Help! - RF Cafe Forums

 Post subject: Help!

Posted: Thu Apr 23, 2009 7:46 am 

Hi everybody!

Excuse me, I'm working on a University research about CAE softwares.. ..and I'm in trouble..

I would be very pleased if somebody could help me about these topics:

1. Which are the main problems you find using them?

2. Have softwares that can do multifield analysis (electric, magnetic, thermic, etc.) reached good performances?

3. Are there softwares that can do structural analysis and multifield analisys at the same time?

Thank you very much!

 Post subject: Re: Help!

Posted: Fri Apr 24, 2009 3:52 pm 

Hi Sonia,

I can answer on some of your questions:

1. CAE tools are used to design electronic, mechanical systems or to solve and analyze problems in these fields (Although there are also other engineering or scientific fields which these tools are useful for). In the use of desiging new systems, the purpose of using these tools is to see the expected results and discover potential risks/problems before investing money on creating prototypes and/or doing mass production.

2. Yes, there are tools which are able to perform multiple simulations for different topics and they are able to reach good solutions.

_________________
Best regards,

- IR

 Post subject: Re: Help!

Posted: Sat Apr 25, 2009 6:09 pm 

Just popping in...

Being an analog engineer who works with sensors, I use three main packages for analysis, and four more for design:

For analysis,
1. Pspice - Circuit analysis / stack up
2. MathCad - General analysis
3. FlexPDE - Partial differential equation solver for fields

For design
1. Paper + Pencil
2. Orcad - Schematic capture
3. Pads PCB - Component documentation / PCB layout
4. ispLever - Design of programmable logic

As for trusting the tools, that varies with the problem. Like most engineers, I reuse models from prior work. These reused parts are fairly trustworthy. New parts are suspect, and I generally run tests with predictable outcomes. For example, I'll check curves for an active device against it's data sheet. I also make a point to have two sets of eyes review the schematic / pads agreement for each new part. Miscommunication between documents / individuals is common and costly.

I NEVER trust field models which haven't been verified. I always build simple geometries first, and check the result against established solutions. Particularly with magnetic and thermal designs, this can save huge headaches.

First and foremost - The tools are resources that are available to improve your yield - as a resource. Your foremost mission isn't to fully understand, or to make perfect - it's to provide return on investment. Keep this in mind when picking and using tools, and you'll be a more valuable employee. - Mike

 Post subject: Re: Help!

Posted: Sun Apr 26, 2009 2:17 am 

For circuit simulation:

I personally used a myriad of RF design tools such as ADS, Microwave Office and Genesys.

For linear simulation, all of these tools are the same as the solve linear equations and therefore provide identical results. For non-linear simulations there might be some difference because different tools use different algorithms to converge to final solutions.

For EM (Electro-Magnetic) simulations, there is a hugh difference between different tools because they use completely different methods to solve, some of them are 2.5D, some of them are 3D. Here the cost of the software plays a major tool on the amount of memory provided, the solving time etc.

For Schematic Entry/PCB layout

I used OrCAD, Dx Designer (Viewlogic), nowadays I am using some low-cost tools and eventually will use Altium Designer. The cost of the software also plays here a major role.

I always start a new design with a paper and pencil to gather my thoughts and to see how I fit the design to meet the requirements, and then I am heading to do some simulations (depending on the complexity of the design).

In any case, and this any engineer in any field will tell you, that once a design is completed, it will be tested - no matter how good a simulation tool is, there is still no alternative to test and verification.

_________________
Best regards,

- IR

 Post subject: Re: Help!

Posted: Mon Apr 27, 2009 7:56 am 

Hi Sonia
My dime's worth, regarding Q1: since I ignore pirates advocating KRAK sources of password cracking stuff, I finding the biggest problem is the price.
Re Q3, since many electro-mag problems can be analyzed by Finite Element analysis software, and such an approach answers most structural analysis too, then a good Finite Element software ought to have bolt-on front-ends that can handle either sort of problem.

_________________
At bottom, life is all about
Sucking in and blowing out.