Menu
▾
▴
Re: [Ngspice-users] How to declare a capacitor value as a function of time and use it in the simulation (Nevo Magnezi)
|
From: Simon H. <sim...@ie...> - 2019-07-12 00:56:22
|
Hello Cap Jockeys, I work with dielectric elastomer energy generators - at the physical level charge is always conserved. If you change the voltage across one of these suddenly they "twitch" like a muscle, so the electrical energy goes into mechanical movement, heat and some remains in the altered capacitance and some is lost to the elastomer material's leakage. Interestingly the change in capacitance effect becomes non-linear at high voltages (in part to it being easier to push the material one way or the other due to elastic forces and compressive forces not quite matching each other - this effect allows you to put one of these into the sole of a boot, put a kV or so bias on it then extract energy from that change in capacitance. Modelling this in SPICE is challenging to say the least, but from my experience I can tell you charge is always conserved. If your model tells you different then you have it wrong and it will come back and bite you when you build one of these. Modelling any real-world capacitor such as those on silicon or in ceramic capacitors correctly will always dictate the use of a non-linear dielectric model - there is an interesting paper from Teverovsky <https://nepp.nasa.gov/files/25847/2013-Taverovsky-paper-NEPPweb-MLCCsVabs-n264.pdf> that gives you a linear model for dielectric absorption (see figure 8.1, page 33) which is "time based" that should get you to start thinking about how you might model the step capacitance effect you are discussing. Just as an aside: - I see this as a as an external application modelling problem and not one the internal primitive behaviour of the simulator needs to replicate as the effects are too complex to solve for all applications. I think in this case the device models are the responsibility of the end users of the Ngspice tool to solve, not the tool writers as there are all the primitives already available to do so. That knowledge is perhaps what differentiates a junior and a senior engineer I suspect? Keep up the good work guys - I have read this forum for 14 years and have always enjoyed the exchanges you have. Best regards, SimonH (= a spicer since '86) Signature ============================================ _Simon Harpham_ <https://www.linkedin.com/pub/simon-harpham/0/b45/a12> _CEng_ <http://www.engc.org.uk/ceng.aspx> __MIET 29958838_ <http://www.theiet.org/membership/profreg/downloads/ceng-comp-synopsis.cfm>_ ============================================ On 12/07/2019 06:11, ngs...@li... wrote: > Send Ngspice-users mailing list submissions to > ngs...@li... > > To subscribe or unsubscribe via the World Wide Web, visit > https://lists.sourceforge.net/lists/listinfo/ngspice-users > or, via email, send a message with subject or body 'help' to > ngs...@li... > > You can reach the person managing the list at > ngs...@li... > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Ngspice-users digest..." > > > Today's Topics: > > 1. Re: How to declare a capacitor value as a function of time > and use it in the simulation (Nevo Magnezi) > 2. Re: How to declare a capacitor value as a function of time > and use it in the simulation (Nevo Magnezi) > 3. Re: How to declare a capacitor value as a function of time > and use it in the simulation (Saurabh Bansode) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Wed, 10 Jul 2019 23:29:54 +0000 > From: Nevo Magnezi <nm...@pr...> > To: pip...@gm..., ngs...@li... > Subject: Re: [Ngspice-users] How to declare a capacitor value as a > function of time and use it in the simulation > Message-ID: > <zud...@pr...> > > Content-Type: text/plain; charset="utf-8" > > I hypothesize that as you physically move the plates of a capacitor apart, you must expend energy as the charges on each plate are attracted to one another. So as C goes down, so does energy and voltage. Amount of charge Q I would imagine stays the same. Though perhaps someone with greater experience in electrodynamics would want to chime in. > > - Nevo > > -------- Original Message -------- > On Jul 11, 2019, 1:56 AM, Anders Andersson wrote: > >> On Wed, Jul 10, 2019 at 8:33 AM <mh...@ia...> wrote: >>> A capacitance that depends on *time* will have "interesting" properties. >>> Let's say at t=0 the capacitor is 1uF and charged to 100V. At time=t, >>> its >>> capacitance changes to 0.1uF. This means its voltage instantaneously >>> changes 10-fold (charge balance), while its stored energy changes >>> 100-fold. Where did that energy come from? And how/why did the >>> capacitance decide to change? Does it have a build-in clock? >> https://en.wikipedia.org/wiki/Variable_capacitor >> >> Common component in every radio during the whole 20th century. There's >> nothing magic about this. >> >> _______________________________________________ >> Ngspice-users mailing list >> Ngs...@li... >> https://lists.sourceforge.net/lists/listinfo/ngspice-users > -------------- next part -------------- > An HTML attachment was scrubbed... > > ------------------------------ > > Message: 2 > Date: Wed, 10 Jul 2019 23:41:11 +0000 > From: Nevo Magnezi <nm...@pr...> > To: pip...@gm..., ngs...@li... > Subject: Re: [Ngspice-users] How to declare a capacitor value as a > function of time and use it in the simulation > Message-ID: > <5O1...@pr...> > > Content-Type: text/plain; charset="utf-8" > > Correction: My last hypothesis is false because C = Q/V and if V goes down C goes up... > > Fundamentally, I think it is likely best to simply perceive a capacitor as a ratio of charge and voltage. We are used to these quantities always moving linearly together, but they don't have to. > > In Marcel's example, reducing capacitance is the same as increasing voltage. One might think this increases energy of a closed system at initial glance, but obviously either charge will necessarily change to balance things out, or energy might be converted from another type. What happens likely depends on the device. > > - Nevo > > -------- Original Message -------- > On Jul 11, 2019, 2:29 AM, Nevo Magnezi wrote: > >> I hypothesize that as you physically move the plates of a capacitor apart, you must expend energy as the charges on each plate are attracted to one another. So as C goes down, so does energy and voltage. Amount of charge Q I would imagine stays the same. Though perhaps someone with greater experience in electrodynamics would want to chime in. >> >> - Nevo >> >> -------- Original Message -------- >> On Jul 11, 2019, 1:56 AM, Anders Andersson wrote: >> >>> On Wed, Jul 10, 2019 at 8:33 AM <mh...@ia...> wrote: >>>> A capacitance that depends on *time* will have "interesting" properties. >>>> Let's say at t=0 the capacitor is 1uF and charged to 100V. At time=t, >>>> its >>>> capacitance changes to 0.1uF. This means its voltage instantaneously >>>> changes 10-fold (charge balance), while its stored energy changes >>>> 100-fold. Where did that energy come from? And how/why did the >>>> capacitance decide to change? Does it have a build-in clock? >>> https://en.wikipedia.org/wiki/Variable_capacitor >>> >>> Common component in every radio during the whole 20th century. There's >>> nothing magic about this. >>> >>> _______________________________________________ >>> Ngspice-users mailing list >>> Ngs...@li... >>> https://lists.sourceforge.net/lists/listinfo/ngspice-users > -------------- next part -------------- > An HTML attachment was scrubbed... > > ------------------------------ > > Message: 3 > Date: Thu, 11 Jul 2019 20:40:55 +0530 > From: Saurabh Bansode <sau...@gm...> > To: Holger Vogt <hol...@un...> > Cc: ngs...@li..., MANISH MEHTA > <man...@gm...> > Subject: Re: [Ngspice-users] How to declare a capacitor value as a > function of time and use it in the simulation > Message-ID: > <CAC...@ma...> > Content-Type: text/plain; charset="utf-8" > > Hi, > This is the function I want to use for capacitor value as a function of > time. > Attaching an image for reference. > > C(t) = -155.68t + 53.57 for 0<=t<0.32 > 276.77t - 84.81 for 0.32<=t<0.5 > 53.57 for 0.5<=t<1 > > Attaching an image for reference. > Thank you in advance. > > > On Wed, Jul 10, 2019, 11:14 AM Holger Vogt <hol...@un...> wrote: > >> What is the equation C=f(t) you want to achieve? >> >> >> >> _______________________________________________ >> Ngspice-users mailing list >> Ngs...@li... >> https://lists.sourceforge.net/lists/listinfo/ngspice-users >> > -------------- next part -------------- > An HTML attachment was scrubbed... > -------------- next part -------------- > A non-text attachment was scrubbed... > Name: IMG_20190711_203901.jpg > Type: image/jpeg > Size: 1387705 bytes > Desc: not available > > ------------------------------ > > > > ------------------------------ > > Subject: Digest Footer > > _______________________________________________ > Ngspice-users mailing list > Ngs...@li... > https://lists.sourceforge.net/lists/listinfo/ngspice-users > > > ------------------------------ > > End of Ngspice-users Digest, Vol 108, Issue 4 > ********************************************* > |
