Forum Replies Created
20. August 2021 at 17:41 in reply to: Opening older QucsStudio schematics in version 4.2.2 #1022
The only way is to load old schematics with version 3.3.2 and save them. After this, you can also load the files with version 4.2.2.
The new version of “CompiledLibrary.qucs” is now available! Please download and try again.
Yes, you’re right! Thanks for the bug report. I will fix it.
Do you want to use the rectangular waveguide in a circuit simulation?
I don’t know how to model this. But if you can help here, it will be quite easy to implement it as an external component.3. May 2021 at 17:03 in reply to: Transient simulation of a bipolar transistor fails #918
There’s indeed a bug in the substrate capacitance that causes this problem. If you set Cjs of the BJT to zero, it should work. You can add the Cjs value to Cj of the diode in order to reach the same value again.
Thanks for reporting the problem! It will be fixed in the next release.
It works for me! Can you attach the schematic? Maybe there’s something else wrong.30. April 2021 at 9:35 in reply to: Transient simulation of a bipolar transistor fails #905
Can you share the schematic files? Otherwise it’s not possible to investigate the problem.
The ports of the hybrid are marked correctly in the schematic: A small “phi” between the ports with the phase shift and zeros between the ports without phase shift.
Furthermore, there are no input or output ports. The device is passive and all ports are bi-directional.
All results I’ve seen so far are correct, too! So I really cannot understand why some people are spreading around reports that claim something different. Indeed, this kills the project!
Of course, you can simulate a coupler with the phase that you want. But as the warning explains, physically, the result makes sense with 180 degree only. This means that in reality, (ideal) devices with other phases does not exist.
If a voltage step is applied to a capacitor, this results in an infinite high current. Especially, the trapezoidal integration method encounter problems here.
Anyway, it would be a good idea to catch such events in order to work around it. This is on the todo list for coming QucsStudio versions.
You can use indexing like var[1,2]. This is often used for s-parameters.
A better choice are usually the functions yvalue(…) and range(…). They are explained in the program help (F1 key).
Yes indeed, this is a bug. Thanks for reporting it!
It will be fixed in the next release.29. March 2021 at 10:57 in reply to: Strange phase angle results using phase versus wphase #844
This is indeed a bug. Thanks a lot for reporting it!
As a workaround: Write the term wphase(…) directly into the diagram. This works correctly.
The data in the smith chart is correct. Of course, the markers show the values of the s-parameters and not the corresponding impedances. Therefore the numbers.
Please use the function rtoz(…) to transform reflection coefficient (e.g. S11) to impedance.
Does this mean that the problem with the corrupt library is fixed now? Or does it still happen?