Updated the description file for the board and added descriptions for the two daughter cards.
One of the reasons FPGA's are better for outputting data, is the 32 output state machines. If you look at the enclosed picture, it is of my controller outputting 32 universes of WS2801 data with 9 channels of data on each universe. The top line in the picture is the input Ethernet data. Lines D0 do D11 correspond to the clock on universes 1 through 12, and D12 to D15 corresponds to the clock on universes 29 thru 32. As you can see each of the outputs has a separate controller, so once the message is received, the controllers starts sending data to the lights. When you calculate the time it takes to output the data for a whole universe at the 781Khz clock, you get about 5mS. This means you have a maximum refresh rate of about 200 hertz, so you can do your 3D output video. 🙂
Very cool.
What are the little breaks in the bursts of the clock lines? Is that a sampling artifact, or are there pauses in the clocking?
If you actually try to refresh your pixels that fast, you might get some interaction between the pwm rate and your refresh rate, depending on the chips. 180x and 2811 have particularly slow PWM rates. Your 2801 should be better.
Zeph wrote:Very cool.What are the little breaks in the bursts of the clock lines? Is that a sampling artifact, or are there pauses in the clocking?
If you actually try to refresh your pixels that fast, you might get some interaction between the pwm rate and your refresh rate, depending on the chips. 180x and 2811 have particularly slow PWM rates. Your 2801 should be better.
The yellow signal is the data valid from the Ethernet physical. Since Ethernet expects breaks between the messages, the low spikes indicate breaks between the messages. Since I am transmitting 32 universes, there should be 32 clock high signals.
I agree, you could have interactions with the PWB if you refresh too fast.
Double Post

