Please guys support the original qucs instead of qucs studio which is not open source but closed and users are dependent on the good will of the author.
Yes, QucsStudio is closed but it is improving, with the sole work of one maintainer. He asked for some support, without much success. Also, although it is closed, check the improvements of QucsStudio over Qucs and see if he only cloned the project and said it is his or if he really care about a better simulator.
Also to be clear, Qucs helped me a lot when I was searching free and open source simulator, and even showed me the world of open source, but the latest release is from 2017. QucsStudio still improving, and improving. Site, docs, tutorials, etc.
As both are good projects, and one is the foundation of the other, I think both deserve some credit and support.
Why? From a quick review of the videos, he has added some impressive features such as harmonic balance and planar EM. Then again, it’s nowhere near the professional tools such as Microwave Office, so I’ll continue to pay for those.
It's amazing how much is in Qucs already. I'll have to take a look. Agreed that the professional tools have a long and deep head start - I couldn't justify spending my time wrangling Qucs either.
What would experts here recommend for simulating 5ghz and 10ghz USB 3 on a particular gerber file and layup?
I've been avoiding it and spinning boards making small changes hoping for the best but still having issues. And don't have the resources to buy a scope capable of doing an eye diagram at 10ghz
Also bonus question.. How good are the simulations? Am I likely to be able to iterate on my layout and improve the eye diagram and order boards which have a good chance of working?
Well, I used it for a class, and it was fine for class purposes.
I did notice significant differences in modeled circuit performance between versions of Qucs Studio, for even very simple models. Instead of trying to get to the bottom of that problem, I just used the same version as the instructor and called it "good enough."
But "good enough" for a class is definitely not "good enough" for the workplace. Either I was doing something stupid (quite likely) or you should stick with ADS.
Actually the development of Qucs is ongoing, it's just that they are a little late on making the jump from Qt 4 to Qt 5, so Qucs isn't in available in some distributions that removed Qt 4 at the moment (e.g. Arch).
It works quite well for linear network simulations (eg: plugging together black box S-parameters from manufacturer data or extracted from a VNA) and it works reasonably well for filter synthesis. I've never used the harmonic balance simulator or the new planar EM sim though.
