On Sat, 22 Jul 2017 12:11:44 -0700, Jim Thompson <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:>On Sat, 22 Jul 2017 12:05:37 -0700 (PDT), Klaus Kragelund ><klauskvik@hotmail.com> wrote: > >>On Saturday, July 22, 2017 at 8:40:24 PM UTC+2, Joerg wrote: >>> On 2017-07-22 07:43, John Larkin wrote: > >[snip] >>> > >>> > Colleges use LT Spice all over the world. It has its own culture. >>> > Culture matters. >>> > >>> >>> I'll second all that. 100%. >>> >>> I don't know how much sales I generated for LTC because of LTSpice but >>> it likely was 7-digit. And I am only one of many. >>> >>I don't believe that for a second >> >>I have worked at 6 different firms and I have never come across any colleagues using a single LT part. Only here at SED >> >>Cheers >> >>Klaus > > {>:-} > >Right down there with Microchip Analog parts ;-)Use them too. Really nice RRIO opamps, for cheap.
LT spice, leakage inductance
Started by ●July 20, 2017
Reply by ●July 22, 20172017-07-22
Reply by ●July 22, 20172017-07-22
On Sat, 22 Jul 2017 12:43:48 -0700, Joerg <news@analogconsultants.com> wrote:>On 2017-07-22 12:05, Klaus Kragelund wrote: >> On Saturday, July 22, 2017 at 8:40:24 PM UTC+2, Joerg wrote: >>> On 2017-07-22 07:43, John Larkin wrote: >>>> On Sat, 22 Jul 2017 08:42:45 +0100, "Kevin Aylward" >>>> <kevinRemovAT@kevinaylward.co.uk> wrote: >>>> >>>>> "John Larkin" wrote in message >>>>> news:2r95nc5ls1rrm7cmp5ptn26kmlrroi6jpp@4ax.com... >>>>> >>>>> On Fri, 21 Jul 2017 21:25:00 +0100, "Kevin Aylward" >>>>> <kevinRemovAT@kevinaylward.co.uk> wrote: >>>>> >>>>>> "John Larkin" wrote in message >>>>>> news:m794nclodso7tu7c9bihmic1cuqtuf4ufa@4ax.com... >>>>>> >>>>>> >>>>>> >>>>>>> Companies buy companies because the bought companies are >>>>>>> eating >>>>>>>> onto the buying companies revenues. The buying company >>>>>>>> wants the, now get this, the *Customers* of the bought >>>>>>>> company, and secondly, their *products* that customers >>>>>>>> *buy*. LT gets billions of revenue from paying >>>>>>>> customers, that's what ADI want and bought. Dah.... >>>>>>>> >>>>>>>> LTSpice is just a nice icing on the cake sort of thing. >>>>>>>> If it ceased to exist, it ut bwould make F'all >>>>>>>> difference. That's reality. >>>>>> >>>>>>> It's a major sales tool. >>>>>> >>>>>> No. Its a sales tool. >>>>>> >>>>>>> People pay for sales tools. >>>>>> >>>>>> Sure, but not $1B. >>>>>> >>>>>>> When I made the suggestion to the LT/ADI transition team, >>>>>>> nobody contradicted me. I'd guess that LT Spice added at >>>>>>> least a billion dollars to the deal. >>>>>> >>>>>>> Complete and utter nonsense. >>>>>> >>>>>>> Like, Cadence do about $2B revenue. Candace is THE >>>>>>> mainstream de-facto system in IC design. You know, the >>>>>>> things that the world runs on. >>>>>> >>>>>>> You're trying to claim that the LT Freebee is on a par in >>>>>>> value. >>>>>> >>>>>>> LT do about $2B in physical sales. You are effectively >>>>>>> trying to claim that if LTSpice didn't exist, LT would lose >>>>>>> like $100Ms of dollars. Bullshit. >>>>>> >>>>>>> You really need to stand back and think about the big >>>>>>> picture. >>>>>> >>>>>>> People buy LT parts because of the performance, not >>>>>>> because of a simulater. Period. >>>>>> >>>>>>> LTSpice only has to improve sales by say, a million or so >>>>>>> to justify its existence, i.e Mike's salary :-). I agree, >>>>>>> it probably does that, and such a salary pisses me off no >>>>>>> end.... >>>>>>> >>>>> >>>>>> In summary, you are jealous of LT Spice and its author. >>>>> >>>>> The summary, is that some here have absolutely no idea >>>>> whatsoever of the value of some marketing tools. >>>>> >>>>> AD do about $3B in product. >>>>> >>>>> How much would it cost to for any company to get the Spice3 >>>>> code, and hire someone to write a GUI for it? Like a few $100k >>>>> ? >>>>> >>>>> Hint: you can buy 3rd party GUI tools that have schematic >>>>> capture as a drop in component. >>>>> >>>>> Do your really think all the other companies are so totally >>>>> stupid that they wouldn't invest a few hundred $ to produce a >>>>> full featured Spice to give away for free if it would net them >>>>> a $1B more a year in revenue. Get real. >>>>> >>>>> >>>>> -- Kevin Aylward http://www.anasoft.co.uk - SuperSpice >>>>> http://www.kevinaylward.co.uk/ee/index.html >>>> >>>> ADI paid 14.8 big ones for LTC. LTC charges at least 2x what >>>> other people charge for similar functions. LT Spice helps them do >>>> that. >>>> >>>> TI's sim stuff won't even run on my PC for some reason. So I use >>>> LTC parts unless the cost impact is extreme. >>>> >>>> TPS54302 is a great little switcher, but I had to spend an >>>> afternoon breadboard fiddling to figure out all my needed >>>> compansations with various caps and voltages. I'll use a lot of >>>> them, so it was worth it. But neither TI's data sheet nor their >>>> software was much help. >>>> >>>> Colleges use LT Spice all over the world. It has its own >>>> culture. Culture matters. >>>> >>> >>> I'll second all that. 100%. >>> >>> I don't know how much sales I generated for LTC because of LTSpice >>> but it likely was 7-digit. And I am only one of many. >>> >> I don't believe that for a second >> > >It's true. I can't vouch for the exact amount, of course, but there were >designs that contain several LTC chips and are still in production. > > >> I have worked at 6 different firms and I have never come across any >> colleagues using a single LT part. Only here at SED >> > >In some markets a thorough simulation is very important. Most of my >designs work quietly in the background somewhere. You may have had them >very close to you, for example while flying in large aircraft. Others >are in hospitals but those machines you'd only encounter after a serious >cardiac event and hopefully that will never happen.I seem to have become a regular user of your sort of product. :-(
Reply by ●July 22, 20172017-07-22
On Sat, 22 Jul 2017 12:09:07 -0700, Jim Thompson <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:>On Sat, 22 Jul 2017 12:01:43 -0700, Jim Thompson ><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: > >>On Sat, 22 Jul 2017 11:40:34 -0700, Joerg <news@analogconsultants.com> >>wrote: >> >>>On 2017-07-22 07:43, John Larkin wrote: >> >>[snip] >>>> >>>> TPS54302 is a great little switcher, but I had to spend an afternoon >>>> breadboard fiddling to figure out all my needed compansations with >>>> various caps and voltages. I'll use a lot of them, so it was worth it. >>>> But neither TI's data sheet nor their software was much help. >>>> >>>> Colleges use LT Spice all over the world. It has its own culture. >>>> Culture matters. >>>> >>> >>>I'll second all that. 100%. >>> >>>I don't know how much sales I generated for LTC because of LTSpice but >>>it likely was 7-digit. And I am only one of many. >> >>Designed any chips with LTspice? >> >> ...Jim Thompson > >TPS54302 Spice model is encrypted... > ><http://www.ti.com/lit/zip/slvmbl9> > >just what I've been harping about. > >Models that won't run on ANY Berkeley-Spice-compliant simulator aren't >worth the powder to blow 'em to hell. > >Me, I can always resort to rolling my own, you peasants are SOL ;-) > >Right now I'm writing a model for a pressure-sensing MEMS... probably >take me a hour or two >:-} > > ...Jim ThompsonI ended up 'writing' a general purpose MEMS capacitor used as a sensor... <http://www.analog-innovations.com/DeviceModelsSubckts/General_Purpose_MEMS_Capacitor_Sensor_2017_07_22.png> You can input, via V(VC+,VC-), _any_ function you want (including its own terminal voltages should you so choose). (The 'dump' beyond a GHz is just my choice of bounding the d/dt. I'm going to go back and parameterize that to make it more general purpose ;-) (For those LTspice 'purists', it meets Mikey's subterfuge about 'charge-controlled' capacitors... BARF >:-} ...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 | I'm looking for work... see my website. Thinking outside the box...producing elegant & economic solutions.
Reply by ●July 23, 20172017-07-23
On Sat, 22 Jul 2017 12:57:59 -0700, Jim Thompson <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:>On Sat, 22 Jul 2017 12:39:45 -0700, John Larkin ><jjlarkin@highlandtechnology.com> wrote: > >>On Sat, 22 Jul 2017 12:11:44 -0700, Jim Thompson >><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >> >>>On Sat, 22 Jul 2017 12:05:37 -0700 (PDT), Klaus Kragelund >>><klauskvik@hotmail.com> wrote: >>> >>>>On Saturday, July 22, 2017 at 8:40:24 PM UTC+2, Joerg wrote: >>>>> On 2017-07-22 07:43, John Larkin wrote: >>> >>>[snip] >>>>> > >>>>> > Colleges use LT Spice all over the world. It has its own culture. >>>>> > Culture matters. >>>>> > >>>>> >>>>> I'll second all that. 100%. >>>>> >>>>> I don't know how much sales I generated for LTC because of LTSpice but >>>>> it likely was 7-digit. And I am only one of many. >>>>> >>>>I don't believe that for a second >>>> >>>>I have worked at 6 different firms and I have never come across any colleagues using a single LT part. Only here at SED >>>> >>>>Cheers >>>> >>>>Klaus >>> >>> {>:-} >>> >>>Right down there with Microchip Analog parts ;-) >>> >>> ...Jim Thompson >> >>Some of their parts are quite good, like their comparators. > >[Begin personal opinion (legal disclaimer) mode ;-] > >Their Spice models are the suckiest of the sucky. > > ...Jim ThompsonWho needs a Spice model for a comparator? I sure don't. -- John Larkin Highland Technology, Inc lunatic fringe electronics
Reply by ●July 23, 20172017-07-23
On Sat, 22 Jul 2017 12:50:57 -0700, Joerg <news@analogconsultants.com> wrote:>On 2017-07-22 12:16, John Larkin wrote: >> On Sat, 22 Jul 2017 11:40:34 -0700, Joerg <news@analogconsultants.com> >> wrote: >> >>> On 2017-07-22 07:43, John Larkin wrote: >>>> On Sat, 22 Jul 2017 08:42:45 +0100, "Kevin Aylward" >>>> <kevinRemovAT@kevinaylward.co.uk> wrote: >>>> >>>>> "John Larkin" wrote in message >>>>> news:2r95nc5ls1rrm7cmp5ptn26kmlrroi6jpp@4ax.com... >>>>> >>>>> On Fri, 21 Jul 2017 21:25:00 +0100, "Kevin Aylward" >>>>> <kevinRemovAT@kevinaylward.co.uk> wrote: >>>>> >>>>>> "John Larkin" wrote in message >>>>>> news:m794nclodso7tu7c9bihmic1cuqtuf4ufa@4ax.com... >>>>>> >>>>>> >>>>>> >>>>>>> Companies buy companies because the bought companies are eating >>>>>>>> onto the buying companies revenues. The buying company wants the, now get >>>>>>>> this, the *Customers* of the bought company, and secondly, their >>>>>>>> *products* >>>>>>>> that customers *buy*. LT gets billions of revenue from paying customers, >>>>>>>> that's what ADI want and bought. Dah.... >>>>>>>> >>>>>>>> LTSpice is just a nice icing on the cake sort of thing. If it ceased to >>>>>>>> exist, it ut bwould make F'all difference. That's reality. >>>>>> >>>>>>> It's a major sales tool. >>>>>> >>>>>> No. Its a sales tool. >>>>>> >>>>>>> People pay for sales tools. >>>>>> >>>>>> Sure, but not $1B. >>>>>> >>>>>>> When I made the suggestion to the LT/ADI transition team, nobody >>>>>>> contradicted me. I'd guess that LT Spice added at least a billion >>>>>>> dollars to the deal. >>>>>> >>>>>>> Complete and utter nonsense. >>>>>> >>>>>>> Like, Cadence do about $2B revenue. Candace is THE mainstream de-facto >>>>>>> system in IC design. You know, the things that the world runs on. >>>>>> >>>>>>> You're trying to claim that the LT Freebee is on a par in value. >>>>>> >>>>>>> LT do about $2B in physical sales. You are effectively trying to claim >>>>>>> that >>>>>>> if LTSpice didn't exist, LT would lose like $100Ms of dollars. Bullshit. >>>>>> >>>>>>> You really need to stand back and think about the big picture. >>>>>> >>>>>>> People buy LT parts because of the performance, not because of a >>>>>>> simulater. >>>>>>> Period. >>>>>> >>>>>>> LTSpice only has to improve sales by say, a million or so to justify its >>>>>>> existence, i.e Mike's salary :-). I agree, it probably does that, and such >>>>>>> a >>>>>>> salary pisses me off no end.... >>>>>>> >>>>> >>>>>> In summary, you are jealous of LT Spice and its author. >>>>> >>>>> The summary, is that some here have absolutely no idea whatsoever of the >>>>> value of some marketing tools. >>>>> >>>>> AD do about $3B in product. >>>>> >>>>> How much would it cost to for any company to get the Spice3 code, and hire >>>>> someone to write a GUI for it? Like a few $100k ? >>>>> >>>>> Hint: you can buy 3rd party GUI tools that have schematic capture as a drop >>>>> in component. >>>>> >>>>> Do your really think all the other companies are so totally stupid that they >>>>> wouldn't invest a few hundred $ to produce a full featured Spice to give >>>>> away for free if it would net them a $1B more a year in revenue. Get real. >>>>> >>>>> >>>>> -- Kevin Aylward >>>>> http://www.anasoft.co.uk - SuperSpice >>>>> http://www.kevinaylward.co.uk/ee/index.html >>>> >>>> ADI paid 14.8 big ones for LTC. LTC charges at least 2x what other >>>> people charge for similar functions. LT Spice helps them do that. >>>> >>>> TI's sim stuff won't even run on my PC for some reason. So I use LTC >>>> parts unless the cost impact is extreme. >>>> >>>> TPS54302 is a great little switcher, but I had to spend an afternoon >>>> breadboard fiddling to figure out all my needed compansations with >>>> various caps and voltages. I'll use a lot of them, so it was worth it. >>>> But neither TI's data sheet nor their software was much help. >>>> >>>> Colleges use LT Spice all over the world. It has its own culture. >>>> Culture matters. >>>> >>> >>> I'll second all that. 100%. >>> >>> I don't know how much sales I generated for LTC because of LTSpice but >>> it likely was 7-digit. And I am only one of many. >> >> Literally millions of kids will get EE degrees where they learned to >> use LT Spice. LTS has mediocre documentation and HELP, and lots of >> hidden features, and the best tutorials are written by college profs, >> for their students. All those kids will go out and continue to use LT >> Spice, and spec LTC parts. [1] [2] >> > >I am afraid that the vast majority of them will have no clue about >anything analog, anything at the transistor level. Yeah, they can press >the run button on LTSpice but when they change something and the thing >goes berserk they don't know why. > > >> They won't be using SuperSpice much, because the student version is >> badly crippled and not free. >> >> [1] not everyone battles to find the lowest-price parts. I've known >> engineers that preferred expensive parts, because they thought they >> might be better, or just because it made them feel more important. >> > >Sometimes I have to design mission-critical stuff. There cost is >secondary but you have to simulate the dickens out of it. That is not >possible when using an IC for which no SPICE model is published. > > >> [2] NI understands this dynamic too. When I was a kid, I worked >> summers in a college electronics lab, for 85 cents an hour. Fairchild >> and TI would send us boxes of their newfangled silicon transistors; >> same idea. >> > >Oh yeah. I have yet to encounter a client where their test lab is not >running on NI modules and LabView. At one we needed to switch some >higher voltage and a simple relay board cost hundreds of Dollars.Yesterday one of my customers was complaining about critical NI modules, used in their test stands, going EOL. They are scouring ebay for more. -- John Larkin Highland Technology, Inc lunatic fringe electronics
Reply by ●July 23, 20172017-07-23
On Sat, 22 Jul 2017 20:35:43 -0700, John Larkin <jjlarkin@highlandtechnology.com> wrote:>On Sat, 22 Jul 2017 12:57:59 -0700, Jim Thompson ><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: > >>On Sat, 22 Jul 2017 12:39:45 -0700, John Larkin >><jjlarkin@highlandtechnology.com> wrote: >> >>>On Sat, 22 Jul 2017 12:11:44 -0700, Jim Thompson >>><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>> >>>>On Sat, 22 Jul 2017 12:05:37 -0700 (PDT), Klaus Kragelund >>>><klauskvik@hotmail.com> wrote: >>>> >>>>>On Saturday, July 22, 2017 at 8:40:24 PM UTC+2, Joerg wrote: >>>>>> On 2017-07-22 07:43, John Larkin wrote: >>>> >>>>[snip] >>>>>> > >>>>>> > Colleges use LT Spice all over the world. It has its own culture. >>>>>> > Culture matters. >>>>>> > >>>>>> >>>>>> I'll second all that. 100%. >>>>>> >>>>>> I don't know how much sales I generated for LTC because of LTSpice but >>>>>> it likely was 7-digit. And I am only one of many. >>>>>> >>>>>I don't believe that for a second >>>>> >>>>>I have worked at 6 different firms and I have never come across any colleagues using a single LT part. Only here at SED >>>>> >>>>>Cheers >>>>> >>>>>Klaus >>>> >>>> {>:-} >>>> >>>>Right down there with Microchip Analog parts ;-) >>>> >>>> ...Jim Thompson >>> >>>Some of their parts are quite good, like their comparators. >> >>[Begin personal opinion (legal disclaimer) mode ;-] >> >>Their Spice models are the suckiest of the sucky. >> >> ...Jim Thompson > >Who needs a Spice model for a comparator? I sure don't.You apparently only do _one_device_at_a_time_ simulations? No whole-system simulations? ...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 | I'm looking for work... see my website. Thinking outside the box...producing elegant & economic solutions.
Reply by ●July 23, 20172017-07-23
On Sat, 22 Jul 2017 21:00:58 -0700, Jim Thompson <To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote:>On Sat, 22 Jul 2017 20:35:43 -0700, John Larkin ><jjlarkin@highlandtechnology.com> wrote: > >>On Sat, 22 Jul 2017 12:57:59 -0700, Jim Thompson >><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >> >>>On Sat, 22 Jul 2017 12:39:45 -0700, John Larkin >>><jjlarkin@highlandtechnology.com> wrote: >>> >>>>On Sat, 22 Jul 2017 12:11:44 -0700, Jim Thompson >>>><To-Email-Use-The-Envelope-Icon@On-My-Web-Site.com> wrote: >>>> >>>>>On Sat, 22 Jul 2017 12:05:37 -0700 (PDT), Klaus Kragelund >>>>><klauskvik@hotmail.com> wrote: >>>>> >>>>>>On Saturday, July 22, 2017 at 8:40:24 PM UTC+2, Joerg wrote: >>>>>>> On 2017-07-22 07:43, John Larkin wrote: >>>>> >>>>>[snip] >>>>>>> > >>>>>>> > Colleges use LT Spice all over the world. It has its own culture. >>>>>>> > Culture matters. >>>>>>> > >>>>>>> >>>>>>> I'll second all that. 100%. >>>>>>> >>>>>>> I don't know how much sales I generated for LTC because of LTSpice but >>>>>>> it likely was 7-digit. And I am only one of many. >>>>>>> >>>>>>I don't believe that for a second >>>>>> >>>>>>I have worked at 6 different firms and I have never come across any colleagues using a single LT part. Only here at SED >>>>>> >>>>>>Cheers >>>>>> >>>>>>Klaus >>>>> >>>>> {>:-} >>>>> >>>>>Right down there with Microchip Analog parts ;-) >>>>> >>>>> ...Jim Thompson >>>> >>>>Some of their parts are quite good, like their comparators. >>> >>>[Begin personal opinion (legal disclaimer) mode ;-] >>> >>>Their Spice models are the suckiest of the sucky. >>> >>> ...Jim Thompson >> >>Who needs a Spice model for a comparator? I sure don't. > >You apparently only do _one_device_at_a_time_ simulations? > >No whole-system simulations? > > ...Jim ThompsonMy systems are way too complex to simulate. I only sim one little section if I think there's any risk. Lots of products don't need simulation at all; design it, lay out the board, and sell it. For a complex product, I can do that sooner than someone could sim it. Comparators seldom need to be simulated; it would just slow things down. Some simpler element will work to do the comparator function and run a lot faster. I use ideal opamps a lot too, in a higher-level simulation. That's fast too. -- John Larkin Highland Technology, Inc lunatic fringe electronics
Reply by ●July 23, 20172017-07-23
On Thursday, July 20, 2017 at 8:39:37 PM UTC+2, John Larkin wrote:> I'm simulating a flyback switcher and noticed that any K < 1 in the > transformer radically slows down the sim, which is annoyingly slow > already... 90 seconds to sim 60 ms of startup and a little pulsed load > blip. Maybe 6:1 slower with a little leakage inductance.With K < 1 the leakage inductor will resonate with the switch node capacitance. When there is no snubber, this will be a very high-frequency oscillation, driving down the time step and thus increasing simulation time. LTspice is very actively looking for such behavior -- some other simulators can be fooled into not seeing it. -marcel
Reply by ●July 23, 20172017-07-23
On Saturday, July 22, 2017 at 9:42:58 AM UTC+2, Kevin Aylward wrote: [..]> How much would it cost to for any company to get the Spice3 code, > and hire someone to write a GUI for it? Like a few $100k ?Looking e.g. at the ngspice open source project, to get from the SPICE3 code to something comparable to LTspice takes at least 50 man-years, and then we are not talking code monkeys. It is hard for me to believe that a good GUI for spice3 will cost only 300K$ if it is written from scratch by a hired gun. I think I'd prefer to find somebody to sell one to me for that price that it is already working :-) Even then, you'd need a few fte's to get all the bugs out. -marcel
Reply by ●July 23, 20172017-07-23
John Larkin <jjlarkin@highland_snip_technology.com> wrote:> I'm simulating a flyback switcher and noticed that any K < 1 in the > transformer radically slows down the sim, which is annoyingly slow > already... 90 seconds to sim 60 ms of startup and a little pulsed load > blip. Maybe 6:1 slower with a little leakage inductance. So I only > include leakage inductance to tweak the snubber.You should always include realistic stray capacitances in the inductor model. Also, the default ESR is 1 milliohm when you do not specify it yourself. This default value does not occur for capacitors. You can end up with a high-Q resonance at very high frequencies that slow down the sim. Always check to see the Spice -> Netlist Options -> Default Integration Method is set to Modified Trap. This is Mike Englehardt's creation and is described in "SPICE Differentiation" in http://cds.linear.com/docs/en/lt-journal/LTJournal-V24N4-01-df- SPICEDifferentiation-MikeEngelhardt.pdf Another factor is Schottky diodes used as rectifiers. Some models have components that greatly slow the simulation. Try another diode or just use the default D.
