Shop OBEX P1 Docs P2 Docs Learn Events
What analog Electronics Simulation Software does everyone use? and why? — Parallax Forums

What analog Electronics Simulation Software does everyone use? and why?

Bob Lawrence (VE1RLL)Bob Lawrence (VE1RLL) Posts: 1,720
edited 2015-03-03 07:34 in General Discussion
What Spice based analogue electronic simulation software does everyone use? and why?

New Wave Live Wire ( inexpensive ) I started with this one and it is easy to use but limited .
http://www.new-wave-concepts.com/pr/livewire.html


Linear - (FREE) LTSpiceIV (then I started to use this one and still use it)
http://www.linear.com/designtools/software/#LTspice


TI.com - Tina (FREE) ( I've had a look at this one but have not tried much)
http://www.tina.com/


Digikey - Parts Sim (FREE) ( I tried one design with this software - I's easy to use but too limited( under development)
http://www.digikey.ca/en/resources/design-tools/partsim


B2.Spice A/D Windows Version 2014 ( I tried the trial a few times and it works good but is too expensive for the few times a year I use software like this)
http://www.beigebag.com/?

Comments

  • Heater.Heater. Posts: 21,230
    edited 2015-02-28 17:48
    I love LTSpice. It's easy is fast it's free. It's a Windows program but it has always worked perfectly under Wine on Linux.

    Not being a professional circuit designer I'm not inclined to pay for such things.

    Anything that won't run on Linux is of no use.
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2015-02-28 19:09
    I use LTSpice, too, for those very rare times when I need to simulate a circuit.

    -Phil
  • LeonLeon Posts: 7,620
    edited 2015-02-28 20:24
    I use SIMetrix Spice:

    http://www.simetrix.co.uk/site/index.html

    It's very good, with good support

    It's integrated with the Pulsonix PCB software that I use.
  • Bob Lawrence (VE1RLL)Bob Lawrence (VE1RLL) Posts: 1,720
    edited 2015-02-28 21:57
    @ Heater

    I like it as well however, I've never tried it on Linux. I'll put that on my list of to do list. :)

    @phil

    re: for those very rare times when I need to simulate a circuit.

    I don't simulate that often either. Most of the time I find it easier and faster to build the circuit on a breadboard and test it in the real world rather than simulate.

    @Leon
    re:I use SIMetrix Spice

    Thanks! I'm downloading it now to try it .
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2015-02-28 23:15
    Bob Lawrence (VE1RLL) wrote:
    I don't simulate that often either. Most of the time I find it easier and faster to build the circuit on a breadboard and test it in the real world rather than simulate.
    I think it was Steve Ciarcia who said that his favorite software was solder -- or something to that effect.

    -Phil
  • Bob Lawrence (VE1RLL)Bob Lawrence (VE1RLL) Posts: 1,720
    edited 2015-02-28 23:26
    @phil

    I think it was Steve Ciarcia who said that his favorite software was solder -- or something to that effect.

    Yes he did. Quote attached :)

    Ciarcia's Circuit Cellar, Volume 7

    From " A Supercomputer" 12

    Steve_Quote.png
    Steve_Quote.png
    388 x 198 - 63K
  • Jeff HaasJeff Haas Posts: 421
    edited 2015-03-01 00:56
    I heard that quote attrributed to Bob Pease (well-known National Semiconductor engineer) - see the tribute to him at TI.com, and scroll down to see the quote on the right side of the page:
    http://www.ti.com/ww/en/bobpease/

    It's possible Ciarcia was quoting Bob.

    Plus there's this clever way of remembering Bob:
    http://blog.atmel.com/2013/09/19/bob-pease-says-my-favorite-programming-language-is-solder/
  • Heater.Heater. Posts: 21,230
    edited 2015-03-01 03:14
    Programming with a soldering iron used to be common. Just use a diode array ROM. Make the connections at the right places to set the bit pattern, code, that you want.

    This is a ROM from an HP9100A calculator.

    attachment.php?attachmentid=113331&d=1425208127

    In fact, such diode array ROMS, although smaller, are used to program the conflict matrix of many old traffic light controllers that are still in use around the world.

    Mind you, I don't think that is the kind of programming Bob had in mind.
    9100diodenest.jpg
    800 x 531 - 298K
  • Heater.Heater. Posts: 21,230
    edited 2015-03-01 03:30
    I have used Spice occasionally to play with a circuit a bit before I get to try building it. It's nice to see that whatever calculations were done to arrive at component values are at least in the right ball park.

    I'm always weary, of the accuracy of the results. Spice is a lumped element simulation, in reality your circuit is a big wobbly eletro-magnetic jelly, a continuum of capacitance, inductance and resistance where all parts can interact with all other parts in unexpected ways. More so as frequencies go up. A simple case is the decoupling capacitor, Spice does not care where you place your capacitor or how long and wide the traces are to it. Reality does!

    More often I just find myself just playing with Spice. When I find some new circuit configuration I have never heard of before it's nice to be able to build one in simulation and play around with it, just for fun and learning.
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    edited 2015-03-01 08:56
    Weary of the accuracy of the results?

    Or wary of the....? Reader beware. Either way, electronics simulation software can be very demanding to use. The actual program, and the quality of the componet database are big issues.

    While everyone provides the right stuff for resistors, coils, and capacitors; the non-linear devices such as op-amps, transistors, and MOSfets get tricky to acquire good data.

    Electro-magnetic jelly? Mmmm.... delicious. The world is full of estimates, guess-timates, and trash. The simulations are only a good as the user and his sophistication.

    ++++++++++
    I have tried several versions of Spice offered in Linux. They mostly seemed incomplete and terrible. I don't think I tried LTSpice.

    I suspect that getting a version from TI or Linear Technologies or such means much better data and better long-term support as this is a sales tool for them.

    My overall impression was that one needed Windows or OSX to acquire the good ones. Linux users have to resort to installing WINE. TINA may be independent of the OS and cloud based, but wants money.
  • Bob Lawrence (VE1RLL)Bob Lawrence (VE1RLL) Posts: 1,720
    edited 2015-03-01 09:36
    re: my favorite programming language is solder

    Ok then my favorite debugger is Desoldering Braid :)

    @ Loopy Byteloose

    re:TINA may be independent of the OS and cloud based, but wants money.
    There is a cloud version that's a paid ver but they have a desktop version that's free

    re:The simulations are only a good as the user and his sophistication.


    or her :)
  • PublisonPublison Posts: 12,366
    edited 2015-03-01 09:55
    Bob Lawrence (VE1RLL) wrote: »
    re: my favorite programming language is solder

    Ok then my favorite debugger is Desoldering Braid :)

    LOL Good one Bob.!
  • Beau SchwabeBeau Schwabe Posts: 6,566
    edited 2015-03-01 09:59
    lol - Heater .... I have used "diode ROM" before in a "Decade to 7-Segment LED decoder" :-)
  • Heater.Heater. Posts: 21,230
    edited 2015-03-01 10:04
    Loopy well spotted, perhaps I'm just too weary to be wary just now. Of course, yes, I may also be weary of the results. :)

    As you say the simulator is just fine in terms of accuracy, it's just that what it is modelling is perhaps not what you are building. A capacitor, famously has equivalent series resistance and inductance. I'm not sure how many capacitor models take this into account. Same goes for inductors and resistors, every component has second order effects that may not be in the models you are using.

    Then how many people build models of their PCB traces and other interconnects into their simulations? Those connections are also components. They also have R and C and L. They are also coupled to each other via electric and magnetic fields. How many people bother to take all that interaction into account?

    So yes, to make a sensible and accurate simulation one has to understand the limitations of the models in use. Especially as frequencies or impedances go out of limits where the lumped circuit element model works. Second order effect and all that.

    LTSpice is a bit special because many years ago Mike Engelhardt came up with a way to speed up the maths of the simulation by a great deal and then went looking for someone to market it. Ending up at LT. It's not just the progiginal Spice code from UCB wrapped up in a GUI like many others have been.

    The Windows limitation is a pain. As it is is so many other application areas.
  • Heater.Heater. Posts: 21,230
    edited 2015-03-01 10:26
    Beau,

    Coincidentally Fran Blanche just released a video describing "diode steering" and demoing the idea with a decade to 7-segment decoder built from diodes: https://www.youtube.com/watch?v=JNbDdnEpAOQ

    Now, what I want to do is build the boot ROM for CP/M on the Propeller using a matrix of LEDs. Then we can watch it boot up!
  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2015-03-01 16:01
    Jeff Haas wrote:
    Plus there's this clever way of remembering Bob:
    http://blog.atmel.com/2013/09/19/bob...age-is-solder/
    Great post! This may change forever the way I do silkscreens!

    bob-pease_my-favorite-programming-language-is-solder_back.jpg?w=1220&h=432

    -Phil
  • Beau SchwabeBeau Schwabe Posts: 6,566
    edited 2015-03-01 19:24
    Found It !! My Diode ROM application ( 1991 ) ... Yes, I never throw away anything :-)

    Back then we simulated our logic on paper !! ... and we liked it !!


    For an assignment in an early electronics class we had to demonstrate our knowledge of designing with logic gates. Here is a design that I was able to fit into something the size of a pack of cigarettes including the battery. The design had a 60Hz time base that I divided down to 10ths of a second or 1 second clocks, or you could manually clock the counter. A 7447 BCD to 7-segment was used, but to get the LED's on the seven segment to display HEX, it was necessary to decode the remaining BCD with discrete logic gates since the 7447 would only go from 0 to 9 .... In order to do this DIODE ROM was used to get "A" "b" "C" "d" "E" and "F". I got an "A" on the assignment. :-)

    http://bscircuitdesigns.com/Electronics.JPG
  • Mark_TMark_T Posts: 1,981
    edited 2015-03-02 07:09
    One issue I see with simulating is that you are tempted to use the "typical" response for
    a component, rather than the worst-case behaviour (which is of course context-dependent),
    giving an unrealistically optimistic result.

    I tend to think designs through worrying about the worst case for each stage, so that there
    are no unpleasant surprises (ie you assume the transistor has a gain of only 100 when the model
    assumes 300, you assume the opamp offsets are the worst case in the worst directions...)

    So my software of choice is brain, pen, calculator, back of envelope...
  • kwinnkwinn Posts: 8,697
    edited 2015-03-02 07:22
    Always a good idea to to take a look at the worst case scenarios.

    BTW, you left napkins off the list ;-)
  • Bob Lawrence (VE1RLL)Bob Lawrence (VE1RLL) Posts: 1,720
    edited 2015-03-03 07:34
    @Publison

    re: Quote Originally Posted by Bob Lawrence (VE1RLL) View Post
    re: my favorite programming language is solder

    Ok then my favorite debugger is Desoldering Braid
    ======================================================
    re: LOL Good one Bob.!


    In 50 years people are going to ask I wonder who came up with that line and they will get a reply like " oh that was a Prop head guy named Publison" LOL

    @Beau

    re:Found It !! My Diode ROM application ( 1991 ) ...

    It sure took allot of chips for that design. How things change :)
Sign In or Register to comment.