Designer’s Guide Community

Tino

New Member

Offline

Posts: 5
Germany

Constant forward power supply for simulation in Spectre
Apr 13^th, 2011, 4:13am

Hi guys, please help a newbie out!!!
For simulation purposes, i have to implement in Spectre (Cadence) a source or power supply which provides a fixed forward power independently of the load (which is complex in most cases) and the frequency. Basically, i want to supply a load which constantly changes its impedance. As an example, let's say the load which is directly to the source has 10+200i at 10MHz and 50+50i at 500MHz. The source should be able to provide a constant power of 10dB to the load under these different conditions.
In Spectre, there is a component called port which unfortunately only delivers a fixed incident power (Pinc). But i am actually interested in providing a fixed forward power (Pfor). Of course the formula Pinc = (1/(1-|gamma|�))*Pfor might be of help, but this implies automatically measuring the impedance of the load for each condition, calculate gamma and then Pinc, but i don't know how this is feasible in spectre Sad

Please, does anybody knows the best way to implement this in Spectre (circuit? script? component?)... Any help, suggestion or advice would be so much appreciate!
Thanks!

Back to top

IP Logged

Tino

New Member

Offline

Posts: 5
Germany

Re: Constant forward power supply for simulation in Spectre
Reply #1 - Apr 13^th, 2011, 1:57pm

Dear pancho_hideboo,
thank you for your answer. Sorry if my question my sound stupid, but what should be the respective parameters or products of the simulations. Did you mean, i should use AC or SP analysis to obtain the load impedance over the frequencies and then use these results in a second step (PSS or QPSS) to calculate the forward power needed for each frequency?
Like i said before, im new to this so please bare with me. I would be very greatful for more details concerning the implementation.
Thanks a million!

Back to top

IP Logged

Tino New Member Offline Posts: 5 Germany	Re: Constant forward power supply for simulation in Spectre Reply #2 - Apr 13^th, 2011, 1:59pm I have forgotten to say that all my models are only available in Spectre, so unfortunately, the implementation can't be performed in other platforms (ADS, etc...) Thanks!
Back to top	IP Logged

sheldon

Community Fellow

Offline

Posts: 751

Re: Constant forward power supply for simulation in Spectre
Reply #3 - Apr 13^th, 2011, 9:20pm

Tino,

� This is a case where you need to think Virtuoso, environment, and
not just Spectre, simulator. There several options for implementing
this measurement. Here are the details on two approaches,

Approach #1: Use Pre-Run Skill scripts [in the environment]
Assuming that you are using IC615 and ADE-XL, then use the pre-run
Skill script functionality. Basically, you want to make a measurement
and then adjust the testbench based on the measurement results. The
pre-run Skill script function is intended for this application.

In one case, I used this approach for simulating Bandgap
Reference variation after trimming
1) Simulator the the bandgap voltage, in your case simulate "gamma"
2) Adjust the trim circuit based on the bandgap voltage, in your case
� �adjust the power level
3) Characterize the bandgap, in your case characterize your design

Approach #2: Use OCEAN [batch mode]
You could write an OCEAN script to do the measurement. I would
dump the OCEAN script for testbench and add the gamma
measurement to the script.

� � Don't think that one approach is better than the other. The
difference is that in the environment, you need to be on IC615.
Using the OCEAN script is version independent.

� � � � � � � � � � � � � � � � � � � � � � � � � � � � � �Best Regards,

� � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �Sheldon

Back to top

IP Logged

Tino

New Member

Offline

Posts: 5
Germany

Re: Constant forward power supply for simulation in Spectre
Reply #4 - Apr 14^th, 2011, 1:07am

Hello sheldon,
thanks for your reply. Your idea sounds pretty good, and i think i will go for OCEAN, since it's version independent. But like already mentioned in my first post, i don't know how to precisely measure the load impedance in Spectre. Would an AC-Analysis be enough, or do you think there might be some better alternative.
Thank again

Back to top

IP Logged

sheldon

Community Fellow

Offline

Posts: 751

Re: Constant forward power supply for simulation in Spectre
Reply #5 - Apr 14^th, 2011, 4:04am

Tino,

� Yes, ac analysis should work for measuring impedance. You might
also want to output the z-parameters.

� � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �Best Regards,

� � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � �Sheldon

Back to top

IP Logged