Spectre : Invalid component name was given as value of parameter

Rather than just pointing us at a resource with a generic capability, please provide an actual example that shows the problem or at least some clear guidelines how to take the link you gave and get the result you're seeing.

The simplest would be to post the input.scs, custombjt.sp and GNTAna.scs files referenced in the spectre log file that you shared a screenshot of.

Andrew

Please see attached.   To use GNT capability, you need to follow the guideline inside this README

gnt_scs_files.zip

Well, the errors are very clear. The alter statements in GNTAna.scs are referring to devices that don't exist. For some it's clear - the instance of gntprobe is called igft not gft and so lines like:

agnt2 alter dev=gft.viy param=mag value=0

should be

agnt2 alter dev=igft.viy param=mag value=0

However, I don't know that the lines with dev=I are supposed to be referring to.

I did briefly download the GNT package, and try out the gnt/noninverting_amplifier_1 example using the dc_gft state in ADE L, and that appeared to work fine - from this I can see that the instance names on the form are correctly igft and VI (so it should have been VI on the form above). 

Rather than me debugging somebody else's tool, I suggest you contact the authors of this utility.

Regards,

Andrew

I just found out what went wrong.  I forgot to put a forward slash symbol before "igft"  and "VI"  and "vo"  during the ADE_L spectre analysis setup

Now the utility tool does not throw any more errors.

However, I am getting entirely different and strange (the opamp seems not working at all) simulation waveform as compared to the waveform on page 12 of getting_started.pdf

Now the utility tool does not throw any more errors.

However, I am getting entirely different and strange (the opamp seems not working at all) simulation waveform as compared to the waveform on page 12 of getting_started.pdf

As I said, wouldn't this be better to raise with the authors of the tool rather than asking here? I'm not sure it makes sense for people on this forum to debug somebody else's work... (I certainly don't have the bandwidth to do that, however interesting it might be!)

Andrew