Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Are PSpice & LTSpice worthwhile or just trouble?

Status
Not open for further replies.

ClydeCrashKop

Well-Known Member
Most Helpful Member
This is from Programmer's Corner in another forum. This seems to be typical.
I get the impression that PSpice & LTSpice just add a whole new layer of problems and misdirection to designing circuits. Are they worthwhile or just trouble?


Spice problems.jpg
 
I use SPICE!
LTSpice is free. People complain greatly about free software.
People will not take the time to learn SPICE so they complain.
If you don't know electronics SPICE is hard.
This is a complicated program. SPICE is much much better than 30 years ago. And still people complain.
People also complain about CAD. Mostly about free CAD.
If you pay $4000.00 for a program it comes with a book. And people actually take the time to learn. But a free program should just run with out any learning.
"How come it is hard?" Well if you won't get the book, and you won't read the .PDF and you have never used it before......
Most hobbyist just clime on the internet and ask for help. Then complain because there is not a free help line.
Some of us use SPICE every week/day. That helps. Some people only use SPICE once a year, that makes SPICE hard.
Life is hard.
SPICE is a tool. Like a hammer. If you don't know how to use it, it is hard. When building a house a hammer will add a level of complexity to your work. Once you get good with SPICE you will find it as valuable as a hammer.
Yes you can get fooled by SPICE. And a hammer can "fix" your thumb real good.
 
There are numerous free Spice packages. Multisim, LTSpice, TINA, pSpice, are but a few of the free programs.

TINA-TI is at the Texas Instrument's Website.

I've used multisim back when it was called Electronics Workbench. I shifted to TINA about the time of the change.

Each program has it's quirks.

A 2012 article from Electronic Design can be found at https://electronicdesign.com/analog...ce-tools-capture-and-simulate-analog-circuits
 
Last edited:
Spice is a somewhat complex but very powerful simulation tool.
If you don't really understand circuits and try to use it to simulate crappy or nonsensical designs then it may give erroneous or weird results (the old garbage in -- garbage out problem).
It also doesn't show all problems that may occur in a real circuit (such as the effects of component tolerances, unless you deliberately include them in the circuit simulation, or that running 10A through a 100mA diode will either cause it to short or turn into a fuse:eek:).
But it's well worth the effort to learn (even though it does have a somewhat steep learning curve) if you do any circuit design and learn the limitations of the simulation.

I never built any circuits, even simple, ones without first simulating them (I now use LTspice).
If the simulation works, then I have a reasonable assurance that the actual circuit will work (even though normal circuit tolerances may cause it to work slightly differently).
If the simulation doesn't work, then it's almost a certainty that the actual circuit won't either.
 
This is from Programmer's Corner in another forum. This seems to be typical.

Programmers are generally not renowned for their superb circuit design skills, so it's not surprising that some of them have difficulty using Spice.

I get the impression that PSpice & LTSpice just add a whole new layer of problems and misdirection to designing circuits.
They might, I suppose, to someone with near-zero electronics knowledge or experience; but to anyone who knows what he's doing, Spice is a powerful, effective design tool that does its job very, very well. I've been using Spice for over 25 years and I would not want to be without it.

Are they worthwhile or just trouble?
They're more than just worthwhile: they're indispensable.

Remember: the posts you cited in your illustration are from people who are having trouble understanding how Spice works and how to use it; don't assume that because a few people are struggling, EVERYBODY is struggling. I assure you, they're not.
 
To which I ask: What the hell do programmers know about circuit design?

Okay, that explains it. That reminds me of something I overheard one time.
"Put down that wheelbarrow. You don't know nothin' about machinery!"
 
Last edited:
Are they worthwhile or just trouble?
It's all relative. First issue is the "Free" aspect. You get what you pay for. Literally. Extremely limited versatility.

For strictly analog circuit design, I'm a big fan of TINA Basic 10.0 (Spice based. Designsoft - $129USD).

For the user that actually intends to build a real circuit with real parts, TINA Basic (Version 9.3.80.273 BV-DS - not free) gives me a tool that mimics a real breadboard effort at designing, at times, remarkably complex (to me) circuits before I even heat up the solder pencil, i.e. (an MP3 player control circuit):
upload_2016-3-13_9-22-58.png

For me (a bread boarder at heart), I like the animated parts (switches can be "switched", relay terminals change positions, motors rotate, lights "light" etc.) that mimic the real thing. And I can translate the TINA circuit to a real circuit with 99% confidence that it will work as expected (note C3 1u cap - without it, that 555 Latch circuit would not "hold" in real circuit, even though it did hold in TINA sim). I will say its PCB app does require a ton of trace re-alignment(s).

I've used Free LTSpice (as have many ETO members). It does give to you very precise tools for investigating circuit performance, but the learning curve is steep. And (best I can tell) it lacks a "real time" mode where the circuit performs it's function while you watch it and make "on the fly" changes. It seems more attuned to the professional EE type than the practical circuit maker.

Anyway, my take on the issue.
 
Last edited:
BTW, this video is a good intro to LTSPICE:

 
Okay, that explains it. That reminds me of something I overheard one time.
"Put down that wheelbarrow. You don't know nothin' about machinery!"

Just to show how I routinely use LTSpice, look at this recent forum thread.

I used my experience ( I've been around the block a few times) to come up with a tenative circuit design that solves the TS's problem. LTSpice didn't do that; it came out of my head. I have done similar things with the TL431, so had that in my bag of tricks...

First thing I used LTSpice for was to draw the beginnings of the schematic. I'll be the first to admit that LTS is not a great schematic drafting tool, but it is very simple and straightforward (Programmers hate it, but this Engineer likes it).

I had the idea of putting the LEDs into what looks like a bridge circuit, where the current through the inverse-parallel connected LEDs reverses when the TL431 switches on/off. LTS didn't suggest that... I had done it before.

I fleshed out the design as I was drafting the schematic, adding needed components to the schematic like current limiting resistors for the LEDs. Since the TS wanted two LEDs, based on previous experience, I extended the design to use the bi-color LED or two different color LEDs. LTS didn't do that; I did...

I knew that TL431 is essentially a voltage comparator (opamp) with a band-gap voltage reference connected to one of its inputs, and could be used to switch a LED as the input voltage changes by a few mV. This application is not what the TL431 "Shunt Regulator" is usually used for.

I knew it could be used this other way because I have done it before, many times. LTS doesn't know that so that it can't suggest that type of usage. It can only show how a circuit works after you have allread created a schematic...

So, as I got closer to having a complete schematic, I ran a couple of quick LTS simulations to see how my design was going to work. The first step was modeling the TS's existing circuit, and then to see what happens if his 28V supply stays fixed, but the load voltage varies.

LTS just does only what it is told, so my experience said that there will be stuff in the schematic that models his existing stuff, and my Voltage Detector circuit would be added on. Note how I used the LTS schematic to show which parts are existing; and which parts need to be added. LTS didn't do that, I did...

Since the purpose of the design is to demonstrate that the LEDs flip just as the load voltage V(load) crosses 10V, I set up a sweep of V(load) from 11V to 9V to see if the trip point is where it is supposed to be. LTS knows how to do lots of different types of simulations; I knew that the .DC simulation was the appropriate one for this test...

So I didn't guess the initial R2/R3 voltage divider ratio quite right, and the trip point wasn't right at 10V, so I did several simulations changing R2 and R3 by selecting standard 1% values that can be purchased and ended up pretty close to a 10V trip point. I did this much quicker than I could have solved two equations in two unknowns. I know that the LM431 switches when its Ref pin has 2.495V on it; LTS can model that, but it can't tell you that directly... You still have to read and understand the TL431's data sheet.

Next, I plotted the current through D1 and D2, and noticed that based on my initial guesses for R1, R5 and R6, the currents were not well matched. A few more trial simulations with changing the values of R1, R5 and R6 brought the currents into balance... LTS made the simulations instantaneous, but my experience told me which values to change, and by how much.

Next, I asked LTS to tell me what power is being dissipated in the TL431 and the various resistors. I didn't post the plots of power, but I checked to make sure all the resistor powers where substantially less than 250mW, so 1/4W resistors can be used. The power dissipated in the TL431 is a tiny fraction of it maximum allowed values. Again, LTS didn't tell me that this should be done, but it makes asking the question and getting the answer almost effortless...

So, hopefully, from this posting, you have figured out that LTS doesn't turn you into a circuit designer if you aren't one already, but it makes a lot of the tasks that circuit designers do practically painless...

I find the "animated" components that cowboybob talks about are worthless to me. In this project, for example, I wanted to see when the Red LED lights up vs when the Green LED lights up as a function of V(load) [not as a function of time], so I simply asked LTS to plot the current through the two LEDs as V(load) sweeps from 11V to 9V. It showed me exactly what I can expect from the circuit, including the little dead spot. Not likely that having animated LED's would have exposed that little detail...
 
Last edited:
One thing that SPICE can do is try all combinations of component error. (resistors are +/-5%) So what happens when some resistors are 5% high and some 5% low. (all combinations) What if capacitors are +/-10%.

You can make tests like, what percent of error does it take to cause a problem.

In production it might take a while to find a unite where R52 is 5% high and R7 is 5% low and C9 is 10% high. But in SPICE it might take seconds to try all combinations.

It also helps me guess what value is best. Lets say I chose 10k ohms for no good reason. I can ask SPICE to try all values form 4.7k to 50k and see what happens. If 9.1k through 47k all work good then my 10k is probably not the best choice. I should choose a value in the middle of the good values.
____________________________________________
Different subject.
On another thread some one connected a 15 volt battery to a LED and got 100 amps of current.
1) A spice "battery" probably can make millions of amps with out smoke. (not a real world battery) You can make a spice battery that is more real world if you know how.
2) A real LED would explode at 100 amps. A spice LED will not. I think this is a problem.
3) Most spice files will not tell you when the voltage is too high. (100 volts on a 5V part) Parts don't brake! This is like when my accounting program says there is $1,000,000.00 in the bank. (not true) (garbage in garbage out = I sold 1000 parts for $1.00 each but I entered 1000 parts for 1000.00 each)
 
Q:Would Multi-sim count as a PSpice and LTspice type tool?
Yes, except that it attempts to add digital simulation and u-processor emulation. It is a very expensive tool if bought legitimately; a lot of people on these forums steal it.

OTOH, LTSpice is totally free.

AFAICR, PSpice has student and hobbyist discounts; but is not free...

All three have roots that go back to Berkley SPICE, running on a VAX computer, using printouts on an ASR33 teletype (the first time I ran it ;))
 
Last edited:
If I remember right (Berkley Spice): I wrote the schematic on paper. Then labeled the nodes. 1,2,3,4.........
To enter the schematic:
R1, 5,8 10k (R1 from node-5 to node-8) Do this 100 time for all 100 parts and see if there are errors. Yes there are typing errors.
Then you start sim and go for lunch.

A graph looks like this:
100......................x.................

150..........x..................x.............

50.....x............................x.........

0..x......................................x...

It will take several minutes for the TTY to print out a page.
 
It looks like the Spice programs are a very good tool for you guys that know what you are doing.

"Put down that wheelbarrow. You don't know nothin' about machinery!"
Was aimed at those that don't know what they are doing. (Me included)

We appreciate you guys sharing from your bags of tricks.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top