LTSpice v26.0.1 convergence / stability issues after upgrading from v26.0.0

Dear 2Torr ,

Every LTspice release is extensively tested. 26.0.1 is by far the most robust version ever.

It is a common misconception that something is "wrong" with a new release simply because some random circuit "worked" in previous releases. Many LTspice models use non-physical idealized components to speed up the simulation, but unfortunately it is very easy to construct ill-formed circuits this way. Improvements in LTspice often expose these flaws, which is why models suddenly fail in a new release. We've seen this a hundret times. This is a very complex matter.

I'm afraid we can't be of much help without a test case that demonstrates the problem. However, I strongly advice against using 26.0.0 for reasons that are beyond the scope of this post.

Best Regards,
Mathias

Hi 2Torr ,

2Torr said:

It's definitely related to an update of the program, not the included SPICE models, as I'm able to simulate the circuits fine in v26.0.0 and earlier with the latest models.

As mborn states, It would be helpful if could upload the specific schematics (and any required files) that show this behavior, so we can confirm the problem and possibly find solutions. Zip up the files and drag to this window.

thanks,

mike

Hi Mike and Mathias,

Thanks for your replies. Would it be possible to share the sample circuits privately, and that they be kept in strict confidence as they contain proprietary information?

If not, I might be able to cobble together some sample circuits that illustrate the recent issues.

Thanks!

Hi 2Torr ,

Yes, I'll set it up.

mike