This seems to me to be such a sensible requirement but the only way I can see to do it is a fudge and is very time consuming. I have had to set up two full versions of the same circuit and run them both at the same time in the same simulation to be able to plot both of their outputs on the same graph. I just can't see a way of doing it other than that despite being a user of LTSpice for some years now.

Is there a way to add plots together in the same plot pane for a number of simulation runs? That implies holding onto the plot of the first run while the second one is being plotted on the same pane.

Let's say I run a sim and plot the response at a particular node. I then go into the circuit and make a fundamental change to one part, say I select a totally different passive filter configuration within the schematic, not just change values of existing components. I then want to add the plot of this new circuit to the existing plot from the previous run.

Can this be done?

 

The nearest thing I can think of to what you're asking is the use of the STEP directive to re-run a sim with a parameter changed.

 

The only way I know of is to add both filters to the circuit and then using SW voltage-controlled switches and the .STEP directive to switch between filters.
LTspice will then make one run with one filter, and make a second run with the other filter, superimposing the two results on the same plot.

You still have to put both filters in the circuit, but the rest of the circuit doesn't have to be duplicated.

If you need an example on how that works, I can post one.

 

No I don't think that is what I need Alec, I'm very familiar with the STEP directive and unless there is something about its use that I am missing that will only make changes to a parameter within the current circuit configuration. I need a process to modify the actual configuration itself within a single simulation run.

I think Crutschow is onto at least a simplification of what I am currently having to do. My needs would be simple enough for that to not add too much complexity. I'll give that a whirl.

Thanks to both of you for responding guys.

 

A simple example:

 

Hahaha! Pretty much how I set things up in my own test Crutschow. Thanks for the heads up. It works fine (as we knew it should) and has solved my problem. It does make things a bit ponderous setting up changeover switches from pairs of the basic SW model but I guess you have to accept what it can do and what it can't. Does anyone know if there is a SPDT model already available as a single entity?

 

Hahaha! Pretty much how I set things up in my own test Crutschow. Thanks for the heads up. It works fine (as we knew it should) and has solved my problem. It does make things a bit ponderous setting up changeover switches from pairs of the basic SW model but I guess you have to accept what it can do and what it can't. Does anyone know if there is a SPDT model already available as a single entity?

Hi

Also, consider that the switches in post #5 are not really ever an open circuit. Each switch has two resistances, a value when open and a value when closed. So...for the circuit in post #5, the filters are always in parallel. How that affects filter response depends on the switch behavior and the ron/roff values.

You can also accomplish the same thing by using one circuit and stepping C1's and R1's value on each pass.

For example:




eT

 

See

 

You can also accomplish the same thing by using one circuit and stepping C1's and R1's value on each pass.

That works for my simple example where the circuit elements are the same, but not where the two filters would have a completely different configuration.

 

That works for my simple example where the circuit elements are the same, but not where the two filters would have a completely different configuration.

Why not?

 

Why not?

How could it if you have two completely different filter circuits?

 

How could it if you have two completely different filter circuits?

You just add the appropriate tables for the component values you want to change.

 

You just add the appropriate tables for the component values you want to change.

You are not getting the point.
He wants to test say, two completely different circuits, not one circuit with different component values.

 

hi bordonbert,
This zip has some LTS switch models.
E

 

You are not getting the point.
He wants to test say, two completely different circuits, not one circuit with different component values.

OK...understand...post #5 is not a good example of what the TS wants.

So how about posting the circuit(s) that need to be tested?

 

May be this utility will help?
Edit: From the author of utility:
...
The features are:
1.
Merge as many raw files you want into one raw file.
The advantage is that you can have different simulations in one graph.
...
https://groups.google.com/forum/#!topic/sci.electronics.cad/y_f1dBVTBPc
Edit: Utility recommended in LTspice XVII / Help / Help Topics / F.A.Q. / Exporting Waveform Data / Exporting/Merging Waveform Data.

 

So how about posting the circuit(s) that need to be tested?

The TS will have to do that, if he's so inclined.

But this thread is about how to do such a simulation in general, not just for a specific circuit.