12-03-2009 05:02 PM
I am looking for help in coming up with a suitable project for my final year masters in Engineering.I would appreciate any suggestion which utilises xilinx resources such as picoblaze, spartan fpga e.t.c
Your help is much appreciated
12-03-2009 11:13 PM
get some AES IP-core and a picoblaze, put some extra registers around this and implement an AES-Function with all modis selectable like ECB, CBC, OFB, CFB, etc..
That should be sufficient for a master thesis, especially when you are targeting high data rates.
Have a nice synthesis
12-04-2009 03:57 AM
Normally the most significant constraint on any project is what hardware you will have access to or will be capable of fabricating in the timescale. If you are learning how to design with Xilinx for the first time then I’d have to suggest that you work with known good hardware (i.e. one of the development boards) to minimise the number of unknowns in your work.
It is probably best to consider working only with the hardware available or building things that connect to an existing development board. For example you might mound a Spartan-3AN starter kit onto a vehicle made out of Lego Technic or something and just build some motor drivers etc.
As to what project then the list is almost endless and so depends on what your direction is. Are you more into software or hardware? Do you have a particular interest in a certain area such as signal processing, encryption, image processing, motor control, robotics, voice recognition, neural networks, self-modifying code, code development tools, graphics, gaming……
Some projects can really be like product designs whilst others can be far more ‘research’ in nature. I suggest you find out what is expected as a deliverable. It may be better to have something simple that works than something complicated that is experimental in nature and form the outset has an unknown outcome.
But to just throw out one idea that comes to mind right now. This one can sit nicely on a standard Spartan Starter Kit and the complexity could be adjusted to fit your time scale but still include aspects of research and require some careful engineering. Power consumption is common concern in our industry and we know that power is directly related to frequency. Power management schemes often turn off circuits that are not actively being used or run them at lower frequencies. So how about investigating this on FPGAs using frequency ‘throttling’ to continuously match the clock rate to the demands of processing at a give time for a given function. You could control the frequency that different (PicoBlaze) processors are operate and also consider when it makes sense to use hardware circuits rather than processors and change the operating frequency of those circuits too. Have a careful look at the reference designs on the 3E and 3A Kits and you will see some useful building blocks and related information. It should not matter that you are working with a relatively small device and a few hundred mW; this whole subject is applicable to the small mW portable market as a mobile phone base station where devices need cooling. You could even say it is an environmentally friendly project!
Do tell us what you decide.
12-06-2009 06:02 PM
Thanks for the response, the deliverable is focusing on a product designs ,so that's the are a i am focusing on. I am good in both software and hardware so an idea would be helpful it can be a simple idea which i can develop into a proper product design. A the moment i have Spartan 3E starter kit however if the project requires something different the starter board is open for change.
12-07-2009 02:09 AM
Try a search for "PicoBlaze" in youtube and this may also give you some idea of what you can achieve with a small board.