Reply-To: "Kevin Aylward"
From: "Kevin Aylward"
References: <email@example.com> <firstname.lastname@example.org> <email@example.com> <12j4LzC+fEp9EwgG@jmwa.demon.co.uk> <firstname.lastname@example.org> <699p9.8961$QY.email@example.com> <firstname.lastname@example.org>
Subject: Re: Problems with Electronics Workbench - timestep errors
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
X-Inktomi-Trace: public1-pete2-5-cust19.pete.broadband.ntl.com 1034261928 29857 188.8.131.52 (10 Oct 2002 14:58:48 GMT)
Date: Thu, 10 Oct 2002 15:58:49 +0100
NNTP-Posting-Date: Thu, 10 Oct 2002 15:58:49 BST
"fred bartoli" wrote in message
> Kevin Aylward a écrit dans le message :
> > "Roger Johansson" wrote in message
> > news:email@example.com...
> > > "Kevin Aylward" wrote:
> > >
> > > >Sure, you can connect up signal generators and scopes and see all
> > > >little bitty pretty pictures, if that makes you happy, but you
> > be
> > > >doing electronics.
> > >
> > > I think that millions of people have worked all their lives with
> > > electronics without even having heard about spice simulators,
> > > netlists, etc..
> > >
> > I was referring to showing pictures in a *spice simulation* with no
> > of the principles of how those pictures came about. If you are a
> > in electronics, then you will have to lean what frequency response
> > time domain response actually means. Once you understand the
> > of what you are doing, running SS is trivial. If you don't know
> > or little about, how can you possible expect to run a simulator
> > a bit of a learning curve. Its exactly the same as the learning
> > a breadboard, it just that the graph screen is not, usually a scope,
> > actually something easier to use.
> > > These people, and service engineers, and students, amateurs, ham
> > > builders, etc will have no problems using EWB, without reading one
> > > single line of docs, without learning anything.
> > >
> > But that simple not true. There is no need to know about netlists to
> > SS. Furthermore I have had many complete novices run SS with no
> > at all.
> > > That is the beauty of EWB, people with knowledge about electronics
> > > using computers can use it straight away, and have a working
> > > within a minute.
> > I don't agree at all. I an expert in using simulators and I found
> > essentially impossible to use. I knows exactly what simulators
> > do, and most variants of how they do it. In EWB, many things were
> > not guessable at all. I would say that additional support for this
> > is that, imo, is that in these NG's EWB often gets a trashing.
> > > To use your program they would need to learn a lot about netlists,
> > > learn a whole new set of terms and how to use them.
> > >
> > As I noted, this is simple *not* true. You simple place components.
> > up runs in the GUI, e.g. stop start run times, press run and all
> > get displayed. Its a no brainier. As I said, you must know enough
> > electronics to know that if you want a signal of 5Volts at a
> > of 1KHz, you have to tell the source that! A simulator can not read
> > mind.
> > > EWB users would be doing spice analysis without even knowing the
> > > "spice".
> > >
> > > Of course they could do a lot more sophisticated and specialized
> > > simulations if they had that knowledge, but with EWB they at least
> > > start working with computer simulations without any initial
> > Not in my book. As I noted, I had amazing difficulty getting EWB to
> > Why do you think that I started SS in the first place?
> > > If you are used to get by with such simple tools as oscilloscopes,
> > > voltmeters, function generators, you feel no need for the more
> > > sophisticated tools your program offers.
> > >
> > Ahmmm. The SuperSpice graph is easier then a scope. Simple moving a
> > point on to wires and pins displays volts and current. This is far
> > easier then connecting up voltmeters. The superspice source *are*
> > function generators, they just don't look like them.
> > > A sophisticated spice simulator could very well have such an easy
> > > use interface even if it also had a lot more sophisticated tools
> > > the user could learn later.
> > >
> > > There is no good reason for omitting the simple user interface
> > > is relatively easy to implement.
> > SuperSpice *does* have a trivial simple user interface. However, it
> > not one designed around virtual instruments.
> > > It would be like writing books with edlin just because nobody
> > > to create a modern text editor.
> > >
> > You have still missed the point. Spice without virtual instruments
> > *the* *modern* way of doing spice simulations. It is a *superior*
> > method. People brought up on scopes and meters need to get out of
> > *dated* mindset. Vendors introduce virtual instruments not because
> > a better way to do spice simulation but because people like to stick
> > with what they know, not what is best.
> > > The first professional spice simulator which offers such a simple
> > > interface will take over a big part of the market.
> > In all honesty, you just don't have any idea of what simulation is
> > about.
> > > The potential market is a lot bigger than just the few
> > > designers who can use spice simulators without such an interface.
> > >
> > As I said, once the novices user grows up a bit and rids himself of
> > preconceived ideas he can get on with the job of real electronic
> > analysis.
> > I should point out here, that while SS is indeed easy to use, imo,
> > simple *not* designed as a student *instruction* tool. It is
> > for doing real analogue design work.
> > > You used the word "must" extensively in your message.
> > > A lot of people would like a world with less "musts" in it.
> > Because must was the right word to use. If you *want* to do and
> > understand electronics you *must* understand the very raw, basic and
> > trivial stuff like what is a frequency response, what is a time
> > response, how to I generate signals to do these. Of course if you
> > *don't* want to understand electronics, then none of this will apply
> > you.
> Coming late in this thread to support you Kevin (not meaning that you
Thanks. I don't often get support on this.
> Have read all the comments above and in the other posts and I 2000%
> agree with you and *must* say to little lost angel and Roger that they
> Sticking with scopes... only show reluctance to change one's way of
> and habits. Designing serious analog stuff require you to do this very
> often. Once you'll changed to use
> I also "tried" EWB one or two years ago, and decided as soon as
> that it's just eligible for the trash can.
> Setting up a schematic is 10 times too long. The models library is
> ridiculously low quality, and the *necessary* features for serious
> lacking, at the "benefits" of useless scopes, meters and so on. For me
> a strictly marketing positionning from a looser firm that *can't*
> build a good enough simulator for the professional market and then try
> make money on credulous amateurs. I've seen a lot in the spice
> domain but I must admit that EWB is the worst I've ever seen.
> To be honest in comparing both, I've just downloaded SS and gave it a
> 3 min for download
> 1 min for installation
> 2 min to play with the menus, toolbars, library structure, options...
> a quick feeling of how SS is designed...
I gather you ran a few examples to see how the "magic" works.
> then 2 min to setup a SK LP filter with a 2N3904 buffer, simulate its
> transfer function, extract the poles and zeros, then build an
> laplace block and so on.
I have not done a pole-zero graph yet, principle because I personally
have never found a use for one.
There is a set of Laplace filter models already in the
\SuperSpice\system folder, but they are not loaded by default, there are
also other vendors models files that are also not loaded by default.
Just drag and drop will load them in.
> Oh, the 2n3904 model was also created in a snap with parameters from
> library, not using Kevin's ones, as I wanted to compare the results
> one I'm used to.
> Definitely simple to use and intuitive. I also reckon that it's a nice
> from a single man that also simulate 40hours a day. Maybe you wrote SS
> during the other 7 days of the week ?
By and large, it was all done whilst I had a day job. Mostly, using
Cadence Analogue Artist doing analogue i.c. design. 600+ files, around
100k active lines, difficult to count, a few files have 20k lines
setting up various data structures.
And of course, software is only a hobby of mine:-)
> FYI, I've noticed some easy to implement small features that could
> schematics capture speed, though I did play with SS only for something
> 30min, so I probably don't have all the good reflexes with SS. If
> interested I could more thoroughly "review" it and give you some
> (from someone that only does simulation 8 hours a day).
I'm always open to feedback. I just put in page borders and a title
block because I as being nagged by a user.However, I like to put in real
key features rather then fluff. I do need to have a single right mouse
click to quick change the waveform colour, some of the defaults are too
> To little lost angel and Roger, again SS is a much more valuable
> than EWB for amateurs (and it's free for small sized projects)
> you have the willing to (only slightly) change your way of thinking.
Indeed. As I noted, the 30 top level limit on a second lower level limit
of 20 per schematic is 600 components!. Real pros, imo, design
schematics object orientated and hierarchically. E.g. A PLL might be a
current pump, a phase detector, a bias block, a divider block, a loop
filter block etc.
> will learn much more this way, but perhaps it's not your point after
Ahhhh at last some sense!
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.