Are unused detached components and instances laying around on the schematic affecting the simulation? I got inconsistent result.

Status
Not open for further replies.
M

melkord

Full Member level 3
Joined
May 18, 2018
Messages
151
Helped
0
Reputation
0
Reaction score
1
Trophy points
18
Activity points
1,772
I am having inconsistent ac and pz simulation result.

From the pole frequency value that should not be there, it makes me think that it happens before but at the end, I got my expected result. So, I moved on.
Today, it happens again and it keeps showing wrong value. It makes the bandwidth change as well.

I suspect it is caused by some instances and component laying around that I use for troubleshooting purpose.
These components are connected with noConn from basic library.

Has it happened to anyone ?


 

Components like this do affect he circuit matrix and it has happened that they influence things like convergence. It should not affect, though the ac and pz results, if the components are completely isolated from everything else.
 

Run a test case to see if you can get the sim to break.


Regards, Dana.
 

Every element affects the DC solution because
there are tolerances, within which anything goes.
The solution is only as good as you say it has to
be. Change the network and where the error is
largest, can change too.

Changing that "residue" can change operating
point if the network is sufficiently sensitive (like
throw a femtoamp against a gigohm, and your
op amp solves with 1mV input offset instead of
zero and small signal gain drops a few dB, etc.).

Now you'd like a circuit design and representation
that -isn't- so fussy. But figuring out whether or
why, is another chore.
 

The simulator may regard detached components as circuits which you know are not complete but which the simulator acts as though they have a power supply and ground and current flow and normal parameters. These incomplete circuits may refuse to 'make sense' to the simulator (regardless of what time-step you choose or other settings). So the simulation may go through a certain number of iterations, then say 'that's enough' and start calculating the next frame but with incorrect values for your main circuit.

It may help if you connect the spare components to your main circuit through high ohm resistors. Change them to low-ohm when you wish to reconnect those components.
 

Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…