On Fri, 16 Feb 2018 20:20:54 -0000, "Kevin Aylward" <kevinRemovAT@kevinaylward.co.uk> wrote:>"Jim Thompson" wrote in message >news:bjbc8d5f855140u283t3hb464n6g31l8u1@4ax.com... > >On Thu, 15 Feb 2018 19:19:39 -0500, krw@notreal.com wrote: > >>On Thu, 15 Feb 2018 08:51:44 -0800, John Larkin >><jjlarkin@highlandtechnology.com> wrote: >> >>>On Thu, 15 Feb 2018 09:40:28 -0700, Jim Thompson >>><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>> >>>>On Thu, 15 Feb 2018 10:50:10 -0500, Phil Hobbs >>>><pcdhSpamMeSenseless@electrooptical.net> wrote: >>>> >>>>>On 02/14/2018 04:24 PM, Jim Thompson wrote: >>>>>> On Wed, 14 Feb 2018 15:07:42 -0600, John S <Sophi.2@invalid.org> >>>>>> wrote: >>>>>> >>>>>>> Here is what I have: >>>>>>> >>>>>>> .model LXHL-BW02 D(Is=4.5e-20 Rs=.85 N=2.6 Cjo=1.18n Xti=200 >>>>>>> Iave=400mA >>>>>>> Vpk=5 mfg=Lumileds type=LED) >>>>>>> >>>>>>> How do I change the Vf at a specified current? I suspect that it is >>>>>>> more >>>>>>> involved than just changing Rs. >>>>>>> >>>>>>> Thanks, Guys. >>>>>> >>>>>> Iave, Vpk and "type" are NOT true Spice parameters... they are >>>>>> toyLTspice fudges that use LTspice-specific models that are NOT usable >>>>>> anywhere else. >>>>> >>>>>Like pspice encryption? >>>> >>>>It's not the same thing. LTspice has encryption too, you twit. >>>> >>>>.MODEL D D(...) is a standard device model format that will "play" on >>>>any other Spice-compliant simulator. Mikey's version will not. >>>> >>>>The proper way to handle such a model is to define it as a subcircuit >>>>with parameter-passing. >>>> >>>>Most LTspice models of their standard products won't "play" on any >>>>other simulator. Is that good business? >>> >>>LT Spice is free! It doesn't need to be "good business" in the >>>paid-for simulator market. What it does is sell LTC chips. >>> >>> >>>> >>>>TI seems to have (somewhat) caught on to the fact that they need to >>>>offer across-platform/simulator models. >>>> >>>>Wonder when Analog Devices (nee Linear) will catch on, toss Mikey, and >>>>make available compliant models? >>> >>>I was told that ADI will be migrating to LT Spice. I think that LT >>>Spice is one reason they bought LTC. >> >>I understand that TI is migrating to PSpice. > >>Giving up TINA? I hadn't heard that. But, of what I've analyzed, >>TINA does not deviate enough from standard Spice to call home about. > >Despite SS using the basic Spice3/XSpice engine, I use it extensively, daily >in parallel with Cadence Virtuoso, because it has button press WC (Worse >Case). I can run 100s of process corners with ease. Without WC, you don't >have a design. Period. >Maybe for ICs. At board level, wc design will either run the cost way up, or result in un-competitive specs. We set the specs so that most units pass test, and fix or trash the few that don't. Real wc basically never happens. It's way too far out the probability curve. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
LTSpice LED parameters
Started by ●February 14, 2018
Reply by ●February 16, 20182018-02-16
Reply by ●February 16, 20182018-02-16
On Fri, 16 Feb 2018 20:20:54 -0000, "Kevin Aylward" <kevinRemovAT@kevinaylward.co.uk> wrote:>"Jim Thompson" wrote in message >news:bjbc8d5f855140u283t3hb464n6g31l8u1@4ax.com... > >On Thu, 15 Feb 2018 19:19:39 -0500, krw@notreal.com wrote: > >>On Thu, 15 Feb 2018 08:51:44 -0800, John Larkin >><jjlarkin@highlandtechnology.com> wrote: >> >>>On Thu, 15 Feb 2018 09:40:28 -0700, Jim Thompson >>><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>> >>>>On Thu, 15 Feb 2018 10:50:10 -0500, Phil Hobbs >>>><pcdhSpamMeSenseless@electrooptical.net> wrote: >>>> >>>>>On 02/14/2018 04:24 PM, Jim Thompson wrote: >>>>>> On Wed, 14 Feb 2018 15:07:42 -0600, John S <Sophi.2@invalid.org> >>>>>> wrote: >>>>>> >>>>>>> Here is what I have: >>>>>>> >>>>>>> .model LXHL-BW02 D(Is=4.5e-20 Rs=.85 N=2.6 Cjo=1.18n Xti=200 >>>>>>> Iave=400mA >>>>>>> Vpk=5 mfg=Lumileds type=LED) >>>>>>> >>>>>>> How do I change the Vf at a specified current? I suspect that it is >>>>>>> more >>>>>>> involved than just changing Rs. >>>>>>> >>>>>>> Thanks, Guys. >>>>>> >>>>>> Iave, Vpk and "type" are NOT true Spice parameters... they are >>>>>> toyLTspice fudges that use LTspice-specific models that are NOT usable >>>>>> anywhere else. >>>>> >>>>>Like pspice encryption? >>>> >>>>It's not the same thing. LTspice has encryption too, you twit. >>>> >>>>.MODEL D D(...) is a standard device model format that will "play" on >>>>any other Spice-compliant simulator. Mikey's version will not. >>>> >>>>The proper way to handle such a model is to define it as a subcircuit >>>>with parameter-passing. >>>> >>>>Most LTspice models of their standard products won't "play" on any >>>>other simulator. Is that good business? >>> >>>LT Spice is free! It doesn't need to be "good business" in the >>>paid-for simulator market. What it does is sell LTC chips. >>> >>> >>>> >>>>TI seems to have (somewhat) caught on to the fact that they need to >>>>offer across-platform/simulator models. >>>> >>>>Wonder when Analog Devices (nee Linear) will catch on, toss Mikey, and >>>>make available compliant models? >>> >>>I was told that ADI will be migrating to LT Spice. I think that LT >>>Spice is one reason they bought LTC. >> >>I understand that TI is migrating to PSpice. > >>Giving up TINA? I hadn't heard that. But, of what I've analyzed, >>TINA does not deviate enough from standard Spice to call home about. > >Despite SS using the basic Spice3/XSpice engine, I use it extensively, daily >in parallel with Cadence Virtuoso, because it has button press WC (Worse >Case). I can run 100s of process corners with ease. Without WC, you don't >have a design. Period. > > >-- Kevin Aylward >http://www.anasoft.co.uk - SuperSpice >http://www.kevinaylward.co.uk/ee/index.htmlYep. I do the same, but use macros... I suppose I could make it a "button-press" with MacroExpress... which I use extensively to ease my workload. ...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | STV, Queen Creek, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | It's what you learn, after you know it all, that counts.
Reply by ●February 16, 20182018-02-16
On 02/15/2018 08:45 PM, krw@notreal.com wrote:> On Thu, 15 Feb 2018 18:07:27 -0700, Jim Thompson > <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: > >> On Thu, 15 Feb 2018 19:19:39 -0500, krw@notreal.com wrote: >> >>> On Thu, 15 Feb 2018 08:51:44 -0800, John Larkin >>> <jjlarkin@highlandtechnology.com> wrote: >>> >>>> On Thu, 15 Feb 2018 09:40:28 -0700, Jim Thompson >>>> <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>>> >>>>> On Thu, 15 Feb 2018 10:50:10 -0500, Phil Hobbs >>>>> <pcdhSpamMeSenseless@electrooptical.net> wrote: >>>>> >>>>>> On 02/14/2018 04:24 PM, Jim Thompson wrote: >>>>>>> On Wed, 14 Feb 2018 15:07:42 -0600, John S <Sophi.2@invalid.org> >>>>>>> wrote: >>>>>>> >>>>>>>> Here is what I have: >>>>>>>> >>>>>>>> .model LXHL-BW02 D(Is=4.5e-20 Rs=.85 N=2.6 Cjo=1.18n Xti=200 Iave=400mA >>>>>>>> Vpk=5 mfg=Lumileds type=LED) >>>>>>>> >>>>>>>> How do I change the Vf at a specified current? I suspect that it is more >>>>>>>> involved than just changing Rs. >>>>>>>> >>>>>>>> Thanks, Guys. >>>>>>> >>>>>>> Iave, Vpk and "type" are NOT true Spice parameters... they are >>>>>>> toyLTspice fudges that use LTspice-specific models that are NOT usable >>>>>>> anywhere else. >>>>>> >>>>>> Like pspice encryption? >>>>> >>>>> It's not the same thing. LTspice has encryption too, you twit. >>>>> >>>>> .MODEL D D(...) is a standard device model format that will "play" on >>>>> any other Spice-compliant simulator. Mikey's version will not. >>>>> >>>>> The proper way to handle such a model is to define it as a subcircuit >>>>> with parameter-passing. >>>>> >>>>> Most LTspice models of their standard products won't "play" on any >>>>> other simulator. Is that good business? >>>> >>>> LT Spice is free! It doesn't need to be "good business" in the >>>> paid-for simulator market. What it does is sell LTC chips. >>>> >>>> >>>>> >>>>> TI seems to have (somewhat) caught on to the fact that they need to >>>>> offer across-platform/simulator models. >>>>> >>>>> Wonder when Analog Devices (nee Linear) will catch on, toss Mikey, and >>>>> make available compliant models? >>>> >>>> I was told that ADI will be migrating to LT Spice. I think that LT >>>> Spice is one reason they bought LTC. >>> >>> I understand that TI is migrating to PSpice. >> >> Giving up TINA? I hadn't heard that. But, of what I've analyzed, >> TINA does not deviate enough from standard Spice to call home about. > > That's my understanding (from a pretty good source). You can't import > models to the (free) TINA. If it's not in their library, forget it. I > have no idea if they're going to cripple it or not but I suspect it'll > only run on the web and without things like optimization and noise > analysis. They're certainly not going to give out free licenses to > the full PSpice. Cadence doesn't work that way. ;-) > >TI's TINA models are mostly horrible. The OPA140 model won't converge unless the supplies are *exactly* symmetrical, and not very often even if they are. Most of the board-level pspice models I've tried worked okay in LTspice. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net https://hobbs-eo.com
Reply by ●February 16, 20182018-02-16
On Fri, 16 Feb 2018 16:35:42 -0500, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote: [snip]> >TI's TINA models are mostly horrible. The OPA140 model won't converge >unless the supplies are *exactly* symmetrical, and not very often even >if they are.Didn't I write-up an OPA140 model for you that worked (~2015)?> >Most of the board-level pspice models I've tried worked okay in LTspice.Yes. Unfortunately a high percentage of LTspice models won't work with PSpice (or any other simulator) :-( That's a huge pain. I might design-in some LTspice switchers if I could simulate them along with non-LT components. So I just lead my clients to other vendors. The Spice models I create run on ANY Spice-compliant simulator. (I check my models on LTspice as well as a number of other cheapy simulators... due to customer requests :-)> >Cheers > >Phil Hobbs...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | STV, Queen Creek, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | It's what you learn, after you know it all, that counts.
Reply by ●February 16, 20182018-02-16
On 2018-02-16, whit3rd <whit3rd@gmail.com> wrote:> On Thursday, February 15, 2018 at 5:55:04 PM UTC-8, John Larkin wrote: > >> Face it: LT Spice does everything right. > > Not on a MacBook, it doesn't.Is there something badly broken in the OSX version of wine? -- This email has not been checked by half-arsed antivirus software
Reply by ●February 16, 20182018-02-16
On Fri, 16 Feb 2018 16:35:42 -0500, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>On 02/15/2018 08:45 PM, krw@notreal.com wrote: >> On Thu, 15 Feb 2018 18:07:27 -0700, Jim Thompson >> <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >> >>> On Thu, 15 Feb 2018 19:19:39 -0500, krw@notreal.com wrote: >>> >>>> On Thu, 15 Feb 2018 08:51:44 -0800, John Larkin >>>> <jjlarkin@highlandtechnology.com> wrote: >>>> >>>>> On Thu, 15 Feb 2018 09:40:28 -0700, Jim Thompson >>>>> <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>>>> >>>>>> On Thu, 15 Feb 2018 10:50:10 -0500, Phil Hobbs >>>>>> <pcdhSpamMeSenseless@electrooptical.net> wrote: >>>>>> >>>>>>> On 02/14/2018 04:24 PM, Jim Thompson wrote: >>>>>>>> On Wed, 14 Feb 2018 15:07:42 -0600, John S <Sophi.2@invalid.org> >>>>>>>> wrote: >>>>>>>> >>>>>>>>> Here is what I have: >>>>>>>>> >>>>>>>>> .model LXHL-BW02 D(Is=4.5e-20 Rs=.85 N=2.6 Cjo=1.18n Xti=200 Iave=400mA >>>>>>>>> Vpk=5 mfg=Lumileds type=LED) >>>>>>>>> >>>>>>>>> How do I change the Vf at a specified current? I suspect that it is more >>>>>>>>> involved than just changing Rs. >>>>>>>>> >>>>>>>>> Thanks, Guys. >>>>>>>> >>>>>>>> Iave, Vpk and "type" are NOT true Spice parameters... they are >>>>>>>> toyLTspice fudges that use LTspice-specific models that are NOT usable >>>>>>>> anywhere else. >>>>>>> >>>>>>> Like pspice encryption? >>>>>> >>>>>> It's not the same thing. LTspice has encryption too, you twit. >>>>>> >>>>>> .MODEL D D(...) is a standard device model format that will "play" on >>>>>> any other Spice-compliant simulator. Mikey's version will not. >>>>>> >>>>>> The proper way to handle such a model is to define it as a subcircuit >>>>>> with parameter-passing. >>>>>> >>>>>> Most LTspice models of their standard products won't "play" on any >>>>>> other simulator. Is that good business? >>>>> >>>>> LT Spice is free! It doesn't need to be "good business" in the >>>>> paid-for simulator market. What it does is sell LTC chips. >>>>> >>>>> >>>>>> >>>>>> TI seems to have (somewhat) caught on to the fact that they need to >>>>>> offer across-platform/simulator models. >>>>>> >>>>>> Wonder when Analog Devices (nee Linear) will catch on, toss Mikey, and >>>>>> make available compliant models? >>>>> >>>>> I was told that ADI will be migrating to LT Spice. I think that LT >>>>> Spice is one reason they bought LTC. >>>> >>>> I understand that TI is migrating to PSpice. >>> >>> Giving up TINA? I hadn't heard that. But, of what I've analyzed, >>> TINA does not deviate enough from standard Spice to call home about. >> >> That's my understanding (from a pretty good source). You can't import >> models to the (free) TINA. If it's not in their library, forget it. I >> have no idea if they're going to cripple it or not but I suspect it'll >> only run on the web and without things like optimization and noise >> analysis. They're certainly not going to give out free licenses to >> the full PSpice. Cadence doesn't work that way. ;-) >> >> > >TI's TINA models are mostly horrible. The OPA140 model won't converge >unless the supplies are *exactly* symmetrical, and not very often even >if they are. > >Most of the board-level pspice models I've tried worked okay in LTspice. > >Cheers > >Phil HobbsOpamp behavioral models seldom do power supplies right. I have one opamp sim where the amp generates 3KV into an unconnected power pin. Probably some pure current source inside. Wish the real part would do that. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
Reply by ●February 16, 20182018-02-16
On 02/16/2018 05:08 PM, Jim Thompson wrote:> On Fri, 16 Feb 2018 16:35:42 -0500, Phil Hobbs > <pcdhSpamMeSenseless@electrooptical.net> wrote: > > [snip] >> >> TI's TINA models are mostly horrible. The OPA140 model won't converge >> unless the supplies are *exactly* symmetrical, and not very often even >> if they are. > > Didn't I write-up an OPA140 model for you that worked (~2015)?Nope, as I recall you hacked with it a bit and then agreed that the TINA one was a mess.> >> >> Most of the board-level pspice models I've tried worked okay in LTspice. > > Yes. Unfortunately a high percentage of LTspice models won't work > with PSpice (or any other simulator) :-(Well, not the proprietary ones.> > That's a huge pain. I might design-in some LTspice switchers if I > could simulate them along with non-LT components. > > So I just lead my clients to other vendors.The only LTC part I've ever designed into a client's board is the LT1677, which had a combination of input CM range and output swing that I couldn't match with jellybeans without a lot of extra gingerbread. I'll probably design it out again next time round, because the next version is going to have an MCU in it that can do some of the fancy footwork.> > The Spice models I create run on ANY Spice-compliant simulator. (I > check my models on LTspice as well as a number of other cheapy > simulators... due to customer requests :-)I'm not that fond of circuit simulation, because the available models are so crappy. It's generally not a good enough use of my time to make better ones, so I've never bothered to get good at making models, just like I don't do my own PCB layouts. That's why I'm not that emotional about any of them. Some small discrete designs are difficult to get really right without modelling, because it's axiomatic that every breadboard contains at least one perfect component that you'll never see again. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics Briarcliff Manor NY 10510 http://electrooptical.net https://hobbs-eo.com
Reply by ●February 16, 20182018-02-16
On Fri, 16 Feb 2018 18:38:19 -0500, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>On 02/16/2018 05:08 PM, Jim Thompson wrote: >> On Fri, 16 Feb 2018 16:35:42 -0500, Phil Hobbs >> <pcdhSpamMeSenseless@electrooptical.net> wrote: >> >> [snip] >>> >>> TI's TINA models are mostly horrible. The OPA140 model won't converge >>> unless the supplies are *exactly* symmetrical, and not very often even >>> if they are. >> >> Didn't I write-up an OPA140 model for you that worked (~2015)? > >Nope, as I recall you hacked with it a bit and then agreed that the TINA >one was a mess.Sort of. A quick check of the E-mails back-and-forth between us shows the problem to have been "LTspice can't cope with a noiseless resistor as a DC path, PSpice can" (as well as any simulator Berkeley-compliant)... it's simply a G-source... why LTspice can't cope with that, who the hell knows ?:-) My complaints about LTspice are that so many are using it, yet it disobeys standard Spice rules right and left. My dad actually had his service trucks labeled "We DON'T service Muntz TV's" Maybe I should put up a banner "LTspice not supported" ... might cost me the business of the number of thumbs on one hand ?>:-}>> >>> >>> Most of the board-level pspice models I've tried worked okay in LTspice. >> >> Yes. Unfortunately a high percentage of LTspice models won't work >> with PSpice (or any other simulator) :-( > >Well, not the proprietary ones. > >> >> That's a huge pain. I might design-in some LTspice switchers if I >> could simulate them along with non-LT components. >> >> So I just lead my clients to other vendors. > >The only LTC part I've ever designed into a client's board is the >LT1677, which had a combination of input CM range and output swing that >I couldn't match with jellybeans without a lot of extra gingerbread. >I'll probably design it out again next time round, because the next >version is going to have an MCU in it that can do some of the fancy >footwork. > >> >> The Spice models I create run on ANY Spice-compliant simulator. (I >> check my models on LTspice as well as a number of other cheapy >> simulators... due to customer requests :-) > >I'm not that fond of circuit simulation, because the available models >are so crappy. It's generally not a good enough use of my time to make >better ones, so I've never bothered to get good at making models, just >like I don't do my own PCB layouts. That's why I'm not that emotional >about any of them. > >Some small discrete designs are difficult to get really right without >modelling, because it's axiomatic that every breadboard contains at >least one perfect component that you'll never see again. > >Cheers > >Phil Hobbs...Jim Thompson -- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | STV, Queen Creek, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at http://www.analog-innovations.com | 1962 | It's what you learn, after you know it all, that counts.
Reply by ●February 16, 20182018-02-16
On Fri, 16 Feb 2018 15:37:25 -0800, John Larkin <jjlarkin@highland_snip_technology.com> wrote:>On Fri, 16 Feb 2018 16:35:42 -0500, Phil Hobbs ><pcdhSpamMeSenseless@electrooptical.net> wrote: > >>On 02/15/2018 08:45 PM, krw@notreal.com wrote: >>> On Thu, 15 Feb 2018 18:07:27 -0700, Jim Thompson >>> <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>> >>>> On Thu, 15 Feb 2018 19:19:39 -0500, krw@notreal.com wrote: >>>> >>>>> On Thu, 15 Feb 2018 08:51:44 -0800, John Larkin >>>>> <jjlarkin@highlandtechnology.com> wrote: >>>>> >>>>>> On Thu, 15 Feb 2018 09:40:28 -0700, Jim Thompson >>>>>> <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>>>>> >>>>>>> On Thu, 15 Feb 2018 10:50:10 -0500, Phil Hobbs >>>>>>> <pcdhSpamMeSenseless@electrooptical.net> wrote: >>>>>>> >>>>>>>> On 02/14/2018 04:24 PM, Jim Thompson wrote: >>>>>>>>> On Wed, 14 Feb 2018 15:07:42 -0600, John S <Sophi.2@invalid.org> >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> Here is what I have: >>>>>>>>>> >>>>>>>>>> .model LXHL-BW02 D(Is=4.5e-20 Rs=.85 N=2.6 Cjo=1.18n Xti=200 Iave=400mA >>>>>>>>>> Vpk=5 mfg=Lumileds type=LED) >>>>>>>>>> >>>>>>>>>> How do I change the Vf at a specified current? I suspect that it is more >>>>>>>>>> involved than just changing Rs. >>>>>>>>>> >>>>>>>>>> Thanks, Guys. >>>>>>>>> >>>>>>>>> Iave, Vpk and "type" are NOT true Spice parameters... they are >>>>>>>>> toyLTspice fudges that use LTspice-specific models that are NOT usable >>>>>>>>> anywhere else. >>>>>>>> >>>>>>>> Like pspice encryption? >>>>>>> >>>>>>> It's not the same thing. LTspice has encryption too, you twit. >>>>>>> >>>>>>> .MODEL D D(...) is a standard device model format that will "play" on >>>>>>> any other Spice-compliant simulator. Mikey's version will not. >>>>>>> >>>>>>> The proper way to handle such a model is to define it as a subcircuit >>>>>>> with parameter-passing. >>>>>>> >>>>>>> Most LTspice models of their standard products won't "play" on any >>>>>>> other simulator. Is that good business? >>>>>> >>>>>> LT Spice is free! It doesn't need to be "good business" in the >>>>>> paid-for simulator market. What it does is sell LTC chips. >>>>>> >>>>>> >>>>>>> >>>>>>> TI seems to have (somewhat) caught on to the fact that they need to >>>>>>> offer across-platform/simulator models. >>>>>>> >>>>>>> Wonder when Analog Devices (nee Linear) will catch on, toss Mikey, and >>>>>>> make available compliant models? >>>>>> >>>>>> I was told that ADI will be migrating to LT Spice. I think that LT >>>>>> Spice is one reason they bought LTC. >>>>> >>>>> I understand that TI is migrating to PSpice. >>>> >>>> Giving up TINA? I hadn't heard that. But, of what I've analyzed, >>>> TINA does not deviate enough from standard Spice to call home about. >>> >>> That's my understanding (from a pretty good source). You can't import >>> models to the (free) TINA. If it's not in their library, forget it. I >>> have no idea if they're going to cripple it or not but I suspect it'll >>> only run on the web and without things like optimization and noise >>> analysis. They're certainly not going to give out free licenses to >>> the full PSpice. Cadence doesn't work that way. ;-) >>> >>> >> >>TI's TINA models are mostly horrible. The OPA140 model won't converge >>unless the supplies are *exactly* symmetrical, and not very often even >>if they are. >> >>Most of the board-level pspice models I've tried worked okay in LTspice. >> >>Cheers >> >>Phil Hobbs > >Opamp behavioral models seldom do power supplies right. I have one >opamp sim where the amp generates 3KV into an unconnected power pin. >Probably some pure current source inside.Sure. Most will supply more power than they consume.>Wish the real part would do that.#MeToo
Reply by ●February 16, 20182018-02-16
On Fri, 16 Feb 2018 18:38:19 -0500, Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> wrote:>On 02/16/2018 05:08 PM, Jim Thompson wrote: >> On Fri, 16 Feb 2018 16:35:42 -0500, Phil Hobbs >> <pcdhSpamMeSenseless@electrooptical.net> wrote: >> >> [snip] >>> >>> TI's TINA models are mostly horrible. The OPA140 model won't converge >>> unless the supplies are *exactly* symmetrical, and not very often even >>> if they are. >> >> Didn't I write-up an OPA140 model for you that worked (~2015)? > >Nope, as I recall you hacked with it a bit and then agreed that the TINA >one was a mess. >> >>> >>> Most of the board-level pspice models I've tried worked okay in LTspice. >> >> Yes. Unfortunately a high percentage of LTspice models won't work >> with PSpice (or any other simulator) :-( > >Well, not the proprietary ones. > >> >> That's a huge pain. I might design-in some LTspice switchers if I >> could simulate them along with non-LT components. >> >> So I just lead my clients to other vendors. > >The only LTC part I've ever designed into a client's board is the >LT1677, which had a combination of input CM range and output swing that >I couldn't match with jellybeans without a lot of extra gingerbread. >I'll probably design it out again next time round, because the next >version is going to have an MCU in it that can do some of the fancy >footwork. > >> >> The Spice models I create run on ANY Spice-compliant simulator. (I >> check my models on LTspice as well as a number of other cheapy >> simulators... due to customer requests :-) > >I'm not that fond of circuit simulation, because the available models >are so crappy. It's generally not a good enough use of my time to make >better ones, so I've never bothered to get good at making models, just >like I don't do my own PCB layouts. That's why I'm not that emotional >about any of them. > >Some small discrete designs are difficult to get really right without >modelling, because it's axiomatic that every breadboard contains at >least one perfect component that you'll never see again. > >Cheers > >Phil HobbsI like Spice because it trains my instincts. I am doing a really gnarly design right now that I really didn't understand until I Spiced it about 50 times and saw where the currents were actually going when. Then I could push things around, to simplify and optimize. The device power dissipation measurement thing is awesome, not something I'd want to do by hand... or on a breadboard. We also leave commented Spice sims in project folders to help document why we did things. That can be really handy to have around a few years later. -- John Larkin Highland Technology, Inc picosecond timing precision measurement jlarkin att highlandtechnology dott com http://www.highlandtechnology.com
