The Cyber-Spy.Com Usenet Archive Feeds Directly
From The Open And Publicly Available Newsgroup
This Group And Thousands Of Others Are Available
On Most IS NNTP News Servers On Port 119.
Cyber-Spy.Com Is NOT Responsible For Any Topic,
Opinions Or Content Posted To This Or Any Other
Newsgroup. This Web Archive Of The Newsgroup And
Posts Are For Informational Purposes Only.
Reply-To: "fred bartoli"
From: "fred bartoli"
References: <firstname.lastname@example.org> <email@example.com> <firstname.lastname@example.org>
Subject: Re: Flash a string of LEDs
Date: Thu, 5 Dec 2002 18:15:02 +0100
X-Newsreader: Microsoft Outlook Express 5.00.2314.1300
Organization: Guest of ProXad - France
NNTP-Posting-Date: 05 Dec 2002 18:12:03 MET
Terry Pinnell a écrit dans le message :
> "fred bartoli" wrote:
> >Yes, the same. I've just set the cap value to 22uF as I recall something
> >like that from an earlier exchange between you and Spehro. However it
> >not change anything about the behaviour, except the flash period of
> >The simulation time is set to 10 seconds.
> >The time step has nothing to do with simulation.
> Not sure I understand that, Fred? If I accept the defaults that CM
> offers, of 5 uS and 20 nS for step and stop respectively, then I
> obviously won't see much! As with most of my sims, I have to set it
> manually. In this case (having reduced cap to 22 uF) I chose a stop
> time of 10 S and a step time of 10 mS. These have a very significant
> effect on whether I see the result in a second or two, or a few days
> But, apart from the *display* (which is maybe what you meant?),
> doesn't the step value also affect simulation itself? I frequently get
> 'timestep too small' or some such message. And I'm pretty sure I
> recall changing this value has sometimes got an oscillator simulating
> (or stopped it).
Now I can see the dialog box on the URL your provided, so :
You have 2 parameters. Timestep, which is the one that I described and has
no effect, as long as your simulation run. The time step too small error may
some time be suppressed by reducing this value as it is a basis of
comparison (with a 10^-?? factor applied) to decide whether the time step is
too small or not. Other than that, it has *no* effect on the simulation (the
algorithm that give the integration time step is an auto adaptative one).
The maxstep parameter allow you to constrain the *max* value of the
integration time step, in order to see some events the adaptative algo could
miss in some special cases.
> >It has effects after the
> >simulation is done and gives the sampling period of the raw datas,
> >the simulation is done, for the ascii output file. I don't know about CM
> >I guess almost all todays spices don't use this file as the interface
> >the scope program as this leads to lots of artifacts problems
> >(undersampling...). So if CM directly use raw data, simply set TimeStep
> >reasonably low value (don't generate huge text files) and all should be
> >You can test this by setting a simple simulation file with a sine wave
> >running 2 sims varying the timestep parameter and looking at what you
> >in your scope.
> >For your simulation still not working don't forget that the oscillator,
> >all, has a start up time (in my case with 22uF it's about 5 seconds), so
> >enough simulation time.
> >Hmmm, wait a minute...
> >Ok, that's it. I've set the cap to 220 uF. It also refuse to start and
> >instead gently settle to the DC bias point I obtain without using the UIC
> >Now build an AC analysis to measure the circuit loop gain (there was a
> >recent thread about this but I can't remember what it was). Then vary the
> >cap and see what happens to your loop gain.
> >In my case with a 22uF cap the loop gain goes above 0dB somewhere -> the
> >circuit starts.
> >When setting the cap to 220uF the loop gain *never* go above 0dB -> no
> >Then I set the 1uF cap to 10uF, the loop gain rise again above 0dB
> >and the circuit starts again.
> >Setting back to 220uF/1uF but increasing the 220R value to 1K, the loop
> >rise again above 0dB somewhere and the circuit still starts.
> Thanks for taking the trouble, Fred. It's made me even more determined
> to get it working (especially as it works so well on breadboard, even
> with 220 uF). But as you see from
> it still eludes me!
Ha... it's not that your flasher don't start, but that it (in CM) stop
Let me think a little to this.
> Terry Pinnell
> Hobbyist, West Sussex, UK
Go Back To The Cyber-Spy.Com
Usenet Web Archive Index Of
The sci.electronics.design Newsgroup