When I get time, I'll take a look at it. It may be awhile. I just
stared a new job in a different city, so I am in the process of getting
relocated.
Chuck
On 1/25/2013 8:56 AM, mike brinson wrote:
> Hello Chuck
>
> Thanks for your email.
>
> Stefan Jahn implemented the HB version which is currently coded in Qucs.
> Stefan is no longer an active member of the Qucs development team.
> The Qucs implementation is not complete and is rather unstable - One
> problem is its convergence
> algorithms are not adequate. The code is well documented by Stefan -
> this explains its limitations.
> Similarly most of the device models are not implemented for HB analysis.
>
> My interest is in the modelling - which I am looking at with a view to
> improving in the near future.
>
> if you have the knowledge concerning implementation of HB would you
> have a look at the Qucs code,
> and if you can, help improve, and stabilise, the current code.
>
> Thanks for your interest.
>
> Mike
>
>
> Mike Brinson
>
> mbrin72043@...
> ------------------------------------------------------------------------
> *From:* Chuck <cmcook3rd@...>
> *To:* mike brinson <mbrin72043@...>
> *Sent:* Friday, 25 January 2013, 0:24
> *Subject:* Re: [Qucs-devel] Bug/logic oddity in Transient solver loop
>
> The basic harmonic balance equation just solves the continuity
> equation for charge.
> Q, dQ/dV, I, dI/dV all must be continuous. That's the theory
> anyway. I have not looked at the QUCS implementation. Who is lead
> on the HB simulator?
>
> Chuck
>
>
> On 1/24/2013 11:44 AM, mike brinson wrote:
>> Hello Chuck
>>
>> I agree for HB we need good continuity in a function , its 1st order
>> derivative and higher order
>> derivatives. The new BSIM 6 model has just these properties, see
>>
>> http://www.mos-ak.org/india/presentations/Chauhan_BSIM6_MOS-AK_India12.pdf .
>>
>> Mike.
>> Mike Brinson
>>
>> mbrin72043@... <mailto:mbrin72043@...>
>> ------------------------------------------------------------------------
>> *From:* Chuck <cmcook3rd@...> <mailto:cmcook3rd@...>
>> *To:* mbrin72043@... <mailto:mbrin72043@...>;
>> fransschreuder@... <mailto:fransschreuder@...>;
>> wackston@... <mailto:wackston@...>;
>> qucs-devel@...
>> <mailto:qucs-devel@...>
>> *Sent:* Thursday, 24 January 2013, 14:06
>> *Subject:* Re: [Qucs-devel] Bug/logic oddity in Transient solver loop
>>
>> HB simulation needs comtimuity in the function & 1st deritive. Models
>> need to be well behaved outside off expected ranges.
>>
>>
>> -------- Original message --------
>> Subject: Re: [Qucs-devel] Bug/logic oddity in Transient solver loop
>> From: mike brinson <mbrin72043@...>
>> <mailto:mbrin72043@...>
>> To: Frans Schreuder <fransschreuder@...>
>> <mailto:fransschreuder@... Stevens
>> <wackston@...>
>> <mailto:wackston@..."
>> <mailto:qucs-devel@...>
>> <qucs-devel@...>
>> <mailto:qucs-devel@...>
>> CC:
>>
>>
>> Hello Frans and Andrew
>>
>> The basic Qucs semiconductor device models are similar to those found
>> in SPICE. The EKV 2.6 MOS
>> model I added around two to three years ago. The same comment
>> applies to the MESFET models. All
>> the existing models appear to work for DC, AC and TRAN simulation.
>> Circuits with convergence problems
>> will, of course, fail sometimes. However, one area where there are
>> real problems is Harmonic Balance. If a
>> model has discontinuities in its I/V or Q characteristics second
>> order or higher derivatives HB simulation will
>> fail and not complete. My plans at the moment are to add a number of
>> new semiconductor device models, starting
>> with the latest upgrades to the HICUM models. I have both BSIM 3 and
>> BSIM 4 MOS models working with QucsStudio. These
>> should port to Qucs without too many problems. However, please note
>> the ADMS XML interface is different for
>> QucsStudio and Qucs so this could take some time. Once BSIM 6 is
>> released to the public domain I will also try and add this to Qucs.
>> I am also thinking of adding some basic BJT and MOS models which
>> work, hopefully without problems, with HB simulation.
>> These should allow those interested in RF and microwave design the
>> opportunity to apply HB simulation to a range of active circuits.
>> Again, please note HB simulation will also require work as its not
>> very stable at the moment.
>>
>> On the subject of models one area that needs work is the development
>> of the microwave models for transient simulation.
>> The existing microstrip models, for example, are AC linear models and
>> do not operate correctly in the transient time domain.
>>
>> Lots to keep the new team occupied !
>>
>> Best wishes.
>>
>> Mike
>>
>>
>> Mike Brinson
>>
>>
>> mbrin72043@... <mailto:mbrin72043@...>
>>
>>
>> ________________________________
>> From: Frans Schreuder <fransschreuder@...>
>> <mailto:fransschreuder@...>
>> To: Andrew Stevens <wackston@...>
>> <mailto:wackston@...>; "qucs-devel@..."
>> <mailto:qucs-devel@...>
>> <qucs-devel@...>
>> <mailto:qucs-devel@...>
>> Sent: Wednesday, 23 January 2013, 22:03
>> Subject: Re: [Qucs-devel] Bug/logic oddity in Transient solver loop
>>
>> Hi Andrew,
>>
>> That sounds busy! But in order to help you can also do small
>> portions, you
>> don't need to commit your whole life to it. Anyway if you say you don't
>> feel like joining that is ok.
>> I also just joined the team. All i know about the mosfet models is that
>> some of them work better than others. Especially in transient simulations
>> there are often problems.
>> The simulator was mainly developed by Stefan Jahn (ela) but he is not
>> active anymore. I know something about simulators but not enough to
>> create
>> completely new models.
>> I have joined qucs mainly to get it back to life and back into the major
>> linux repositories, for that to happen we needed a qt4 port which i
>> made. I
>> want to concentrate now on bug fixes and to get it at least as stable as
>> the qt3 version, after that i want to learn something about the simulator
>> as well. I have hardly touched the simulator (qucs-core) except for the
>> microstriplange coupler model which was derived from the coupled
>> microstrip lines.
>> For the future we are also looking for people who can help improving the
>> simulator backend.
>>
>> Frans
>> Op 23 jan. 2013 22:50 schreef "Andrew Stevens"
>> <wackston@...> <mailto:wackston@...>
>> hetvolgende:
>>
>> > On 20/01/13 17:10, qucs-devel-request@...
>> <mailto:qucs-devel-request@...> wrote:
>> > >
>> > > Hi Andrew,
>> > >
>> > > We are now building a new team to get Qucs back on the road, at least
>> > > the activity has been boosted.
>> > > A couple new people are working on it now (I am also new).
>> > > I have to get a little more familiar with the way the simulator
>> works, I
>> > > have now concentrated on the Qt4 part.
>> > > It seems like you know what you are talking about, why don't you
>> join us?
>> > >
>> > > Frans
>> > >
>> >
>> > Hi Frans,
>> >
>> > Sorry about the delay responding. I've been laid out with an evil cold
>> > for a few days.
>> > Joining the team sounds fun but with two young kids, a working partner
>> > and non-trivial day job I'm not sure I can make any commitments about
>> > any kind of sustained contributions.
>> >
>> > I'm afraid I know only enough about Newton-Raphson solver based circuit
>> > simulation and numerical analysis to know that I don't know much at all
>> > ;-). I'm a C++ EDA verification/synthesis tooling guy who ended up
>> > doing some time architectingSoC's more or less 'by accident'.
>> >
>> > However, I would like to get a cosimulation interface up and running.
>> >
>> > Regards,
>> >
>> > Andrew
>> >
>> >
>> > PS
>> > Whats the state of the MOSFET Models in Qucs? Some quick tests with a
>> > circuit using a MOSFET to switch a diode-clamped inductive load
>> produced
>> > some pretty weird behaviour. I can't yet rule out a oops in my
>> circuit
>> > - if I get a moment I'll try to cross-check both on SIMetrix - but it
>> > looked damn strange.
>> >
>> > Trying the same MOSFET in a flyback converter model that produce
>> > beautiful "classical" wave with ideal switches also behaved very
>> > strangely with a MOSFET replacing the switch.
>> >
>> >
>> >
>> >
>> ------------------------------------------------------------------------------
>> > Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
>> > MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
>> > with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
>> > MVPs and experts. ON SALE this month only -- learn more at:
>> > http://p.sf.net/sfu/learnnow-d2d
>> > _______________________________________________
>> > Qucs-devel mailing list
>> > Qucs-devel@...
>> <mailto:Qucs-devel@...>
>> > https://lists.sourceforge.net/lists/listinfo/qucs-devel
>> >
>> ------------------------------------------------------------------------------
>> Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
>> MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
>> with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
>> MVPs and experts. ON SALE this month only -- learn more at:
>> http://p.sf.net/sfu/learnnow-d2d
>> _______________________________________________
>> Qucs-devel mailing list
>> Qucs-devel@...
>> <mailto:Qucs-devel@...>
>> https://lists.sourceforge.net/lists/listinfo/qucs-devel
>> ------------------------------------------------------------------------------
>> Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
>> MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
>> with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
>> MVPs and experts. ON SALE this month only -- learn more at:
>> http://p.sf.net/sfu/learnnow-d2d
>> _______________________________________________
>> Qucs-devel mailing list
>> Qucs-devel@...
>> <mailto:Qucs-devel@...>
>> https://lists.sourceforge.net/lists/listinfo/qucs-devel
>>
>>
>
>
>
|