Improvements ... suggestions for newbies and circuit simulators

Thread Starter

JoeJester

Joined Apr 26, 2005
4,390
WBahn stated in another thread:
I really hate how most of these education-oriented simulators teach bad practices by ignoring real world reality. Particularly since student increasingly get little or no hands on experience building actual circuits.
This begs the question how can we, the community in the homework help area, help alleviate this problem identified by WBahn?

One thing I know we can do is enforce using standard values for resistors and capacitors from the EIA standard values. This can be accomplished two ways, have the TS find their own Exx charts or provide them one. I'd probably go with providing one to start them off on a good foot.

I can also see providing them application notes to reinforce any "real world" recommendations. I'd have not problem recommending something and then telling them to read the application note as to why I recommended it.

Get them to look at the specifications of the equipment they are using, for real world effects.

Granted even all the suggestions to the simulator still can not replace "building" the circuit, but hopefully it will translate to fewer errors when the person builds the circuit. Hopefully associative learning takes hold and the builder will recall why certain things were recommended.

Like all tools, the simulators have limitations.
 

LDC3

Joined Apr 27, 2013
924
Like all tools, the simulators have limitations.
One problem with simulator is that they always calculate the same values for a given situation.
How about including temperature dependence or components within the statistical spread?
 

Thread Starter

JoeJester

Joined Apr 26, 2005
4,390
One problem with simulator is that they always calculate the same values for a given situation.
How about including temperature dependence or components within the statistical spread?
What simulator doesn't have a worst case, monte carlo, temperature stepping, or parameter stepping functions these days?
 

WBahn

Joined Mar 31, 2012
29,976
Most analog simulators (I'm sure there are exceptions) CAN do a decent job of modeling these effects. The problem here is that they don't do this by default (nor would you really want them to, as nice as that might seem at first blush). So the student has to go out of their way to model effects that, in many cases, they aren't aware of or, at best, have only been given a few minutes worth of 'lecture awareness' in class. One that is seen repeatedly is falling prey to the notion that two discrete transistors are perfectly matched.

The worse problem are the digital simulators that have no problem with unconnected inputs and simply declaring that any input that is not actively driven HI is deemed to be LO. Many don't have any notion of an "Unknown" state and many that do don't realize that "Unknown" does not mean that you know nothing about that state in the future. For instance, some simulators will power up with Flip Flops in an "X" state, which is fine. But if I have a FF configured as a TFF then if I clock it I might still not know what state it is in, but I DO know that it is no longer in the state is WAS in. Simulators can't grasp this notion. So if I have a counter that cannot be reset and powers up in an unknown state, then it WILL count as a counter and the circuit will work fine as long as it can deal with that first partial cycle starting at an undefined point. But a simulator will show it forever being in an unknown state. What is needed (and perhaps there are simulators out there that can do this) is to be able to configure the simulator so that registers start up in a RANDOM state, just like they do in the real world.

As long as the simulators allow significant non-realworld situations to be accepted as the default, students will not learn about the importance of those issues except through real-world experience, usually involving smoking parts and burned fingers. This, sadly, is going the way of the dodo bird in engineering education.
 

#12

Joined Nov 30, 2010
18,224
I am seriously afraid you're flogging a dead horse. Having the students play video games with a simulator is cheaper than actually teaching them. Which do you want to fix? Designing a good simulator or making the schools actually teach?

We can illustrate the weak points for the students that ask, but a lot of them are going to walk out with a useless diploma and alternate between washing dishes and saying, "Would you like fries with that? (I need to pay a LOT of student loans.)

It hurts like letting go of your children to watch the noobies walk out into the real world when you know it's going to knock them down first and ask questions later. :(
 

WBahn

Joined Mar 31, 2012
29,976
I am seriously afraid you're flogging a dead horse. Having the students play video games with a simulator is cheaper than actually teaching them. Which do you want to fix? Designing a good simulator or making the schools actually teach?
We at least have a CHANCE of designing a good simulator....:(
 
Top