LTspice/SwitcherCAD III now runs on Linux
Mike Engelhardt
This is an unlimited free schematic capture/SPICE
simulation program. Attached is the relevant part
of the FAQ from the help system:
Q. Do you have a version of LTspice for Linux?
A. Not a separate edition, but it does run under
WINE. The program has been tested on Linux
RedHat 8.0 with WINE version 20030219.
Q. OK, I've never used WINE, how do I install that
on my Linux box?
A. 1. Check with www.winehq.com to find the current
version of WINE for your system. At the time
of this writing, for RedHat 8.0, this points
to http://mecano.gme.usherb.ca/~vberon/wine
2. Copy the appropriate .rpm file to your machine
and open it from nautilus.
3. Get the file swcadiii.exe from www.linear.com.
In an xterm, execute "wine swcadiii.exe" to
install LTspice.
4. There will now be a Linear Technology Logo on
your gnome desktop. Double click it to start
or type "wine scad3.exe" from an xterm to
start the program. That's it!
Q. The schematic fonts don't scale as smoothly under
WINE as Windows. Why is that?
A. WINE is doing the best best it can with the fonts
it finds. It will do better if you tell it how
to find the files arial.ttf and cour.ttf from
your Windows system.
Q. The PWL additional point editor doesn't look
right under WINE?
A. Try using the native Windows .dll from your
Windows system. The command line to then invoke
LTspice from WINE is:
wine -dll commctrl,comctl32=n scad3.exe.
Q. It seems LTspice is running slightly differently
under WINE/Linux than windows. Why is that?
A. LTspice detects whether or not it's running under
WINE. If so, it works around a few WINE issues.
You can force LTspice to think it's running under
WINE with the command line switch -wine. You can
force it to think it's not with the command line
switch -nowine in case you're interesting in
debugging WINE.
Q. Under Windows, LTspice extends the virtual
address space for waveform viewing to 64 bits.
Does that work under WINE?
A. Yes. It has been tested on waveform files over
5 Gigabytes in size.
Q. Under Linux, does it support unlimited component
count and depth of schematic hierarchy?
A. Yes. Opps, no. The maximum depth of hierarchy
is 64 levels but that limit is just there to
allow detection of infinite subcircuit recursion.
(Most "unlimited" SPICE's "limit" you to about 21
levels.) LTspice has no limit on node or
component count either per page or for the
entire circuit.
Q. Does cross-probing while simulation work with
those slick marching waveforms while running
under Linux?
A. Yes.
Q. So from what version on is LTspice supported
under WINE?
A. 2.01g.
--Mike
PS. Thanks to UB and CC for their inputs.
Christopher R. Carlen
> LTspice/SwitcherCAD III now runs on Linux under WINE.
>
Mike, may I express my deep appreciation for your efforts.
You should post this on comp.os.linux.announce too, if you haven't already.
Thanks for the great program.
I'll be providing periodic feedback, bug reports, and the like.
Good day.
--
____________________________________
Christopher R. Carlen
Principal Laser/Optical Technologist
Sandia National Laboratories CA USA
crc...@sandia.gov
Mike Monett
> LTspice/SwitcherCAD III now runs on Linux under WINE.
The is great news! Thanks, Mike.
Mike Monett
Paul Burridge
thusly:
Mike! Where have you been lately?
I eventually got a demo copy of Microcap to run properly and have to
admit it was far superior to B2 Spice. The front-end was particularly
impressive - gorgeous schematic representation which a lot of other
programs could benefit from. Super-easy parts placement &
connecting-up on the schematics too.
But now you appear to be won over by LTSpice. I'm currently trying
that too and find it the best overall thus far by a long way. Can't
seem to move components once they've been placed though! Must be my
fault but if you or anyone else knows the secret, I'd be keen to know!
--
"What is now proved was once only imagin'd"
- William Blake, 1793
anthony wooldridge
"Paul Burridge" <red...@waitrose.notthisbit.com> wrote in message
news:99ib6vcp4j90svi5n...@4ax.com...
Yes its not that intuitive in editing, but I like the speed, and the neat
probe for voltage or current.
To move use edit/move edit/delete etc.
It should be more intuitive like click a click operation.
Paul Burridge
<arwooldrid...@genie.co.uk> opined thusly:
>Yes its not that intuitive in editing, but I like the speed, and the neat
>probe for voltage or current.
>To move use edit/move edit/delete etc.
>It should be more intuitive like click a click operation.
Thanks. I'll make a mental note of that.
One other thing that bugs me is that when I run an AC sweep of a
circuit's output, I get two traces. One is gain v. frequency and the
other is - I can only assume - frequency v. phase difference. However,
they're both referred to by one node number and are both the same
colour, so very hard to tell apart especially at large frequency
spreads. Has anyone else had difficulties with this and is there a way
around it? I'm not particularly interested in phase difference and
fail to see the relevance of it.
<confused>
anthony wooldridge
Phase can be very relevant in closed loop applications.
I haven't used AC in LTspice yet. Looking at it now though I can't even see
it.
are you useing the .ac spice command?
There's some colour controls under tools have you tried those?
Anthony
analog
Paul Burridge, Anthony Wooldridge, Paul Burridge wrote:
>>> But now you appear to be won over by LTSpice. I'm currently trying
>>> that too and find it the best overall thus far by a long way. Can't
>>> seem to move components once they've been placed though! Must be my
>>> fault but if you or anyone else knows the secret, I'd be keen to know!
>>
>> neat probe for voltage or current.
>> To move use edit/move edit/delete etc.
>> It should be more intuitive like click a click operation.
Yes, the editing deviates almost completely from the windows standard
so it seems non intuitive at first - backwards, in fact. It is kind
of like the situation with reverse Polish notation and calculators -
less keystrokes once you get used to it. First select the tool
(preferably via a keystroke with the left hand) and then click on or
drag a selection box around the target of the action with the mouse.
> Thanks. I'll make a mental note of that.
> One other thing that bugs me is that when I run an AC sweep of a
> circuit's output, I get two traces. One is gain v. frequency and the
> other is - I can only assume - frequency v. phase difference. However,
> they're both referred to by one node number and are both the same
> colour, so very hard to tell apart especially at large frequency
> spreads. Has anyone else had difficulties with this and is there a
> way around it? I'm not particularly interested in phase difference
> and fail to see the relevance of it.
Yes, this is somewhat confusing. The phase trace should be dashed or
otherwise visually differentiated (should keep the matching color,
though). You can use the cursor to trace along the waveform to
distinguish between the two, or simply turn off the phase axis display
altogether (unfortunately this must be done each time). -- analog
Kevin Aylward
> Paul Burridge, Anthony Wooldridge, Paul Burridge wrote:
>
>>>> But now you appear to be won over by LTSpice. I'm currently trying
>>>> that too and find it the best overall thus far by a long way.
>>>> Can't seem to move components once they've been placed though!
>>>> Must be my fault but if you or anyone else knows the secret, I'd
>>>> be keen to know!
>>>
>>> Yes its not that intuitive in editing, but I like the speed, and
>>> the neat probe for voltage or current.
>>> To move use edit/move edit/delete etc.
>>> It should be more intuitive like click a click operation.
>
> Yes, the editing deviates almost completely from the windows standard
> so it seems non intuitive at first - backwards, in fact.
Therefore, it *is* non-intuitive. It doesn't do what you expect it to
do.
> It is kind
> of like the situation with reverse Polish notation and calculators -
> less keystrokes once you get used to it.
Why do you say that?. What could be more simpler then a single grab with
the mouse to move a component? Its less key strokes. Or a simple move of
the mouse over a component and the delete key?
Kevin Aylward
sa...@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.
analog
Kevin Aylward, analog, Paul Burridge, Anthony Wooldridge, Paul B wrote:
>>>>> But now you appear to be won over by LTSpice. I'm currently trying
>>>>> that too and find it the best overall thus far by a long way.
>>>>> Can't seem to move components once they've been placed though!
>>>>> Must be my fault but if you or anyone else knows the secret, I'd
>>>>> be keen to know!
>>>>
>>>> Yes its not that intuitive in editing, but I like the speed, and
>>>> the neat probe for voltage or current.
>>>> To move use edit/move edit/delete etc.
>>>> It should be more intuitive like click a click operation.
>>
>> Yes, the editing deviates almost completely from the windows standard
>> so it seems non intuitive at first - backwards, in fact.
>
> Therefore, it *is* non-intuitive. It doesn't do what you expect it to
> do.
You are absolutely correct. In this regard LTspice is hobbled with a
legacy interface. Mike E explained that LTspice must be maintain
operative compatibility with older, non-windows based in-house software
used by LTC. In my opinion, he would win over more potential users by
providing a keyboard macro capability to provide a more standard
interface for the outside world, or at least include in the introductory
chapter of the help file a brief explanation of how and why the LTspice
interface differs from the windows standard. Fortunately it only takes
about an hour to get used to the RWN (reverse windows notation) user
interface. Not a big problem.
>> It is kind of like the situation with reverse Polish notation and
>> calculators - less keystrokes once you get used to it.
>
> Why do you say that?. What could be more simpler then a single grab
> with the mouse to move a component? Its less key strokes. Or a simple
> move of the mouse over a component and the delete key?
You're not going to get an argument from me there, but I will say that
the LTspice interface is very quick when used with the left hand hovering
over the keyboard and making all the tool selections and the right hand
driving the mouse. I just wish all of the most used tools were selectable
without having to reach way up to the function keys or over to the right
hand side of the keyboard. (Mike E, are you reading this? And have I
been blackballed because of my "what they don't want you to know posts"?)
analog
anthony wooldridge
"analog" <ana...@ieee.org> wrote in message
news:3E661D0E...@ieee.org...
OK I got the AC going now you just need to add an AC source somewhere.
On my version (2.00p ) the phase is already dotted and the gain is solid.
So download the latest version.
As a general rule phase normally increases but gain decreases with frequency
so you can't really muddle them up.
Anthony
Mike Engelhardt
standard" is that it differs from the windows
standards used in editors. There you select first
and then say what you want to do to it next. Most
CAD editing tools are like that because most are
hobbled together schematic editing tools with
added-on simulators.
LTspice is not. Its UI was carefully considered
to be appropriate for an integrated schematic
capture and SPICE. This makes it deviate from
some editing conventions used in some windows
software. For example, the default behavior
in LTspice when you click on something in the
schematic is to plot the data associated with it.
The default behavior should not be to select it
so that you can change the schematic while the
simulation is running so that the relation
between schematic and simulation is broken. The
convention used by programs that edit only or
have some junk SPICE hobbled on as an after
thought is simply the incorrect one to use.
The net effect of LTspice, with it's xprobing
and marching waveforms, is to give an experience
analogous to scoping out board while it is
running. It breaks out of the confinement
many SPICE programs give in having to carefully
compose a simulation and then see what happened
after it's done running.
And it still uses fewer key and mouse strokes
for editing because usually one needs to move
a bunch of things around not do may different
editing operations on one set of selected
objects. Besides, the select first software
clears the selection after the operation so
there is no possibility to save key/mouse strokes
anyway. That "convention" really seems to be
made to insult the users' intelligence. Anyway,
LTspice's UI stands up to expert use by people
who edit in it all day long day after day, year
after year.
Now, I find it a perverse form of narcissism
to get enthralled in one's own UI so I don't
usually reveal my opinion on the subject.
However, I do believe that LTspice's UI is
better than any other out there. It can't be
immediately intuitive for everyone because not
everyone comes from the same background and
even fewer have seen an integrated Schematic
capture/SPICE program. We get *much* more
feedback from users that LTspice is intuitive
than it is not intuitive. It also uses many
techniques to help one quickly develop
intuition with the program.
--Mike
Ned Konz
> Now, I find it a perverse form of narcissism
> to get enthralled in one's own UI so I don't
> usually reveal my opinion on the subject.
> However, I do believe that LTspice's UI is
> better than any other out there. It can't be
> immediately intuitive for everyone because not
> everyone comes from the same background and
> even fewer have seen an integrated Schematic
> capture/SPICE program. We get much more
> feedback from users that LTspice is intuitive
> than it is not intuitive. It also uses many
> techniques to help one quickly develop
> intuition with the program.
Hi Mike,
I really like LTSpice, but don't have any function keys on my keyboard.
Would it be possible to have remappable keystroke commands?
Thanks,
Ned Konz
Mike Engelhardt
Yes, this is something I've been needing to do for awhile.
Can't promise a release date, though. It shouldn't be too
hard now after what I needed to do to get it running under
Linux/WINE.
--Mike
Maxwell S. Diamond
Hit F7 or the Move button on the toolbar. To move an already connected
part along the wire use F8 or the Drag button. Works for rearranging
the wiring too.
Maxwell S. Diamond
--
All opinions are the responsibility of the author and do
not in any way reflect the views of his employer, spouse or
parents, who have their own windows anyway.
Paul Burridge
<arwooldrid...@genie.co.uk> opined thusly:
>OK I got the AC going now you just need to add an AC source somewhere.
>On my version (2.00p ) the phase is already dotted and the gain is solid.
>So download the latest version.
I do have the latest version, Anthony. And although I have a few minor
quibbles, I remain profoundly impressed by this outstanding piece of
software. Yes, my version does appear to differentiate by dotted and
solid traces, but it's very hard to tell one from the other at
wide-spread sweeps at least.
>As a general rule phase normally increases but gain decreases with frequency
>so you can't really muddle them up.
>Anthony
Agreed, but I dare say there are exceptions. I keep finding myself
expanding the plots along the x-axis during the early part of their
traces to see the characteristic phase signature of a sideways 'Z',
mentally erasing it, and only then being able to read-off the trace I
want.
Still, all in all, very minor beefs in relation to such a fabulous
overall achievement.
anthony wooldridge
"Mike Engelhardt" <pm...@concentric.net> wrote in message
news:b45fl0$a...@dispatch.concentric.net...
> The reason that LTspice "differs from the windows
> standard" is that it differs from the windows
> standards used in editors. There you select first
> and then say what you want to do to it next. Most
> CAD editing tools are like that because most are
> hobbled together schematic editing tools with
> added-on simulators.
>
> LTspice is not. Its UI was carefully considered
> to be appropriate for an integrated schematic
> capture and SPICE. This makes it deviate from
> some editing conventions used in some windows
> software. For example, the default behavior
> in LTspice when you click on something in the
> schematic is to plot the data associated with it.
> The default behavior should not be to select it
> so that you can change the schematic while the
> simulation is running so that the relation
> between schematic and simulation is broken.
On that subject I notice that in simulation or xprobing mode,
which I love incidently, if you hover over a transistor you get the pointing
hand with invitation to right click and edit it, but not for any other
component ie resistor capacitor inductor etc. Why is that?
One could of course argue that you could have two different modes
of operation one for editing and one for xprobing. ie click and drag
works in editing mode but not in simulation or xprobing mode.
> The
> convention used by programs that edit only or
> have some junk SPICE hobbled on as an after
> thought is simply the incorrect one to use.
>
> The net effect of LTspice, with it's xprobing
> and marching waveforms, is to give an experience
> analogous to scoping out board while it is
> running. It breaks out of the confinement
> many SPICE programs give in having to carefully
> compose a simulation and then see what happened
> after it's done running.
Yes I love it and the blinding speed of the simulation.
Also the efficiency calculation and printout of power dissipations
is a really neat idea. Its a pity you can't use it without an LT switching
device on the circuit. I have to resort to the conivance of putting an LT
SW reg somewhere down in the corner of the circuit just to get that
list of power in each device. Can you please enable it for all circuits?
SuperSpice has gone one up on you there with a readout of power
as you hover over a component.
Anthony
Mike Engelhardt
> On that subject I notice that in simulation or xprobing
> mode, which I love incidently, if you hover over a
> transistor you get the pointing hand with invitation
> to right click and edit it, but not for any other
> component ie resistor capacitor inductor etc. Why is
> that?
It tells you on the status bar to right click to edit
for all devices. In the case of a resistor, you can
left click on the body to plot the current, so the cursor
turns into a ammeter to show you can click to plot a
current. On the transistor, you have to point a pin
to plot the current into that pin. Since you can't right
click on the body of the transistor, the cursor doesn't
turn into an ammeter. It shows the the hand which let's
you know you're pointed at a component but can't plot a
current just has you have the mouse positioned. No such
ambiguity exits for a 2-terminal device like a resistor.
> Yes I love it and the blinding speed of the simulation.
> Also the efficiency calculation and printout of power
> dissipations is a really neat idea. Its a pity you can't
> use it without an LT switching device on the circuit.
> I have to resort to the conivance of putting an LT
> SW reg somewhere down in the corner of the circuit just
> to get that list of power in each device. Can you please
> enable it for all circuits? SuperSpice has gone one up
> on you there with a readout of power as you hover over
> a component.
Efficiency of a SMPS doesn't have meaning if it's not
in steady-state operation. Detecting steady state is
somewhat non-trivial. Detection is written into the
models, usually by detecting the error amp's output
current going to zero as integrated over one switching
cycle. Note that it needs to know only the clock edges
but when the error amp is not in steady-state even at
no output current because it's pegged to a rail or turned
off. LTspice's efficiency also makes a note of the energy
stored in all the circuit reactances at the beginning
and end of the simulation and corrects the computed
efficiency by reckoning out the change in energy stored
in the circuit. I've found the simple-minded algorithms
one sees used in other SPICE programs as problematic.
--Mike
anthony wooldridge
> Anthony,
>
> > On that subject I notice that in simulation or xprobing
> > mode, which I love incidently, if you hover over a
> > transistor you get the pointing hand with invitation
> > to right click and edit it, but not for any other
> > component ie resistor capacitor inductor etc. Why is
> > that?
>
> It tells you on the status bar to right click to edit
> for all devices. In the case of a resistor, you can
> left click on the body to plot the current, so the cursor
> turns into a ammeter to show you can click to plot a
> current. On the transistor, you have to point a pin
> to plot the current into that pin. Since you can't right
> click on the body of the transistor, the cursor doesn't
> turn into an ammeter. It shows the the hand which let's
> you know you're pointed at a component but can't plot a
> current just has you have the mouse positioned. No such
> ambiguity exits for a 2-terminal device like a resistor.
OK I understand now.
>
> > Yes I love it and the blinding speed of the simulation.
> > Also the efficiency calculation and printout of power
> > dissipations is a really neat idea. Its a pity you can't
> > use it without an LT switching device on the circuit.
> > I have to resort to the conivance of putting an LT
> > SW reg somewhere down in the corner of the circuit just
> > to get that list of power in each device. Can you please
> > enable it for all circuits? SuperSpice has gone one up
> > on you there with a readout of power as you hover over
> > a component.
>
> Efficiency of a SMPS doesn't have meaning if it's not
> in steady-state operation. Detecting steady state is
> somewhat non-trivial. Detection is written into the
> models, usually by detecting the error amp's output
> current going to zero as integrated over one switching
> cycle. Note that it needs to know only the clock edges
> but when the error amp is not in steady-state even at
> no output current because it's pegged to a rail or turned
> off. LTspice's efficiency also makes a note of the energy
> stored in all the circuit reactances at the beginning
> and end of the simulation and corrects the computed
> efficiency by reckoning out the change in energy stored
> in the circuit. I've found the simple-minded algorithms
> one sees used in other SPICE programs as problematic.
Yes but still...
Why not include power calculations for all circuits?
If its not an SMPS then there's no problem. you just need the
operator to choose a start and a stop time over which to average the power.
Its pretty obvious looking at the waveform when the circuit has reached
steadty state.
I think it would be a really good feature.
One other query. Once you've done the circuit schematic and the simulation
is OK how do you get the design onto a PCB there seems to be no export
to PCB layout programs. Am I misunderstanding this? Is it in the pipeline to
incorporate this? Are there any PCB layout programs that read in .asc files
Anthony
Maxwell S. Diamond
wrote:
And here's another. Since I haven't tried them all I can't affirm it's
the 'best' but it certainly is a very good UI. Especially that
click-and-plot feature is very handy. I believe most of the griping
about non-standard comes from people who are too lazy to read up on
the help files and run the mouse across the tooltips and menu-items.
People seem to have gotten stuck on what they learned in the 'Word for
Dummies" course. Of course, any new software does take a bit of
getting used to, standard or not, nothing new there. But I've been
around since the DOS days when there were no standards at all, you
either got to grips with it on your own or gave up on it.
Paul Burke
>
> Yes, the editing deviates almost completely from the windows standard
> so it seems non intuitive at first - backwards, in fact.
I thought the only standard thing about Windows is that the standard is
whatever Microvirus have decided they are going to change the standard
to. (c.f. lightbulbs)
> > One other thing that bugs me is that when I run an AC sweep of a
> > circuit's output, I get two traces. One is gain v. frequency and the
> > other is - I can only assume - frequency v. phase difference. However,
> > they're both referred to by one node number and are both the same
> > colour, so very hard to tell apart especially at large frequency
> > spreads. Has anyone else had difficulties with this and is there a
> > way around it? I'm not particularly interested in phase difference
> > and fail to see the relevance of it.
One (the phase) dotted and the other (the gain) is solid on my version.
If Mike is reading, please could the trace display setup retain derived
traces (e.g. V(N005,N004)/V(N010) ) when the simulation command is
changed?
Paul Burke
Kevin Aylward
> The reason that LTspice "differs from the windows
> standard" is that it differs from the windows
> standards used in editors.
Were appropriate, SS does this as well. But for the most part, the
Windows standards makes sense, even for CAD. for example, there is no
need to use the ctrl modify for many command, so SS doesn't, unless
usefull.
>There you select first
> and then say what you want to do to it next. Most
> CAD editing tools are like that because most are
> hobbled together schematic editing tools with
> added-on simulators.
>
And some are written from scratch, specifically for circuit design,
based on the writers extensive experiance as an analogue engineer, not
as a software engineer.
> LTspice is not. Its UI was carefully considered
> to be appropriate for an integrated schematic
> capture and SPICE.
Oh...
This makes it deviate from
> some editing conventions used in some windows
> software. For example, the default behavior
> in LTspice when you click on something in the
> schematic is to plot the data associated with it.
I find this method rather annoying personally. I don't like locking out
schematic editing, but as it appears some may like this, and its only a
few
lines of code, I have today posted an update to SS supporting this
feature.
> The default behavior should not be to select it
> so that you can change the schematic while the
> simulation is running so that the relation
> between schematic and simulation is broken.
Again, your personal opinion. I find it very useful to do this, and do
it
all the time with no problems, and so disagree
with your statement on what the default behaviour "should" be. Indeed,
it is the method EWB uses, and many others, way before LTSpice existed,
so there is nothing
original in this method. Again, as a real analogue user, rather than
software
engineer, if I thought that this method was the "best" I would have done
it this way. I'm not daft. I have spent 10,000's of hours *using*
simulation tools rather than writing them. But I take your point,
*some* might like modal probing, so as noted above, I thank you for
re-highlighting this point out to me such that I have now included this
feature as an
alternative in SuperSpice.
The
> convention used by programs that edit only or
> have some junk SPICE hobbled on as an after
> thought is simply the incorrect one to use.
>
You do have this habit of trashing other products with unsupported
rhetoric. The
reality is that some others indeed design their GUI with a little
thought. However, analogue engineers rather than software engineers have
a
different view on what is the best GUI.
> The net effect of LTspice, with it's xprobing
> and marching waveforms, is to give an experience
> analogous to scoping out board while it is
> running.
I agree, that for technicians having virtual instruments and imitating a
lab bench can make using a spice easier for those less experienced.
It breaks out of the confinement
> many SPICE programs give in having to carefully
> compose a simulation and then see what happened
> after it's done running.
Some of us have dreadful memories, and therefore like to modify values
while the simulation is still current, ready for the next simulation,
but still analysing the current simulation. A general "good" practise is
to to not unduly restrict what a user can do, unless necessary.
>
> And it still uses fewer key and mouse strokes
> for editing because usually one needs to move
> a bunch of things around not do may different
> editing operations on one set of selected
> objects.
But the windows standard of enclosing and and uses the fewest key
strokes.
In addition, multi-selecting and changing say, all W and L of a set of
mos in one go is a feature that is pretty much mandatory in pro cmos
design. A feature supported in Cadence analog artist and SS, and indeed
used extensively.
>Besides, the select first software
> clears the selection after the operation so
> there is no possibility to save key/mouse strokes
> anyway. That "convention" really seems to be
> made to insult the users' intelligence. Anyway,
> LTspice's UI stands up to expert use by people
> who edit in it all day long day after day, year
> after year.
>
Oh? You don't say...
This would seem to stretch the facts a bit. As was pointed out a few
months ago. At that time LTSpice had no hierarchy support. With a GUI
circuit
size limited to what would fit on one rather small page, the use of
LTSpice in that condition was arguble somewhat limited for pro use.
Indeed, you admitted that the analogue engineers at LT used an
alternative GUI, only using LTSpice as the engine.
> Now, I find it a perverse form of narcissism
> to get enthralled in one's own UI so I don't
> usually reveal my opinion on the subject.
Oh...You obviously don't read your own posts then.
> However, I do believe that LTspice's UI is
> better than any other out there.
Good for you. What about Santa Claus?
>It can't be
> immediately intuitive for everyone because not
> everyone comes from the same background and
> even fewer have seen an integrated Schematic
> capture/SPICE program. We get *much* more
> feedback from users that LTspice is intuitive
> than it is not intuitive.
Whilst, SS don't have 100,000 downloads, the % that say it is
non-intuitive is < 1%.
> It also uses many
> techniques to help one quickly develop
> intuition with the program.
>
Ahmm...
Mike Engelhardt
If you're so sure of your product why do you
have to defend it in LTspice threads that don't
mention it? Anyway, it's clear you haven't
spent much time with LTspice or you wounldn't
have made so many errors in what you're saying.
LTspice never locks you out of editing a
schematic. In fact it certainly does let you
edit schematics in typically fewer mouse and
keystrokes.
> > ME:
> > Besides, the select first software
> > clears the selection after the operation so
> > there is no possibility to save key/mouse strokes
> > anyway. That "convention" really seems to be
> > made to insult the users' intelligence. Anyway,
> > LTspice's UI stands up to expert use by people
> > who edit in it all day long day after day, year
> > after year.
>
> KA:
> Oh? You don't say... This would seem to stretch the
> facts a bit. As was pointed out a few months ago.
> At that time LTSpice had no hierarchy support. With
> a GUI circuit size limited to what would fit on one
> rather small page, the use of LTSpice in that
> condition was arguble somewhat limited for pro use.
> Indeed, you admitted that the analogue engineers at
> LT used an alternative GUI, only using LTSpice as
> the engine.
You are mistaken. LTspice has been used for four
years as a single sheet tool. Many SPICE
investigations can be done on a single sheet. You
probably realize that or you won't have done
something as stupid as putting waveform data on a
schematic page! Anyway, in the past, when LTspice
was publicly released as a single sheet schematic
the IC designers used the released engine and waveform
viewer for their own GUI's. The waveform viewer's
GUI was critical for its small files due to compression
and the extension of virtual address space to 64bit
on a PC and now also on Linux.
Anyway, it seems there's no hope to talking reason
with you. I'm not out to get you. I really don't
care what kind of product you want to try to sell
people. Yes, I realize that people can get a better
program for free, but hey, let the buyer beware,
right?
--Mike
Kevin Aylward
> Kevin,
>
> If you're so sure of your product why do you
> have to defend it in LTspice threads that don't
> mention it?
The US are so sure that that they can beat Iraq, so why do they feel
they have to defend themselves from Iraq?
Get real, you continually make biased, misleading information. I am
correcting this misinformation. I'm not daft, neither are the other
readers of these posts. Do you really think that they are silly enough
to not realise that many of your "LTSpice is better than these other
unnamed spices" is a reference to specifically include SS. Your so
naive. Everyone can see through what your trying to do. Its plain as the
nose on Michael Jackson's face. "LTSpice....Historic occasion...blah
blah...". You truly believe all of this drivel don't you. Your what we
in the UK call a right bloody wanker.
>Anyway, it's clear you haven't
> spent much time with LTspice or you wounldn't
> have made so many errors in what you're saying.
> LTspice never locks you out of editing a
> schematic. In fact it certainly does let you
> edit schematics in typically fewer mouse and
> keystrokes.
>
Fewer than what other spice? You obviously have not spent much time with
SS or "you wouldn't have made so many errors in what you're saying",
especially in spelling "wounldn't".
Translation of "stupid". "LTSpice dose not support this feature".
I hardly ever use this feature myself, but the main motivation for it
was other competition that also has this feature. Its called business.
ever use that fe
> Anyway, in the past, when LTspice
> was publicly released as a single sheet schematic
> the IC designers used the released engine and waveform
> viewer for their own GUI's.
So you agree, I was not mistaken. IC designers used a *different* GUI.
> Anyway, it seems there's no hope to talking reason
> with you.
Indeed, as *you* are unable to talk reasonably. Its all "I'm great,
everything
else is crap". Go and read your posts. Its plain as day. You don't budge
an
inch.
>I'm not out to get you. I really don't
> care what kind of product you want to try to sell
> people.
Then why criticize it.
>Yes, I realize that people can get a better
> program for free,
Dream on.
Kevin Aylward
You
>> probably realize that or you won't have done
>> something as stupid as putting waveform data on a
>> schematic page!
>
> Translation of "stupid". "LTSpice dose not support this feature".
>
> I hardly ever use this feature myself, but the main motivation for it
> was other competition that also has this feature. Its called business.
>
Oh bugger... I forgot the other reason why you put your foot in it here.
The other motivation for this was the annoyance that *all* other spices
don't allow one to manually position and size the graph areas. It can be
useful on the dedicated *graph* views to place different types of run
data graphs together, e.g. AC and TRAN. Oh shit, LTSpice can't run
multiple runs...oh well...
The graph pages and schematic pages, although they have different
default behaviour, have the same base classes, so having graphs on the
schematic is a freebee. In fact, pasting schematic sections on to the
dedicated graph views can be useful for documentation purposes. Indeed,
I have had specific feedback on the attractiveness of this feature.
Actually, I remember that if you tried pasting components in the Cadence
Analog Artist waveform window, it crashed it.
I suppose the fundamental issue here is that whilst I have spent around
20 years as a professional analogue engineer, you have spent the same as
a professional software engineer.
Mike Engelhardt
> I suppose the fundamental issue here is that
> whilst I have spent around 20 years as a
> professional analogue engineer, you have
> spent the same as a professional software
> engineer.
I sure hope that I'm both a more accomplished
analog engineer and software developer than you.
While I've been writing simulators on and off
for over 25 years, I have done my share of
analog design. One company did a stock
split due to the significance of one of my
analog circuits.
But why get into who you and I are in a thread
about running LTspice on Linux? Why don't
you ask yourself if what is going on is that
you feel implicitly threatened because
there's a better program then you try sell
that available for free? There could be a
worthwhile spiritual experience waiting for
you.
--Mike
Kevin Aylward
> Kevin,
>
>> I suppose the fundamental issue here is that
>> whilst I have spent around 20 years as a
>> professional analogue engineer, you have
>> spent the same as a professional software
>> engineer.
>
> I sure hope that I'm both a more accomplished
> analog engineer and software developer than you.
Indeed, hope is what you need.
>
> But why get into who you and I are in a thread
> about running LTspice on Linux? Why don't
> you ask yourself if what is going on is that
> you feel implicitly threatened because
> there's a better program then you try sell
> that available for free?
Any business would be daft to not recognize competition. I am not daft.
An issue here is not on what program is better, it is what is free or
not. People simply do not evaluate freebees in the same way as they do a
purchased product. They will accept many limitations and colour their
judgment without even being aware of it. As they say, the only valid
marketing survey is a signed purchase order. You claim that LTSpice, is
the best simulator in every shape and form. If this was indeed the case,
I would not get any sales at all from people who have used LTSpice. The
reality is, is that if LTSpice cost, say $300, and did not have the
backing of a 1/2 billion dollar per year company, I doubt that it would
get many sales, and that is the real proof of a products value. Feel
free to have your comfort that people pick up on something you give away
for free.
Kevin Aylward
> Kevin,
>
There could be a
> worthwhile spiritual experience waiting for
> you.
There are no "spiritual " experiences. This is just fantasy invented by
peasants 1000's of years ago who knew no better. Everything we think do
and say is a direct result of the fact that we are manufactured from the
most successful replicators generated by random chance. We are
biological machines, nothing more. There is no "higher" understanding of
life.
Mike Engelhardt
> > But why get into who you and I are in a thread
> > about running LTspice on Linux? Why don't
> > you ask yourself if what is going on is that
> > you feel implicitly threatened because
> > there's a better program then you try sell
> > that available for free?
>
> Any business would be daft to not recognize competition.
> I am not daft. An issue here is not on what program is
> better, it is what is free or not. People simply do not
> evaluate freebees in the same way as they do a purchased
> product. They will accept many limitations and colour their
> judgment without even being aware of it. As they say, the
> only valid marketing survey is a signed purchase order.
> You claim that LTSpice, is the best simulator in every
> shape and form. If this was indeed the case, I would not
> get any sales at all from people who have used LTSpice.
> The reality is, is that if LTSpice cost, say $300, and
> did not have the backing of a 1/2 billion dollar per
> year company, I doubt that it would get many sales, and
> that is the real proof of a products value. Feel free to
> have your comfort that people pick up on something you
> give away for free.
I doubt the cost of a program has much to with its quality.
I sure do hope that LTspice is the best SPICE program,
but I've never claimed it to be the "is the best simulator
in every shape and form" Heck, if there weren't any more
simulators to write I'd be put out of business.
Also, I'm convinced the best way to fund SPICE development,
or any simulator, is *not* by selling licenses for the
software. The trend that I've experienced is that best
simulators are developed by the concerns who need them.
It's all but unheard of for such simulators to be given
distributed free -- let alone even disclose they exist --
so LTspice is in a very special situation.
--Mike
Mike Engelhardt
> > Mike Engelhardt wrote:
> >
> > There could be a worthwhile spiritual
> > experience waiting for you.
>
> There are no "spiritual " experiences. This is just fantasy
> invented by peasants 1000's of years ago who knew no better.
> Everything we think do and say is a direct result of the fact
> that we are manufactured from the most successful replicators
> generated by random chance. We are biological machines,
> nothing more. There is no "higher" understanding of life.
A spiritual experience can be a subtle as a change in perspective.
--Mike
Kevin Aylward
Sure, if you want to change the usual meaning of "spiritual".
Dale Chisholm
programs, but it's not so much different that you can't learn it -
even by trial-and-error methods - in a short time. It DOES have
similarity to other CAD/CAE tools I've used: CAM350 and Cadence
Capture. People with only modest intelligence - certainly less
intelligence than necessary to design analog circuits - can commonly
write in either the cursive, or the printed handwriting styles. They
have little difficulty switching from one to the other. I fail to see
how the fairly minor differences in UI's amount to anything as
significant as the differences in the two handwriting styles. The
effort expended by some people to either "defend" or "attack" the
LTSpice UI strikes me as, frankly, quite childish.
Can the UI be improved? Perhaps. Certainly changed, but not all
change is an improvement. It may be years before a change is
confirmed to be an improvement, and then not everybody may agree on
that. Furthermore, the tradition in software is that changes are
(almost) never reversed, even when they subsequently prove to have
little value or even to be detrimental.
Can I suggest some changes? Sure - but I suspect all of these would
complicate the code, without really adding any features or
funtionality:
- Create 2 (or more) editing conventions, maybe called "Windows
Mode" and "CAD Mode", selectable from the " Tools > Control Panel "
menu.
- Allow keyboard operations to be re-defined via a macro language.
This is really a superset of the above idea, and I'm sure the
respective C++, BASIC, EMACS, etc camps will check in with their
respective versions of what an "intuitive" macro language should look
like.
- Allow LTSpice to have, say, a "Scematic Edit" mode and a "Circuit
Simulate" mode. The responses to actions like left click, right
click, hover, etc could be different in the two modes. If no waveform
data was available, the program would be in "Edit" mode; if a
simulation was in progress it would be in "Simulate" mode; and at
other times you could toggle betweem them with the "ALT" key for
instance. Of course, there would have to be a banner message, or
change of background color, or other means to show what mode you were
in . . .
See how complicated it gets? And the program really isn't any more
useful when it's all said and done.
I'd rather see Mike put his time into other things.
Dale
Maxwell S. Diamond <ma...@notmyself.com> wrote in message news:<s0sc6v86cs5rc52sq...@4ax.com>...
> - - - snip - - -
Robert Hickey
"Kevin Aylward" <ke...@anasoft.co.uk> wrote in message
news:NVlaa.759$t4.69583@newsfep2-gui...
> There are no "spiritual " experiences. This is just fantasy invented
by
> peasants 1000's of years ago who knew no better. Everything we think
do
> and say is a direct result of the fact that we are manufactured from
the
> most successful replicators generated by random chance. We are
> biological machines, nothing more. There is no "higher"
understanding of
> life.
Since I know of no way to prove what you said (perhaps you do) then it
must be regarded as an "item of faith".
And while I don't happen to share that particular "religion" I don't
object to yours. As you seem to object to others.
Robert
Kevin Aylward
> "Kevin Aylward" <ke...@anasoft.co.uk> wrote in message
> news:NVlaa.759$t4.69583@newsfep2-gui...
>> There are no "spiritual " experiences. This is just fantasy invented
>> by peasants 1000's of years ago who knew no better. Everything we
>> think do and say is a direct result of the fact that we are
>> manufactured from the most successful replicators generated by
>> random chance. We are biological machines, nothing more. There is no
>> "higher" understanding of life.
>
> Since I know of no way to prove what you said (perhaps you do) then it
> must be regarded as an "item of faith".
I agree, that in the limit, all has a faith based component. However, do
you want to accept 2 or 3 axioms, or 10234. Furthermore the axioms of
evolution are more easily verified.
>
> And while I don't happen to share that particular "religion" I don't
> object to yours. As you seem to object to others.
>
I object to people ramming it down ones throat. The fact that anyone
puts forward any statements based on religion, opens up the
justification for refuting those statements. If someone wants to believe
in religion privately, then he wont get someone attacking that view.
Religions universally try and an convert others, its in the design of
that particular meme. Its fair game for atheists to do likewise.
Bill Sloman
> "Kevin Aylward" <ke...@anasoft.co.uk> wrote in message
> news:NVlaa.759$t4.69583@newsfep2-gui...
> > There are no "spiritual " experiences. This is just fantasy invented
> > by peasants 1000's of years ago who knew no better. Everything we think
> > do and say is a direct result of the fact that we are manufactured from
> > the most successful replicators generated by random chance. We are
> > biological machines, nothing more. There is no "higher"
> > understanding of life.
>
> Since I know of no way to prove what you said (perhaps you do) then it
> must be regarded as an "item of faith".
Kevin's attitude is reductionist, and classifying "spiritual"
experiences with other conditions - delusions, hallucinations,
paranoid fantasies - experienced by imperfectly functioning human
brains has the great virtue of conforming to William of Occam's
principle of economy in explanation.
If you want to differentiate "spiritual" experiences from other - more
obviously irrational - changes of mood that can't be directly coupled
to changes in the outside world, the burden of proof is on you, rather
than Kevin.
His attitude can't be dismissed as an "item of faith" - it is entirely
consistent with what little we know about how the mind works.
------
Bill Sloman, Nijmegen
Haines Brown
Spice. But when I tracked the link in the original article, I just
became bewildered by all the varieties and files that are not
executables.
Which Spice, hopefully that will run under Wine, is good for simple
circuit analysis?
--
Haines Brown
bro...@hartford-hwp.com
kb1...@arrl.net
www.hartford-hwp.com
Haines Brown
a evengelical minister, I can claim expertise in these matters ;-)
Any argument arises from a set of initial axioms, and these are in
turn supported by other arguments resting on their own axioms. Because
of this infinite regression, an argument carries weight only for those
who happen to share biases and ideology. But do we have to accept this
sad thesis? It seems to imply a radical subjectivism or solipcism,
which no one really wants.
Is it possible to for us all to reach some agreement? One approach
might be to suggest there are probably certain criteria that are
applicable to any body of knowledge, whether or not it rests on
naturalistic or supernatural presuppositions.
For moral and existential reasons, I happen to prefer naturalistic
knowledge and here's why, briefly. I believe that any body of knowlege
ought to be coherent if it is to represent a body of knowledge. But
since "bodies" of knowledge are rather artificial constructs, it
really means that all knowledge should be coherent. When all knowledge
is coherent, it means a necessary reconciliation of the contradiction
between natural and supernatural; coherence seems basic for effective
action in the world and offers a basis for moral behavior.
This is hardly an original point. It is what distinguishes European
Renaissance science and prior Medieval science. Renaissance science
was less "scientific" in that it injected platonic supernaturalism
into a long and highly developed tradition of "cook book" science (of
which DaVinci was an exemplar), and it unfortunately represents the
emergence for the first time in Europe of an elitist culture (the
point of science not being practical effectiveness, but superior
intelligence). However, it had a powerful virtue. The platonic
mysticism implied that all our particular knowledges have a
determinate relationship because the reality they describe was
presumed to be mystically coherent. What is learned about one field is
likely to have implications for other fields.
In other words, modern science arose from a supernatural axiom! The
resulting coherence of all knowledge could support moral
responsibility and effective worldly action, and so might be
preferred, but an interesting question is whether that supernatural
core remains important today.
This is actually a difficult question to answer. The initial
singularity associated with Big Bang cosmology might suggest a
naturalistic cosmic coherence in all that derives from it. But
historically, the bringing of science to earth seems associated with
the bourgeois revolution. However, close attention suggests a mystical
core remained despite it (Newton was a mystic, for example). I suspect
it was the ideological success of the Second Industrial Revolution
that finally put the issue to rest because it its power carried with
it the paradigm that successful practice validated the presumed
coherence of scientific knowledge from which it arose.
But of course my wife denies there was any such singular event as the
Big Bang and cares not at all for the wealth generated by the Second
Industrial Revolution, and so where are we? For the Christian or
Moslem fundamentalist (actually, "fundamentalism" is the wrong word in
Islam), one substitutes for the Platonic coherence of all reality the
singularity of god, to which all else must or should conform. This
primacy of faith in a godhead seems to make perfectly good sense in
itself. In fact, effective action (called "prosperity" by evangelicals
here) is expected to result, and moral behavior is firmly based a
literal interpretation of god's Word. That evangelicals often do not
prosper or behave morally is irrelevant.
In short, I don't think the believer-non-believer argument can ever be
resolved. I don't think much of agnosticism, but I don't know how true
believers of the naturalisic and supernaturalistic type can
communicate on this matter. My effort to disprove my initial thesis
has failed.
Paul Burridge
<bro...@hartford-hwp.com> opined thusly:
>Back to the thread. I run Linux and would like to use Wine with
>Spice.
That may be why I've been having so much trouble with Spice up until
recently.
:-)
Mike Engelhardt
"Haines Brown" <bro...@hartford-hwp.com> wrote in message
news:m2k7f46...@hartford-hwp.com...
> Back to the thread. I run Linux and would like to use Wine with
> Spice. But when I tracked the link in the original article, I just
> became bewildered by all the varieties and files that are not
> executables.
>
> Which Spice, hopefully that will run under Wine, is good for simple
> circuit analysis?
It's called LTspice/SwitcherCAD III availible from www.linear.com/software
Below was the original post.
--Mike
---- snip ----
LTspice/SwitcherCAD III now runs on Linux under WINE.
This is an unlimited free schematic capture/SPICE
simulation program. Attached is the relevant part
of the FAQ from the help system:
Q. Do you have a version of LTspice for Linux?
A. Not a separate edition, but it does run under
WINE. The program has been tested on Linux
RedHat 8.0 with WINE version 20030219.
Q. OK, I've never used WINE, how do I install that
on my Linux box?
A. 1. Check with www.winehq.com to find the current
version of WINE for your system. At the time
of this writing, for RedHat 8.0, this points
to http://mecano.gme.usherb.ca/~vberon/wine
2. Copy the appropriate .rpm file to your machine
and open it from nautilus.
3. Get the file swcadiii.exe from www.linear.com.
In an xterm, execute "wine swcadiii.exe" to
install LTspice.
4. There will now be a Linear Technology Logo on
your gnome desktop. Double click it to start
or type "wine scad3.exe" from an xterm to
start the program. That's it!
Q. The schematic fonts don't scale as smoothly under
WINE as Windows. Why is that?
A. WINE is doing the best best it can with the fonts
it finds. It will do better if you tell it how
to find the files arial.ttf and cour.ttf from
your Windows system.
Q. The PWL additional point editor doesn't look
right under WINE?
A. Try using the native Windows .dll from your
Windows system. The command line to then invoke
LTspice from WINE is:
wine -dll commctrl,comctl32=n scad3.exe.
Q. It seems LTspice is running slightly differently
under WINE/Linux than windows. Why is that?
A. LTspice detects whether or not it's running under
WINE. If so, it works around a few WINE issues.
You can force LTspice to think it's running under
WINE with the command line switch -wine. You can
force it to think it's not with the command line
switch -nowine in case you're interesting in
debugging WINE.
Q. Under Windows, LTspice extends the virtual
address space for waveform viewing to 64 bits.
Does that work under WINE?
A. Yes. It has been tested on waveform files over
5 Gigabytes in size.
Q. Under Linux, does it support unlimited component
count and depth of schematic hierarchy?
A. Yes. Opps, no. The maximum depth of hierarchy
is 64 levels but that limit is just there to
allow detection of infinite subcircuit recursion.
(Most "unlimited" SPICE's "limit" you to about 21
levels.) LTspice has no limit on node or
component count either per page or for the
entire circuit.
Q. Does cross-probing while simulation work with
those slick marching waveforms while running
under Linux?
A. Yes.
Q. So from what version on is LTspice supported
under WINE?
A. 2.01g.
--Mike
PS. Thanks to UB and CC for their inputs.
Mike Engelhardt
mention the term "spiritual experience", let me
point out what I said that caused a debate about
religion:
> I wrote:
> A spiritual experience can be a subtle as a change
> in perspective.
See the problem. God isn't mentioned. Nothing is.
You can believe in God and not disagree and you can
not believe in God and not disagree. It's about as
self-evident of truth as saying, "You can tell when
you've hit bottom by when you've put down the shovel."
It's just the sort of saying that doesn't say anything
but can mean alot.
--Mike
Mike Engelhardt
> > > Hi Mike,
> > >
> > > I really like LTSpice, but don't have any function keys on my
keyboard.
> > > Would it be possible to have remappable keystroke commands?
> >
> > Yes, this is something I've been needing to do for awhile.
> > Can't promise a release date, though. It shouldn't be too
> > hard now after what I needed to do to get it running under
> > Linux/WINE.
OK, version 2.01j, availible now has user-definable hotkeys.
Goto Tools=>Control Panel=>Drafting Options=>Hot Keys
--Mike