problems with co-simulating oscillator in GENESYS

[majorbra]

co-simulating in GENESYS

I´m simulating an oscillator in genesys. Simulating just with harmonic balance and schematic OK, but when I run a co-simulation using Layout and EMPOWER, don´t work.

messsages:

Warning: Element OSC1 REAL.SG1 has 2 nodes in the schematics and 1 ports in the EM simulation. The extra nodes in the co-simulation will be grounded.

HB: Oscillator frequency not found run simulation(s) aborted

anybody can give me an Idea?

circuit is attached.

tks all.

 

[RF-OM]

Re: co-simulating in GENESYS

Unfortunately I cannot download any file that is more than 500 kB, so I cannot check your design file. There should be special port for each node when you do co-simulations. The best results when you use Modelithics models. I think you need to check the manual for Empower and you will find how to do co-simulation and assign the ports. What version of Genesys do you use? If I remember right latest versions do it automatically when synthesize the circuit and layout. It may be harder when you create schematic and layout.
Error messages are talking that there is not as many ports as nodes and that extra nodes therefore will be grounded. This is why HB simulation is aborted.

 

[majorbra]

Re: co-simulating in GENESYS

TKS RF-OM.


My genesys version is 2004.07
The problem with extra nodes, I fix it. My problem now is in find the oscilator port.
I made the file smaler, Can you see it now?

regards.

Majorbra.

 

[RF-OM]

Re: co-simulating in GENESYS

I download and try to open the file. Here I am out of luck. File was open but with long notes about Workspace Conversion to my version 2007.03. Simulation started but then stopped with error message: "Cannot find specified design (EM1) for simulation". This simulator is very busy now and unfortunately I cannot investigate the problem.

Do you have manuals for your version? I have them in pdf for 2004.03 (we used it before new one), so if you need it please let me know. I believe they should be the same as for 2004.07.

 

[majorbra]

Re: co-simulating in GENESYS

TKS again RF-OM.

OK, I have the version 2007 too, I Will Instal in my machine and try solve this problem. If I have the same problem with version 2007, can i contact you?

regads.

 

[RF-OM]

Re: co-simulating in GENESYS

Of course. This is why I joined this board. I cannot promised to solve 100% of your problem, but if I can help I'll do my best.

Be careful, if you work only with 2004 version, this conversion problem will be new for you (as for me too) and may require some time and efforts.

Best regards

 

[majorbra]

Re: co-simulating in GENESYS

RF-OM,

I made a worspace in genesys 2007.08, The problem in find the oscport remain, the simulatos says that the design no include an oscport element.
But in sch i put one.

the file is attached again,

regards,

majorbra.

 

[RF-OM]

Re: co-simulating in GENESYS

I got the file and unrar it. But I can check it only on Monday in office when Genesys will be available. My version is 2007.03 but I think it should be able to open your file.

 

[RF-OM]

Re: co-simulating in GENESYS

I was able to launch the file today. I have some questions about it.

1. Layout looks strange, like it is just after Genesys synthesized it. Probably it is necessary to move parts in order to eliminate shortages and metal crossings as well as metal overlays. Then you need to include EM ports into layout. Program cannot simulate without ports. But may be this mess was created by conversion from your file to my version.

2. Your footprints for L,C and R are all 1206 which is probably not right despite it is Genesys' default footprint. Check other parts footprints also.

3. Layout properties have dimensions in mils, but substrate has mm.

4. In schematic simulation you used only one port for harmonic balance. Usually when oscillator is analyzed it is done with so called open loop analysis. May be this is important point for you. Please let me know if you need additional information on this issue.

5. Physical loss simulation is turned off, this may be a reason of slightly shifted frequency and higher amplitudes.

I think that first of all you need to organize layout: use right footprint; put part in right orientation and right order; eliminate shortages and overlapping; add EM ports; carefully check all simulation parameters. Definitely in the shape what I can see now EM simulation cannot work.

 

[majorbra]

Re: co-simulating in GENESYS

RF-OM,

Answering your questions :

1-certainly there was a problem in conversion. My circuit looks like attached file image.

2- problem of conversion too.

3- Really I forgot this detail. BUT the parameters appear right in in layout properties, I suppose genesys make the conversion internaly.

4- I suppose the problem is Here.
I´m trying to simulate in close loop directly beause that is more realistic.
however, genesys can not "see" the oscilator port. Do you agree?

5- turn off just as first step and time saving.

 

[RF-OM]

Re: co-simulating in GENESYS

You are right. There is huge difference between your layout and what I saw on my screen. I am not in computer with Genesys right now, but remember this picture well. Frankly speaking I never try to do one port EM simulation and cannot tell you is it possible or not. It may be not because s2p S-parameter file cannot be used in such a case. But I am not sure. I use EM simulator Empower only for full two port simulations. Even when simulated just a piece of PCB trace. Some simulators require to terminate unused port when it does not used. But I am sure you need EM port(s) that are missing on your picture. You need to check this point. Do you have Randal Rhea book on Oscillator Design? I think there you can find all the answers as well as into Empower manual.