开云体育

--- In LTspice@..., "danfly09" <da.nc3262@...> wrote:

I'm looking for a hall effect current sensor like the ACS712.

You are in luck. Here is a subcircuit I made, but you will have
to make your own symbol for it. Change the gain parameter to
corresponding to the dash number of interest. -- a.s.

.subckt ACS712 I+ I- Filter Out V+ V-
.param gain=185m; 185m -05B, 100m -20A, 66m -30A
V1 I+ I- 0
F1 V- 2 V1 {gain}
T1 1 V- 2 V- Td=1u Zo=1
R1 2 V- 1
R2 Filter 2 1k7
G1 V- 3 Filter V- 1
C1 3 V- 1?3 Rpar=1
R3 Out 3 50
G2 V- 3 V+ V- 0.5
R4 V+ V- 2k
.ends ACS712

For those wishing to create 'permanent' LTspice parts that appear in
your drop down menu of LTspice, please reference Adding a component to
LTspice - Ltwiki.org
<;
ice> and a section on already completed parts following this approach
at Components Library - LTwiki
<> Further
contributions to this library are welcomed, as well. This approach has
also already been included in the messages here over the years, so I
won't repeat it now. It's a little different that what is called out
for in scad3.pdf.

These parts are portable, but you must be aware of what the symbol and
subcircuit names are, that you put in a folder of the &#92;sym directory and
&#92;sub directory. I personally add a folder below the &#92;sym directory for
my own permanent parts. Then I make subcircuit names that are easily
recognizable to sync to my other computers using LTspice.

To make these parts portable, and part of any design you'd want to
share, just include those specified sym and sub files in the same folder
(all zipped together) as your shared schematic. The receiver can keep
these in the same folder as the provided schematic and all goes on as
before. Alternately, they can copy these provided symbols and
subcircuits to their respective folders as outlined in the wiki, and
they will become their own custom permanent parts. (You must restart
LTspice to see them.)

Sometimes, when exchanging schematics between my desktop and laptop I'll
get a little sloppy and forget I'd used one of my permanent parts. So,
LTspice tells me if I try to do an analysis. The bigger embarrassment
is to professionally provide a schematic, or upload on to this group -
just to have it be incomplete.

I hate to elaborate on this topic, as everyone here as justified strong
opinions about it. But you asked - and here is my complete answer.

--- In LTspice@..., eaneonakis@... wrote:

Dear Sir,
Will you please tell us how to do it? There might be cases when it is

not such a bad idea.

Best Regards
E.A.Neonakis
--- In LTspice@..., "analogspiceman" analogspiceman@

wrote:

You can more or less do this now if you know how,

I forgot to add..
The flux doesn't collapse instantaneously... This can be modeled by
putting V(X) through an RC low pass.
Cheers
AG

I have modeled an ideal transformer with time varying coupling
coefficient..
File is in Temp--> Non_ideal_transformer.asc
<>
Your feedback will be appreciated.
Cheers
A. Ganesan

Google is your friend. Search on the phrase "spice hall effect"

Howard

Another problem with the LTspice Help, is that just about everything
starts with the old non-graphical input format of traditional (i.e.
Berkeley) SPICE:

Syntax: Sxxx n1 n2 nc+ nc- <model> [on,off]

Example:

S1 out 0 in 0 MySwitch

.model MySwitch SW(Ron=.1 Roff=1Meg Vt=0 Vh=-.5 Lser=10n Vser=.6)

And then we tell everyone to use the schematic editor, and upload
schematic files instead of netlists.

So the Help file is a little like using DOS in a Windows world.

Plus the references to a SPICE "deck", a model "card", etc. Imagine
if the Help for MS-Windows gave everything in terms of DOS commands
and command-line switches.

It probably makes little sense to anyone coming to LTspice for the first time.

For those of us who used SPICE before, the relationship between the
element and the model, by way of the model name, was obvious. And we
take for granted that the element is now a schematic symbol, but the
model is a "SPICE directive" ... which is mentioned in the Help
description but not really explained.

So yes, there are lots of problems.

Andy

Hi, i'm looking for a hall effect current sensor like the asc712... there isn't a .lib that i could simulate?? or anything possible solution but without use a shunt resistor

--- In LTspice@..., "ttakeshian" <ttakeshian@...> wrote:

I need to model the coupling coefficient of two inductors as a function of time. I tried using combination of .func and .param statements to pass a time varying variable to the spice K component without success. I also have tried the trick used for defining a time varying resistor using a voltage of a node in the schematic.

Any suggestion would be greatly appreciated.

tony

Tony,

LTspice can be used to model any system of ordinary (not partical) differential equations using "analog computer" techniques. The coefficients can be made time varying by using BV elements that are a function of time.

Rick

In message <j50avg+6kln@...>, dated Fri, 16 Sep 2011, analogspiceman <analogspiceman@...> writes:

Well, in the switch example, Help offered,

"See the schematic file .&#92;examples&#92;Educational&#92;Vswitch.asc to see an example of a model card placed directly on a schematic as a SPICE directive.",

I would add a few more words. 'Open the downloaded LTspiceIV folder, open examples and then Educational, and open the schematic file Vswitch.asc'....

In doing that, I've assumed no more than beginner level understanding of Windows (at least in English) as well. That dot in

.&#92;examples&#92;Educational&#92;Vswitch.asc

hides a lot of meaning.

which the original poster seemed not to have read (or at least ignored, hence my legitimate question about the possibility of there being either a reading comprehension problem or laziness at play). Note that this example would have answered any lingering questions about what "a model card" could mean.

It would certainly have minimised doubt anyway, but I agree with you that 'card' [and 'deck'] are archaisms that should be eliminated.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
When I point to a star, please look at the star, not my finger. The star will
be more interesting.

--- In LTspice@..., John Woodgate wrote:

A constructive, serious answer! Thank you.

[Y]ou have initially to forget all you know about the product,
program or application and consider ONLY what questions the
user is likely to ask.

This much I already supposed, but the reemphasis is appreciated.

Looking back at the original poster's difficulty in processing
this sentence from Help,

"A model card is required to define the behavior of the switch.",

which to me clearly indicates that the user must create a model
statement in which he assigns values to the switch parameters.
Also implied is that this model statement is usually created in
the form of a SPICE text directive and placed upon the schematic.
A complete novice probably wouldn't know this without an example
or pouring over the manual (which I did when I was brand new to
LTspice, by the way).

Also the sentence refers to a model card, which is most likely a
meaningless term to those who weren't around in the days of punch
card input (I wish Mike would expunge that term from any and all
non-historical documentation). Still, these shortcomings would
be overcome and answered by an example, a link to which Help *did*
provide in this case.

For each question, you then 'switch on' ONLY that part of your
product knowledge needed to answer the question. If you feel
compelled to add more, formulate another question to introduce
the addition.

This is really good advice, as it would tend to overcome the
problems from reading comprehension and information overload that
were mentioned earlier. I wonder if this technique can be done
in a way that is compatible with Help remaining an efficient
reference manual. Anyway, thanks for the idea.

Hundreds of questions are likely to be required. The limited
lists of up to 20 or so, offered by Microsoft and other Helps,
are pitifully inadequate.

True, but a well done wiki could offer unlimited possibilities,
both in links and graphics.

Answers should include examples wherever possible.

Well, in the switch example, Help offered,

"See the schematic file .&#92;examples&#92;Educational&#92;Vswitch.asc to see
an example of a model card placed directly on a schematic as a
SPICE directive.",

which the original poster seemed not to have read (or at least
ignored, hence my legitimate question about the possibility of
there being either a reading comprehension problem or laziness at
play). Note that this example would have answered any lingering
questions about what "a model card" could mean.

Regards -- analogspiceman

In message <075FBDBF-7FC5-4AE8-921F-42D73C76DF0D@...>, dated Fri, 16 Sep 2011, Jim Wagner <wagnerj@...> writes:

As a further example, there is not a single screen snapshot in the whole documentation.

I found several. 'Adding attributes', 'Adding the pins' and 'Attached cursors', for example, all include graphics if not exactly screen-shots. 'Colour palette editor' is another.

It just says "do this... " or "do that...". For MANY people, that is simply not sufficient. That is not the way that many people learn.

Agreed; it's not always the best way to communicate.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
When I point to a star, please look at the star, not my finger. The star will
be more interesting.

In message <744FF547-789E-48F5-9AE6-A4FF04DCCB25@...>, dated Fri, 16 Sep 2011, Jim Wagner <wagnerj@...> writes:

As far as I can see, "constructive criticism" does no good. The documentation appears to be the sole perview of Mike and it will be what he wants it to be.

Mike obviously has a responsibility to ensure that the Help is completely accurate. But to extend that to saying that no improvement will be entertained is not justified.

End of story.

By no means.

I have offered several positive suggestions about documentation improvement, including the "go check PSpice documentation, its mostly the same" and put what needs to be put into single document. That was shot down by you and others. So, what else is one to do?

Simply don't accept the 'shoot down'. You can write, if you choose, your own 'Jim's Guide to LTspice' and put it on your web site or even in the Files or Files -> Tut folders on the list's web site.

I would be delighted to contribute to the Wiki, but its not clear how. The front page of the Wiki says "create an account to contribute" but there is no link or information about creating an "account".

If so, it needs to be fixed.

Clearly,j genuine input is not wanted.

Now, that is NOT a constructive criticism. I hope this message will de-frustrate you, at least a bit.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
When I point to a star, please look at the star, not my finger. The star will
be more interesting.

On Sep 16, 2011, at 11:12 AM, John Woodgate wrote:

In message <j500ob+lpdp@...>, dated Fri, 16 Sep 2011,
analogspiceman <analogspiceman@...> writes:

Well, people keeping complaining about Help, but so far no one,
including the complainers, has been willing or able to offer any
concrete constructive criticism. It remains a puzzle. -- a.s.

It needs a complete re-write, but the necessary combination of
skills is
extremely rare - a way to put over intricate concepts in simple
language
AND a deep understanding of LTspice. I have had some success in the
first, but I am nowhere in the second.

To do the first, you have initially to forget all you know about the
product, program or application and consider ONLY what questions the
user is likely to ask. For each question, you then 'switch on' ONLY
that
part of your product knowledge needed to answer the question. If you
feel compelled to add more, formulate another question to introduce
the
addition.

Hundreds of questions are likely to be required. The limited lists
of up
to 20 or so, offered by Microsoft and other Helps, are pitifully
inadequate. Answers should include examples wherever possible.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
When I point to a star, please look at the star, not my finger. The
star will
be more interesting.

As a further example, there is not a single screen snapshot in the
whole documentation. It just says "do this... " or "do that...". For
MANY people, that is simply not sufficient. That is not the way that
many people learn.

Another example: the documentation has few REAL examples. Step-by-
step, picture-by-picture, of how you add a model, or a sub-circuit.
Again, words are fine, words are necessary, but words are not
sufficient.

Jim Wagner
Oregon Research Electronics

[Non-text portions of this message have been removed]

On Sep 16, 2011, at 10:24 AM, analogspiceman wrote:

--- In LTspice@..., John Woodgate wrote:

--- In LTspice@..., analogspiceman wrote:

How to make Help continue to function as a compact and efficient
reference while also being able to effectively provide answers
that neophytes can actually see, process and put to use?

This is, in my experience, a problem with ALL Helps. I could write
a 3-screen post but I won't.

Well, people keeping complaining about Help, but so far no one,
including the complainers, has been willing or able to offer any
concrete constructive criticism. It remains a puzzle. -- a.s.

As far as I can see, "constructive criticism" does no good. The
documentation appears to be the sole perview of Mike and it will be
what he wants it to be.

End of story.

I have offered several positive suggestions about documentation
improvement, including the "go check PSpice documentation, its mostly
the same" and put what needs to be put into single document. That was
shot down by you and others. So, what else is one to do?

I would be delighted to contribute to the Wiki, but its not clear how.
The front page of the Wiki says "create an account to contribute" but
there is no link or information about creating an "account".

Clearly,j genuine input is not wanted.

Jim Wagner
Oregon Research Electronics

[Non-text portions of this message have been removed]

In message <j500ob+lpdp@...>, dated Fri, 16 Sep 2011, analogspiceman <analogspiceman@...> writes:

Well, people keeping complaining about Help, but so far no one, including the complainers, has been willing or able to offer any concrete constructive criticism. It remains a puzzle. -- a.s.

It needs a complete re-write, but the necessary combination of skills is extremely rare - a way to put over intricate concepts in simple language AND a deep understanding of LTspice. I have had some success in the first, but I am nowhere in the second.

To do the first, you have initially to forget all you know about the product, program or application and consider ONLY what questions the user is likely to ask. For each question, you then 'switch on' ONLY that part of your product knowledge needed to answer the question. If you feel compelled to add more, formulate another question to introduce the addition.

Hundreds of questions are likely to be required. The limited lists of up to 20 or so, offered by Microsoft and other Helps, are pitifully inadequate. Answers should include examples wherever possible.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
When I point to a star, please look at the star, not my finger. The star will
be more interesting.

Agree whole-heartedly, Helmut.

--- In LTspice@..., John Woodgate wrote:

--- In LTspice@..., analogspiceman wrote:

How to make Help continue to function as a compact and efficient
reference while also being able to effectively provide answers
that neophytes can actually see, process and put to use?

This is, in my experience, a problem with ALL Helps. I could write
a 3-screen post but I won't.

Well, people keeping complaining about Help, but so far no one,
including the complainers, has been willing or able to offer any
concrete constructive criticism. It remains a puzzle. -- a.s.

--- In LTspice@..., Tony "ttakeshian" wrote:

I need to model the coupling coefficient of two inductors as
a function of time. I tried using combination of .func and
.param statements to pass a time varying variable to the
SPICE K component without success.

As you have noted, K must be a fixed value. However, LTspice
includes a behavioral inductor that may be a function of time.
I suppose you could set up a T-network of three such inductors
to vary with time the percentage of inductance shared between
each side of the "T" while keeping the total inductance "seen"
from each side fixed (this is what K effectively does). If
you require other than a one-to-one turns ration, you could
add an ideal transformer to one side of the T. -- a.s.

PS: The behavioral inductor is well described in Help and in the
many examples available here in the group archive, both in past
messages and in files posted.

In message <j4vu3g+ano2@...>, dated Fri, 16 Sep 2011, analogspiceman <analogspiceman@...> writes:

How to make Help continue to function as a compact and efficient reference while also being able to effectively provide answers that neophytes can actually see, process and put to use?

This is, in my experience, a problem with ALL Helps. I could write a 3-screen post but I won't. It's a good PhD project on communication for someone.
--
OOO - Own Opinions Only. Try www.jmwa.demon.co.uk and www.isce.org.uk
John Woodgate, J M Woodgate and Associates, Rayleigh, Essex UK
When I point to a star, please look at the star, not my finger. The star will
be more interesting.

An emphatic yes. In one of my courses, Network Synthesis, one learns that for minimum phase networks (no zeros in the right half plane), the phase response of the network can be determined from the magnitude frequency response. As to A.S.'s remarks, I do not believe you have a physically realizable circuit by replacing a Laplace transform by its magnitude response. Complex poles and zeros need to occur in pairs in order for the time function to be a real valued function -- and to have a physically realizable circuit.

Hubert