Looking for a best antenna software

[tomath]

Hello,
The antenna I want to simulate is the following

Probe fed patch antenna (on 9cm x 9cm dielectric) will reflect off on 1m dish. There is a 3U satellite structure on the bottom of the patch antenna. The antenna will have frequency at 8.4 GHz.

I have tried to run this on HFSS, but I think the size of the antenna was too much to handle for FEM solver. It eventually ran out of memory. What program would be the best for this simulation? I know FEKO is one, but is there any others? Or is there a way to set up HFSS (Like special ways to set up Mesh, Analysis, Sweep etc.) to make the program run?


Thank you in advance!

 

[BigBoss]

HFSS is a good 3D simulator but it's memory eager..
If you use 64bits OS and sufficient memory, you may simulate your system.

 

[tomath]

Thank you for your response. I will try it on PC

 

[vfone]

Yes, HFSS and CST also, both are ridiculous hungry resources programs. During years, always their software developers were few steps back understanding the average processing power of the computers on the market at that moment.
In other words, most of the time doesn't matter on what computer you run these programs, you have to take your time (break, nap, go home) when running complicated structures.

 

[biff44]

memory is ridiculously cheap for a PC nowadays. Just max it out. If you can not do it yourself, take it to a computer store and have them install it.

Also, change the mesh size so that so much memory is not needed.

Or better still, replace your hard drive with a solid state hard drive and reinstall your OS and emag program solely onto it.

 

[vfone]

I just proved to myself, there is no chance to make it MUCH better even if the computer is an i7 with 16Mb of RAM and using SSD HDD. There are minimum requirements for a platform, but if you use a platform better than this, you don't see bigger improvement when simulate very complex structures.
To make an EM simulator running faster when doing very complex structures it is a full time job for a big group of software developers. Always they claim that new program versions are running faster and faster, but in fact this is just a marketing strategy.
The increase in overall speed of new program versions most of the time is based on new computer performances, and not because they spent time and money in this regards.

 

[tomath]

Thank you all for great tips. I am still a student and am learning a lot from this website about antenna software and computer usage.

Also, are there any tips on meshing? I just set my radiation box to have maximum length of lambda by 10 and for the reflector structure 60 degrees. Is this ok or is there a rule of thumb?

 

[biff44]

yes, there does seem to be a limit you run into. I am not an expert on emag programs, but it seems that some are not written to efficiently use all the cores and threads available, and a bottleneck happens. Heck, with some of these kepler video cards available...one could have thousands of cores available for processing--that SHOULD mean an analysis would take a second or two...but does not happen in real life.

One thing that ALWAYS does work is to overclock the CPU. Get a fast cpu, water cool it, and overclock that puppy.