatible with LTspice ?
On Fri, Apr 18, 2025 at 02:32 AM, Say Die. wrote:
I Am trying for 20nm FinFET design and i have 45 and 180nm done but i cannot find any file for 20nm with 54 or any other level for ltspice.
One thing to keep in mind about transistor models, is that they are meant to describe the fabrication process.? They are inherently linked to only the FAB that you use.? A model for one FAB line does not work for another FAB line or from another manufacturer.? Normally, one gets the SPICE model from the manufacturer - not by finding something on the Internet!
?
When talking about (say) a 20nm process, it might be that everyone gets their chips from the same manufacturer made to the same specifications, and then there would be only one model that everyone uses for that step.? But don't count on it.? In general, you should ask the people who will make your chips, for their SPICE models.? Specify that you need LEVEL=X models (pick one from the list), or only LTspice-capable models, but they are less likely to understand LTspice than LEVEL numbers.? I would not count on models from anyone else to work for your 20nm chips.
?
Andy
?
|
Re: Does anyone have the model file for FinFET that will be compatible with LTspice ?
I looked online and it appears that there are freely available SPICE simulators that support Verilog-AMS (and thus Verilog-A).
Since you would have to decide which (if any) ?would suit your needs, I'd suggest doing an internet search using the string "spice simulators that support verilog-a" or something such as that.
The FinFET is typically modeled using the BSIM-CMG model which is not available in LTspice.? I do not know of a free simulator that supports the BSIM-CMG model unless it is through a Verilog-A representation.
?
BB
|
Re: Does anyone have the model file for FinFET that will be compatible with LTspice ?
The FinFET is typically modeled using the BSIM-CMG model which is not available in LTspice.? I do not know of a free simulator that supports the BSIM-CMG model unless it is through a Verilog-A representation.
?
BB
|
Re: Does anyone have the model file for FinFET that will be compatible with LTspice ?
On Fri, Apr 18, 2025 at 02:32 AM, Say Die. wrote:
also there's something called BSIM model file if you have this then it's also fine
A number of the MOSFET LEVELs are BSIM.? BSIM is not one single model, there are several, a few major levels? and several sub-levels.??Level=4 is BSIM,? Level=5 is BSIM2.? Level=7 is BSIM3.? Level=8 is BSIM3, version 3.3.0.? Level=9 is BSIM3 version 3.2.? Level=14 is BSIM4 version 4.6.1.? Level=49 is BSIM3.? Level=54 is BSIM4.
?
Some of this you could have ascertained by reading LTspice's Help.? Did you try looking there?
?
but I don't know how we use this file in LTspice.
And what "file" is that?? Do you have a model file that can be identified by its LEVEL number?? Use that number to see if LTspice can use it.? As far as I know, these MOSFET LEVELs are accepted:
?
1,?2, 3, 4, 5, 6,?7, 8,?9,?12,?14,?49,?54, 73
?
All other LEVELs are not accepted and cannot be used in LTspice.
?
But be aware that one or two SPICE programs might use different LEVEL numbers for the same model.? Sigh....
?
Andy
?
|
Does anyone have the model file for FinFET that will be compatible with LTspice ?
I Am trying for 20nm FinFET design and i have 45 and 180nm done but i cannot find any file for 20nm with 54 or any other level for ltspice.? if anyone have this file please provide it to me also there's something called BSIM model file if you have this then it's also fine but I don't know how we use this file in LTspice. Please, Reply ASAP
?
|
Re: Have anyone built black hole model already ?
On Thu, Apr 17, 2025 at 09:11 PM, <ericsson.sunshine@...> wrote:
Change the 'rise' time from long time to 1 second (the shortened sweep time), will solve all problems, no more confusion, simulation pain.
That does not solve *all* problems.? You will always have the fact that the input signal to the notch filter is a frequency sweep, so it is not one unique frequency.? Attempting to pass that through a high-Q filter will have artifacts.? They can be diminished (not eliminated) by slowing down the frequency sweep, as Tony said already.? This is very well known in filter theory.
?
And I guess 'Heisenberg's Uncertainty Principle' also apply , in the measuring of distance variation due to 'gravitational wave' of black hole. But this relation , don't know how to well describe in SPICE. By random() ???
I cannot speak to the physics or space-time aspect.? But in college an instructor told us that Heisenberg's uncertainty principle was first described, or applied, to problems involving frequency and time, e.g., Fourier spectra.? It was later (or around the same time) applied to particle physics.
?
I rather doubt that randomness has anything to do with this, SPICE or not.? The stuff we are talking about here is very deterministic (even if the measurements are imprecise) and randomness need not - should not - be applied - except as a way to blind the people you are talking to, and make them think that it is too mysterious for them to comprehend.? Don't go there.
And, I think the frequency sweep should be done well by both, PWL & PULSE, I was used to use PULSE.
Well, yes.? I was just suggesting that a PWL source is better suited for generating a straight-forward voltage ramp, than a PULSE source is.? The PULSE source has too many parameters that interact.? (Did you notice that LTspice rejected your value for its Period?)? Whereas a PWL is ideally suited for generating a voltage ramp when that is all that you need.
?
I am still missing the connection, if any, to a black hole.? This is just an audio filter, is it not?
?
Andy
?
|
Re: Have anyone built black hole model already ?
Hi, :
?
Thank you all who replied, it's nothing special , just to make some relevant between SPICE & physics, which should be able to describe by math equations.
?
The lib file can be found in folders,?
?
Change the 'rise' time from long time to 1 second (the shortened sweep time), will solve all problems, no more confusion, simulation pain.
?
And I guess 'Heisenberg's Uncertainty Principle' also apply , in the measuring of distance variation due to 'gravitational wave' of black hole. But this relation , don't know how to well describe in SPICE. By random() ???
?
And, I think the frequency sweep should be done well by both, PWL & PULSE, I was used to use PULSE.
I have no idea what's supposed in this 'Black hole' simulation, maybe just for fun, I slowly felt nothing can do day by day.
?
But , I guess will still try harder to make more simulation.
?
Thank you all again.
Best regards.
|
Re: creating gm/Id vs Id/W plot in ltspice
Change the whole file. I gave you a complete .plt file that
provides the plot you wanted (if I understood your request
correctly; there seems to be some potential language confusion).
Donald.
On 4/17/25 10:35, john23 via groups.io
wrote:
toggle quoted message
Show quoted text
[DC transfer characteristic]
{
? ?Npanes: 2
? ?{
? ? ? traces: 1 {34668548,0,"Is(M1)"}
? ? ? X: (' ',1,0.01,0.2,1.8)
? ? ? Y[0]: ('?',1,0,8e-07,8e-06)
? ? ? Y[1]: ('_',0,1e+308,0,-1e+308)
? ? ? Amps: ('?',0,0,1,0,8e-07,8e-06)
? ? ? Log: 0 0 0
? ?},
? ?{
? ? ? traces: 1 {524290,0,"-D(Is(M1))/D(V(Vgs))"}
? ? ? X: (' ',1,0.01,0.2,1.8)
? ? ? Y[0]: ('?',0,0,1e-06,1.3e-05)
? ? ? Y[1]: ('_',0,1e+308,0,-1e+308)
? ? ? Units: "ohm-1" ('?',0,0,0,0,1e-06,1.3e-05)
? ? ? Log: 0 0 0
? ?}
}
|
Re: creating gm/Id vs Id/W plot in ltspice
Sorry to hear that. I often tweak the .plt file after figuring
out what I really want it to display. (Though I admit to starting
with a Ctl-S in the plot pane to provide a starter file.)
Donald.
On 4/17/25 10:45, Andy I via groups.io
wrote:
toggle quoted message
Show quoted text
On Thu, Apr 17, 2025 at 10:35 AM, john23 wrote:
Hello ,My original PLT file of the attached photo is
shown below.
Where do I need to change this file ? and to what line do I
need to change?
I strongly recommend that you do NOT ever edit your .PLT
file.? That is not the way to change your LTspice plots.? It is
too dangerous, for one thing.
?
Instead, make the changes the normal way: by editing settings
in the program (LTspice) itself.
?
You told us that you are doing a .DC sweep, but that is not
nearly enough information to give you sound advice.? Plus your
questions are somewhat spread out over multiple topics and
multiple simulations, and I find it difficult to track which
one(s) you are talking about.? What EXACTLY are you trying to
do, this time?? Did you follow Tony's advice earlier?? Did it
work?? If it did not work, what happened instead?
?
Sending us a screenshot of a simulation, as you have done
this time, is absolutely unhelpful.? How does that screenshot
relate to your question?? How can I simulate your screenshot on
MY computer to help answer your question?
?
And what is your objective, anyway?? You have lost me.
Andy
?
|
Re: Have anyone built black hole model already ?
The PULSE()'s Period is pretty much irrelevant, because it simulates only in the first rise time.? Would be better to use PWL instead of PULSE, IMHO.
?
But what's it supposed to do?? More to the point, what is it?not doing?
?
Andy
|
Re: Have anyone built black hole model already ?
I parameterized the run time as Period, and used that in the .TRAN and Pulse Rise Time
Set to Period 10, and it all sorta works.
(2x5k in place ot the pot, for a quick fix)
?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]>
On Behalf Of Andy I via groups.io
Sent: Thursday, April 17, 2025 10:59 AM
To: [email protected]
Subject: EXTERNAL: Re: [LTspice] Have anyone built black hole model already ?
?
Now it's missing the model file, "potentiometer_standard.lib".
Should we change PULSE(...) to PULSE(0 1 0 1 1) like in the photo?? Or should we change .TRAN to .TRAN 4730400000 UIC ?
|
Re: Have anyone built black hole model already ?
ericsson.sunshine, I remain confused about your objective here.
?
The note attached to your upload .zip says "Can't see the singularity of .TRAN time domain".
?
And yet, the simulation and the photo you uploaded both circled a point in the waveform identified as the "Zero Point of Black Hole take effect".
?
Is something mysterious - perhaps something non-physical - supposed to happen when the incoming frequency passes through the notch frequency?? Should the computer's on-screen colors invert or become ultraviolet or something unusual?? Should the simulation freeze at that point?? I don't see a reason why the circuit should not behave in exactly the way that it does.? (Also noting Tony's cautions about timestep and sweep speed.)
?
The AC plot points to the notch and labels it "Singularity".? OK....
?
Like I said, everything looks normal to me.? If something is not working the way you expected, making it so that you can't "see the singularity", can you please explain that to me?
?
Andy
?
|
Re: Have anyone built black hole model already ?
On Thu, Apr 17, 2025 at 12:53 PM, <ericsson.sunshine@...> wrote:
Now it's missing the model file, "potentiometer_standard.lib".
?
What else is missing?
?
Should we change PULSE(...) to PULSE(0 1 0 1 1) like in the photo?? Or should we change .TRAN to .TRAN 4730400000 UIC ?
?
Andy
?
|
Re: creating gm/Id vs Id/W plot in ltspice
On Thu, Apr 17, 2025 at 11:48 AM, John Woodgate wrote:
Thanks. I do occasionally get one right.
You do get them right!? And I mistakenly credited Tony instead of you - which I already fixed in the online message but I should give you the credit. ?
?
|
Re: Have anyone built black hole model already ?
Already uploaded.
Can't recognize what's default built-in symbol, what's not.
?
?
Sorry for the.
?
On Thu, Apr 17, 2025 at 11:14 PM, Andy I wrote:
Did you forget the symbol potentiometer_standard.asy?? It seems to be missing.
?
Andy
?
?
|
Re: Have anyone built black hole model already ?
Ah...
?
I see, it's the copy-paste error from another dinosaur's life-time simulation.
?
Thank you very much. Correct it now. possibly eye's tired.
?
On Thu, Apr 17, 2025 at 10:56 PM, Bell, Dave wrote:
Well, at first glance, not knowing quite what you¡¯re looking for,
- Plot V(Out) on the horizontal axis
- This: ¡°PULSE(0 1 0 {1*60*60*24*365*150} 1 0 2)¡± takes 150 YEARS ro sweep the frequency, and you¡¯re only running a 1 SECOND .TRAN\
?
Dave
?
?
|
Re: creating gm/Id vs Id/W plot in ltspice
Thanks. I do occasionally get one right.
On 2025-04-17 15:54, john23 via
groups.io wrote:
Hello John Woodgate,Your method worked perfectly as shown in
the photo.
Thank you very much.
?
--
Best wishes John Woodgate RAYLEIGH Essex OOO-Own Opinions Only
If something is true: * as far as we know - it's science *for
certain - it's mathematics *unquestionably - it's religion
|
Re: Have anyone built black hole model already ?
I don't recall the discussion from 7 years ago.
?
But I wonder, is the point of this exercise to demonstrate a "notch filter"?? And then give it a funny false name like "black hole"?
?
In electronics and filters, "notch filters" are normally called "notch filters".? I never heard anyone refer to them as "black holes".? But I suppose it is true that a perfect notch filter would suck out all the signal at that one frequency, assuming of course that you stimulated it with only that one frequency and not a signal with a varying frequency or amplitude, since neither of those contains just that one frequency.
?
Are you learning how notch filters work?? Is that the point of this exercise?? Or do you have something else in mind?
?
Andy
?
|
BB
Bell, Dave
John Woodgate