From: Stefan Jahn <stefan@gr...>  20050706 06:28:26

On Mi, 6.07.2005, 05:24, John Steenbergen sagte: > Stefan, Hello John! > I tried to put this on the mailing list but I could not manage to get > subscribed to the help list for some reason. I was wondering if you could > help me with this. I am doing my best. > I would be willing to write up some > documentation/example > on this subject if you would like. This would be great. See qucs/qucshelp/docs/tutorial/ about how to do this. :) > I am trying to calculate constant gain circles using a 2port sparameter > file and the equation solver. The schematic is able to produce stability > circles just fine unsing the quation solver. Simulating the schematic > does give me an warning that S is not a defined varible and an error that > Gp is an undefined varible. The equation I am using is: > Gain GaCircle(S,Ga) > When only these equations are used it simulates just fine but with some > warnings: > StabSource StabCircleS(S) > StabLoad StabCircleL(S) I'll look at this tonight... But a first glance tells me: Ga is really undefined. In order to compute gain circle(s) with constant available gain Ga in the source plane you need to specify for which gain. Thus Ga=10 OR Ga=linspace(5,15,11) OR Gain=GaCircle(S,10) OR Gain=GaCircle(S,linspace(5,15,10)) may work better. > I am wondering if I need to define Gp and/or Gs or if they need to be > calculated? Not calculated but spcified. > I am attaching the schematic and s2p file just in case I > missed anything. Also I am using Qucs 0.0.6. I think you missed the s2p file to attach... Can you send it to me? > Thanks in advance. This program is really great. Thanks for all the hard > work. You're welcome. Thanks for trying Qucs, Stefan. 
From: Stefan Jahn <stefan@gr...>  20050706 07:25:18

On Mi, 6.07.2005, 09:08, John Steenbergen sagte: > Thanks, Hi! > The last program (SuperStar) I used plotted the maximum gain at the > calculated frequency so that was never a parameter. That clears a lot up. Hm, that may be an option... Possibly we can do that as well. Michael: What do you think about it? > I > really appreciate all the help. > Once I get this to work I am going to try to convince my old professors to > use this instead of SuperStar which they pay a lot in licences each year. If Qucs fits your needs... Cheers, Stefan. 
From: <Michael.M<argraf@al...>  20050708 06:09:38

Hello, >> The last program (SuperStar) I used plotted the maximum gain at the >> calculated frequency so that was never a parameter. That clears a lot >> up. > > Hm, that may be an option... Possibly we can do that as well. > Michael: What do you think about it? The gain needs to be a parameter! Not only the maximum gain is important. But perhaps the maximum gain can be the default value if none is specified ? Bye, Michael 
From: Stefan Jahn <stefan@gr...>  20050708 08:20:07

On Fr, 8.07.2005, 08:09, Michael.Margraf@... sagte: > Hello, Hi! >>> The last program (SuperStar) I used plotted the maximum gain at the >>> calculated frequency so that was never a parameter. That clears a lot >>> up. >> >> Hm, that may be an option... Possibly we can do that as well. >> Michael: What do you think about it? > > The gain needs to be a parameter! Not only the maximum gain is > important. But perhaps the maximum gain can be the default > value if none is specified ? That's what I meant. Using the maximum availale gain if no specific gain is given. Stefan. 
From: Stefan Jahn <stefan@gr...>  20050822 07:11:16

On Fr, 8.07.2005, 21:33, John Steenbergen wrote: > Stefan, Hello John! > I tried to use the equations: > K = Rollet(S) > Gmax = abs(S[2,1]/S[1,2])*(Ksqrt((K^2)1)) > > Also reference eq. 1.146 in the technical documents. > > To calculate the max gain circles, but the result was way off the smith > chart. > > Any ideas? I am sorry that I did not respond for such a long time. But now I had again a look at my mailinbox... I assume the mails were about your 'NE85634107.S2P' file. If you look at the Rollet (K) factor you see that it is uncondiionally stable at about 1GHZ (K > 1 and det(S) < 1). That is why the the gain circles are off the smith chart. They are inside the smith chart if the twoport is unconditionally stable and the given gain is less or equal than the maximum gain. If you compute the maximum gain at a stable point and then compute the gain circle the radius is near zero. So GaCircle (S) with no gain given, looking for the maximum gain and computing the gain circle makes no sense, because it computes a point, not a circle. Now I wonder what all this is about? Please send an example what you have done, where the problem is and what you wanted to simulate and what you expect to see. >> According to my notes the max gain is: >> >> Gmax = mag or abs (S21/S12) * (K  (K^21)^(1/2) >> >> Where K is the stability factor. >> >> Hope that is a help. >> >> John >> >> >> On 7/8/05, Stefan Jahn <stefan@...> wrote: >> > >> > On Fr, 8.07.2005, 08:09, Michael.Margraf@... sagte: >> > >> > > Hello, >> > >> > Hi! >> > >> > >>> The last program (SuperStar) I used plotted the maximum gain at >> the >> > >>> calculated frequency so that was never a parameter. That clears a >> > lot >> > >>> up. >> > >> >> > >> Hm, that may be an option... Possibly we can do that as well. >> > >> Michael: What do you think about it? >> > > >> > > The gain needs to be a parameter! Not only the maximum gain is >> > > important. But perhaps the maximum gain can be the default >> > > value if none is specified ? >> > >> > That's what I meant. Using the maximum availale gain if no specific >> gain >> > >> > is given. Cheers, Stefan. 