When is the industry going to give us models whereby you
choose worst case (data sheet worst case values), or typical ,
a check box when you do sim.

Of course we have to fight the fact many are incomplete. Good
example is which CMOS models are we aware of that include
the input crossover distortion elements in RRI input diff pair to do a
proper sim for a 20+ bit A/D signal path? I would bet NONE.

Regards, take your pill Dana, Dana.

 

The problem is, what's a worst-case value for one circuit may be a best-case value for another.
But it would be nice to have some models that at least give you some sort of option for at least some of the parameter limits.

 

When is the industry going to give us models whereby you
choose worst case (data sheet worst case values), or typical ,
a check box when you do sim.

Of course we have to fight the fact many are incomplete. Good
example is which CMOS models are we aware of that include
the input crossover distortion elements in RRI input diff pair to do a
proper sim for a 20+ bit A/D signal path? I would bet NONE.

Regards, take your pill Dana, Dana.

That should be fairly simple to do. Set a param that sets the appropriate variables to change the behavior.

There’s already a device in LTSpice that does this. The universal opamp device has a selectable behavior setting via a pulldown list box attribute.

eT

 

When is the industry going to give us models whereby you
choose worst case (data sheet worst case values), or typical ,
a check box when you do sim.

Of course we have to fight the fact many are incomplete. Good
example is which CMOS models are we aware of that include
the input crossover distortion elements in RRI input diff pair to do a
proper sim for a 20+ bit A/D signal path? I would bet NONE.

Regards, take your pill Dana, Dana.

They will do it when they perceive that it is in their economic interest to do so. Explain what you think the economic interest would be in having a model for every part. What would you pay for the model? What would you pay for a part with a model? Free has both benefits and limitations.

I'm surprised that LTC continues to support LTSPICE after having been acquired by their corporate overlords.

 

I'm surprised that LTC continues to support LTSPICE

.... and I for one am grateful that they do. It's an invaluable tool.

 

.... and I for one am grateful that they do. It's an invaluable tool.

A special shout-out to Mike Engelhardt, of LTC who has and continues to make this project his life's work. Just think how many active deices he has saved by allowing simulation to "smoke" the part instead of allowing it to be done for real. THANKS MIKE!

 

Seconded!

 

A special shout-out to Mike Engelhardt, of LTC who has and continues to make this project his life's work. Just think how many active deices he has saved by allowing simulation to "smoke" the part instead of allowing it to be done for real. THANKS MIKE!

I just came across this note from Mike in the LTspice forums. People there were wondering where he has gone and what he might be doing since leaving ADI. Great news!

I've been writing an all new simulator with dramatically improved performance and capability.

I wrote LTspice 20 years ago. I know now how it should have been done. The only way for me to move forward was to delete the entire LTspice source code tree and start over from scratch.

I'll announce here when there's something to try out.

Stay healthy and the best to you all,

--Mike

 

it would be nice to have some models that at least give you some sort of option for at least some of the parameter limits.

i guess the lack of the Vceo and the lack of the presence of the zener effect of the EB CB junctions are all for allowing the simulator engine do do something in real time rather than halt at first over-current/-voltage event - which may be and not present in real world circuitry due the parasitic L , C , R of the wires , ESR , etc.

↑ it would be cool to simulate all in atomic-/quantum-level assuming everione has a supercomputer with unlimited hw resources ... ??? ... ↑

 

Is this maybe him?

 

Is this maybe him?

Maybe a colleague if not him. I love that they’re soliciting user input. That shows Mike’s long experience with helping users. After a while you learn that heading off user problems is the goal.

 

Is this maybe him?

I don't think so. He uses his given name on the LTSpice group on Yahoo/groups.io Don't know why he would need to use an alias, but I could be wrong.

 

I know you like LTSpice ... however, TINA has these options ...



but I am not 100% certain the free version TINA-TI has them.

On edit ... in a training series, it shows TINA-TI with that same options.