## Re: [Qucs-help] Some comments and questions

 Re: [Qucs-help] Some comments and questions From: Stefan Jahn - 2009-05-08 06:27:59 ```Am Mo, 4.05.2009, 23:27, schrieb Alejandro Exojo: > Hi. hello Alejandro, > That brings my first comment: if not were for the filter synthesis, or > because > there are schematics posted on the web, I could not find how to do a > s-parameter simulation (and right now, I'm only interested in S-parameter > simulation, sorry). It's kind of tricky to find the proper component for > the > source and the load, and I find awkward that both the load and the source > are > a source-shaped component. I've only used Agilen'ts ADS before, so I don't > know if there is a de facto standard on this, or any rationale behind > this, > but in ADS the needed component is called "term", and is shaped like a > resistor. Well there is no defacto standard, but you can imagine a NWA port to be 50Ohm load with a switchable small signal source... that's why. For some tutorial about how to use Qucs goto: ; The S-parameter simulation explained in ; > Now my first question. In my first successful try, I wanted to do a > simulation > of a normalized low-pass prototype filter. The normalization makes the cut > frequency equal to 1 rad/s, but the plot is in Hz, so the important point > is > at 1/2*pi Hz, which is ugly. I would like to know if it's possible to > display > the plot in the angular frequency domain, or to tweak the X-axis to be > omega. > IIRC, in ADS you can change the axis to be 2*pi*frequency, which scales > everything, and does the trick. I tried with equations, without luck. You can do this by equations, e.g. S_Omega=PlotVs(S[2,1], frequency/2/pi) Then plot S_Omega. > And about the new version: I'm running 0.0.14, and I see that 0.0.15 is > out, > but I can't find any changelog, or list of features that I can check to > see > if it's worth building the source, or I will wait till Debian ships this > version. > The 0.0.15 announcement on the list should be on the archives, but > clicking on > that message you see a SPAM, not the announcement. Which makes me wonder: > what if the list is turned to subscribers only? I don't know if > sourceforge > allows it, but Mailman certainly does, and it will make the archives > usable > again. The ChangeLog has been posted on mailing list. Also in a NEWS item on SourceForge. And also you can watch changes at ; and ; > Oh, and, is there any plan to a Qt 4 port? Qt 4.x has many improvements, > and I > understand that porting QCanvas to GraphicsView should be a lot of work, > but > I think it's worth it just for the multiplatform factor. There have been plan. And Gopala started it 2 years ago (or something) but never reached a ready-to-use stage up to now. The Multiplatform factor is given already with Qt3. Cheers, Stefan. ```

 [Qucs-help] Some comments and questions From: Alejandro Exojo - 2009-05-04 21:27:37 ```Hi. Today I ran my first Qucs simulation. I'm studying filters right now at university, and I've used the filter synthesis tool, and it's quite convenient. Good! That brings my first comment: if not were for the filter synthesis, or because there are schematics posted on the web, I could not find how to do a s-parameter simulation (and right now, I'm only interested in S-parameter simulation, sorry). It's kind of tricky to find the proper component for the source and the load, and I find awkward that both the load and the source are a source-shaped component. I've only used Agilen'ts ADS before, so I don't know if there is a de facto standard on this, or any rationale behind this, but in ADS the needed component is called "term", and is shaped like a resistor. Now my first question. In my first successful try, I wanted to do a simulation of a normalized low-pass prototype filter. The normalization makes the cut frequency equal to 1 rad/s, but the plot is in Hz, so the important point is at 1/2*pi Hz, which is ugly. I would like to know if it's possible to display the plot in the angular frequency domain, or to tweak the X-axis to be omega. IIRC, in ADS you can change the axis to be 2*pi*frequency, which scales everything, and does the trick. I tried with equations, without luck. And about the new version: I'm running 0.0.14, and I see that 0.0.15 is out, but I can't find any changelog, or list of features that I can check to see if it's worth building the source, or I will wait till Debian ships this version. The 0.0.15 announcement on the list should be on the archives, but clicking on that message you see a SPAM, not the announcement. Which makes me wonder: what if the list is turned to subscribers only? I don't know if sourceforge allows it, but Mailman certainly does, and it will make the archives usable again. Oh, and, is there any plan to a Qt 4 port? Qt 4.x has many improvements, and I understand that porting QCanvas to GraphicsView should be a lot of work, but I think it's worth it just for the multiplatform factor. I think that's all. Thanks in advance, and greetings. -- Alex (a.k.a. suy) | GPG ID 0x0B8B0BC2 http://barnacity.net/ | http://disperso.net ```
 Re: [Qucs-help] Some comments and questions From: Stefan Jahn - 2009-05-08 06:27:59 ```Am Mo, 4.05.2009, 23:27, schrieb Alejandro Exojo: > Hi. hello Alejandro, > That brings my first comment: if not were for the filter synthesis, or > because > there are schematics posted on the web, I could not find how to do a > s-parameter simulation (and right now, I'm only interested in S-parameter > simulation, sorry). It's kind of tricky to find the proper component for > the > source and the load, and I find awkward that both the load and the source > are > a source-shaped component. I've only used Agilen'ts ADS before, so I don't > know if there is a de facto standard on this, or any rationale behind > this, > but in ADS the needed component is called "term", and is shaped like a > resistor. Well there is no defacto standard, but you can imagine a NWA port to be 50Ohm load with a switchable small signal source... that's why. For some tutorial about how to use Qucs goto: ; The S-parameter simulation explained in ; > Now my first question. In my first successful try, I wanted to do a > simulation > of a normalized low-pass prototype filter. The normalization makes the cut > frequency equal to 1 rad/s, but the plot is in Hz, so the important point > is > at 1/2*pi Hz, which is ugly. I would like to know if it's possible to > display > the plot in the angular frequency domain, or to tweak the X-axis to be > omega. > IIRC, in ADS you can change the axis to be 2*pi*frequency, which scales > everything, and does the trick. I tried with equations, without luck. You can do this by equations, e.g. S_Omega=PlotVs(S[2,1], frequency/2/pi) Then plot S_Omega. > And about the new version: I'm running 0.0.14, and I see that 0.0.15 is > out, > but I can't find any changelog, or list of features that I can check to > see > if it's worth building the source, or I will wait till Debian ships this > version. > The 0.0.15 announcement on the list should be on the archives, but > clicking on > that message you see a SPAM, not the announcement. Which makes me wonder: > what if the list is turned to subscribers only? I don't know if > sourceforge > allows it, but Mailman certainly does, and it will make the archives > usable > again. The ChangeLog has been posted on mailing list. Also in a NEWS item on SourceForge. And also you can watch changes at ; and ; > Oh, and, is there any plan to a Qt 4 port? Qt 4.x has many improvements, > and I > understand that porting QCanvas to GraphicsView should be a lot of work, > but > I think it's worth it just for the multiplatform factor. There have been plan. And Gopala started it 2 years ago (or something) but never reached a ready-to-use stage up to now. The Multiplatform factor is given already with Qt3. Cheers, Stefan. ```