Spiral inductor simulation software

[chunlee]

spiral inductor simulation

What is the best spiral inductor simulation software? (compare from price,accuracy,simulation time and etc.)

Is it HFSS, CST, Maxwell 2D and 3D, momentum, OEA and others?

Thanks,
Chun Lee

 

[itaifrenkel]

inductor simulation

Hello Mr. Lee

(proper disclosure: I work for Ansoft's representative in Israel )

I am assuming that you enquiring for RFIC spiral inductors. If you are using a different technology, please specify it.

RFIC engineers in Israel, and leading RFIC manufacutrers around the world, use HFSS for spiral inductor S-Parameters extractions. It allows you to accuratly simulate the Q-Factor which requires accurate simulation of the losses. HFSS can accuratly calulate the silicon losses (dielectirc losses) and the metal losses (inside the metal) and provide a broadband model of the spiral inductor (including DC losses).

The simulation time depends on the accuracy you would like to achieve, and your hardware. For about 2 percent accuracy, on a one year old computer, spiral inductor simulation would take about 10 to 15 minutes, and a complicated ballun about 20 to 30 minutes

Please note that in order to get fast and accurate results you will need to get HFSS training. The training material includes a spiral inductor example. In addition, you can ask your FAB representative to give you an example of an HFSS simulation. That way you will be using the correct stackup materials and HFSS settings that the FAB use. This is a great advantage since the FABs compare the simulated results to measurements constantly.

Another advatage for using HFSS is for capacitor modeling. In some cases the FAB does not provide capacitor models, and HFSS allows you to extract a model for the capacitor.

Then, if you have enough RAM (I'd start with a 3-4GB PC) - you can simulate complete filters or VCO layouts. For example two spiral inductos and 2 capacitors should take about 4 hours - without sacraficing accuracy.

One note about accuracy. CMOS processes has manufacturing variations. The role of accurate passive component extractions is to tune your design and increase the yield (the percent of manufactured chips that work), and thus increase the profit margin of your company.

Finally, if you require a robust harmonic balance and transient circuit solver that can properly handle boradband S-parameters (from DC up to GHz) ask your local Ansoft representative about Nexxim.

Hope this helps,
Itai

 

[mmicman]

circular inductor rfic

You might also consider looking at Helic. I have herad of people getting very good results vs HFSS, and in much less time. Spiral from OEA is also quite fast and accurate.

 

[rautio]

maxwell inductor simulation

If you are doing spiral inductors on silicon a paper I recently wrote with Rob Groves of IBM Fishskill is very important to read and understand:

A potentially significant on-wafer high-frequency measurement calibration error
Rautio, J.C.; Groves, R.;
Microwave Magazine, IEEE
Volume 6, Issue 4, Dec. 2005 Page(s):94 - 100

The described measurement error is particularly insidious because EM analysis can be modifed (by either user or vendor) to make the analysis agree with measurement, but the measurement is wrong. The article tells you how to detect and avoid this problem. I have visited numerous Si RFIC facilities over the last year, at least half of them have the described problem, and some of them have actually modified their EM software to agree with the bad measurement!

If you have circular or octagonal inductors and want a high accuracy analysis, read:

David I. Sanderson, James C. Rautio, Robert A. Groves, and Sanjay Raman, "Accurate Modeling of Monolithic Inductors Using Conformal Meshing for Reduced Computation," IEEE Microwave Magazine, Vol. 4, No. 4 December 2003, pp. 87 - 96.

For cases where you have very thick metal and narrow gaps, you should check out:

James C. Rautio, "A Space-Mapped Model of Thick, Tightly Coupled Conductors for Planar Electromagnetic Analysis," IEEE Microwave Magazine, Vol. 5, No. 3, September 2004, pp. 62 - 72.

I can email pdfs of any of these paper on request, or go to IEEE Xplore.

All the above work was done using Sonnet (I work for Sonnet), which you can try out for free with no time-out, SonnetLite at www.sonnetsoftware.com.

As for which tool is best...they all are, depending on what you consider to be best. They all have advantages and disadvantages, which would require some time to go through. Just keep in mind that every advantage is paired with a mating disadvantage. They go together like heads and tails on a coin. If you are told an advantage without being told the mating disadvantage, you are not getting the complete story.

 

[e.Horus]

inductor simulator

For quick simulation of inductors up to few GHz you can use ASITIC from Berkeley. Not really as accurate as HFSS, yet again I doubt the transistor models are as accurate. It comes free!!!

 

[rodzski]

inductor measurement berkeley

The newest and best EDA SW for this problem comes from Lorentz Solution in Silicon Valley. They have a product called PeakView that addresses the Passive Component Design (EM Synthesis) and the on-chip EM coupling challenge. PeakView is deeply embedded within the Cadence RFIC design environment making it very easy to use. The underlying core technology is a new IC-focused Full-wave 3D EM engine that runs at least 10-100X faster then other general purpose EM Solvers with accuracies that match silicon measurements. You guys should check it out.....see www.lorentzsolution.com for more info.

 

[doug freeman]

emss feko price

rodzski,

Please do not come on here and spam your product without telling everyone your relationship to the company. Jim Rautio, Ray from Feko and others always state that they work for a certain EM company. To not do so is rude and deceitful.

Also to boldly state that it is the best is ridiculous unless you can back this up technically which doesn't seem to be the case from your website.

 

[RayFEKO]

inductor simulation cst

Hi Chunlee, just for record for others who might also be interested(as your post is kind of old), at the link below, there is a video demo showing how FEKO can be used to simulate a spiral inductor. This demo should be also useful for those who want to see how the new version of FEKO looks like.

http://www.feko.info/video_gallery.html

BTW, as Doug Freeman indicated, I am working with EMSS(the company for FEKO). So feel free to contact me if you want to learn more about FEKO. Thanks.

 

[mwmmboy]

simulation sofeware for inductor

OEA eda tools can be ok for your purposes.

Bye

 

[haisuki]

inductor simulation software

I do agree with e.Horus, ASITIC is a very good starting point and it's free. I myself have been using it and there is just a small deviation in the inductive value.

For a more accurate tool, you could use momentum (I do not know the other tools you mention), but take into account that all the results depen strongly on how you describe all the parameters (characteristics of the substrate, layers, ...)

 

[khouly]

ads momentum peakview and helic

u can start with ASICTIC , but u must write the technology file for it carefully

then u have to use some EM simulator like
Momentum "ADS"
IE3D of zeland , it is great and easy to use
Sonnet : never used it , ppl saying it is great
HFSS : full 3D EM simulator ,
CST : also Full 3D Simulator


note
"many ppl used momentum with great accuracy"

khouly

 

[flywke]

software spiral inductor

Maybe you can try IE3D. It is a good simulation tool for planar structure.