felix-salfelder
commented
1 month ago On Wed, Oct 23, 2024 at 09:33:14AM -0700, domi8space wrote:
I am new to GitHub, which i join to help improving Qucs-based softwares (including Qucs-S), and this is my first post.
(Sorry about Github. Many would prefer keeping the stuff together. E.g. on the mailing list. But switching off the message boards here might leave us with even less ...)
Thanks for sharing. Can you give a quick explanation how you would use the function? Do you intend to use it in device modelling or in postprocessing? Do you need real or also complex arguments? How about the branch selection? Please provide an example use, e.g. in terms of a Qucsator netlist?
Most numerical computation softwares (Maple, Mathematica, Matlab, Scilab, Octave, Python, etc.)
I found a C++ implementation in scipy under a suitable license. According to a comment in the code, it has an issue "close to the branch point". It uses some of the scipy infrastructure. But it might serve as a starting point.
ra3xdh
domi8space
Hi all,
I am new to GitHub, which I join to help improving Qucs-based softwares (including Qucs-S), and this is my first post. Unfortunately, despite being an extensive user, I fear I cannot bring much in coding skills...
A famous mathematical function, often found when it is needed to solve equations with a linear term and either a logarithmic or exponential term, is lacking in the Qucs family's set of mathematical functions: the Lambert's W function. https://en.wikipedia.org/wiki/Lambert_W_function
There are some papers which give some examples of such electronic circuits, and even Wikipedia mentions the "diode and resistor" circuit solving with this function.
Most numerical computation softwares (Maple, Mathematica, Matlab, Scilab, Octave, Python, etc.) have this feature. Because this function can solve many problems that can be found in electronic circuits involving diodes or bipolar transistors, I think it would be useful to see it in Qucsator-based simulation softwares. I personally have to export the simulation data from Qucs(-S) and use an external software to plot the simulated results with the theoretical ones that use the Lambert W function, which makes it painful for repetitive iterations.
Thanks!