View Single Post
Old Jan 08, 2013, 06:39 AM
Gary Evans is offline
Find More Posts by Gary Evans
FPV Desert Beta Test Center
Mesa, Arizona
Joined Nov 2006
2,577 Posts
Quote:
Originally Posted by Ascended View Post
Sorry ElCheapo but I have to totally disagree. Processing power and complexity are no problem at all these days. If you're designing in EaglePCB and can't afford a Multi-thousand dollar IDE - then yes, it is too complex and impractical. The hobby community just doens't have the skill set to do it, and those who do make more money than this hobby could provide on other projects - it's not worth it for us.

It would not be hard to do this at all, I could do it with two IC's and a small ARM chip + a ram chip. The solution requires 3 different voltages to work, and would be a bit bigger than a credit card, but it would be doable very easily.

If i wanted to do it right I'd just use a FPGA. You can even buy video decoder/encoder cores which would do 90% of the work for you for the overlay. These however cost $15k each just to get started (the core, no the fpga) and need double the ram of the discrete solution - however would fit into a much smaller package. This would give you a frame of lag however.

Then there is always Qualcom and TI's DSP ARM Cortex A8s. They have video decoders built in, and someone with enough time can simply program the OSD onto the video picture then send it to the analogue video output. Again, requires 2 powersupplies and RAM, and also provides a single frame of lag. In a 12x12mm package with a 1ghz cortex A8 and a 800mhz DSP there isnt very much you can't do in a small space with a DaVinci processor. You can easily do PIP and multiple video overlay streams with one. At $30 for the chip alone and at least a 6 layer board it's not going to be a cheap OSD to manufacture.

Either way, this is more like 6 months of working evenings on a project, vs for a black and white osd not using a max7456 a week or so (just for the overlay portion.) You'll then need to program in all the nice to have's like return to home, graphics, sensor handling and likely an AHRS system. The return on investment is not really there to make it worth while investing enough time to make this system. If its a labour of love as this seems to be (considering 4 years of development with no release), rather than something developed for sale - it's a totally different story. For those of us making a living from electronics.. its not really worth it.

For not agreeing with him your last line states the same conclusion.
Gary Evans is offline Find More Posts by Gary Evans
Reply With Quote