Hardware Firmware
JohnP
Toolchain Ishani
Bring the Awesome
At long last (BatchPCB shipped them over two months ago), the 16 x 16 boards have arrived. They were most likely a victim of the postal strikes here in the UK.
Big Chill
I dragged Ishani’s Rev2 to the Big Chill this past weekend. Initially, we were planning on making a run of the stripped-down Rev3 to hand out, but a stack of other committments got in the way.
In the process, I learned some valuable lessons:
- Battery life “in the wild” will always be shorter than expected. We got between two and three hours out of it with a typical pattern. Field-replaceable batteries might be a sensible option.
- Perceived brightness varies widely with ambient light (duh!). While having panel intensity embedded in the animation data is nice, some sort of brightness control on the device would be nice as well as useful for extending battery life.
- Laptops are not particularly festi-friendly. Even my little EEEPC was a bother to tote around. Aside from swapping patterns from time to time, it was only used for recharging.
- Finally, there was so much going on to hear and see I wasn’t hugely interested in playing with silly electronics!
It's alive!
Rev2 is up and running, blinking and blending away on Ishani’s desk.
After much cogitation, it would appear I swapped data in and data out to the MAX6960 while swapping pins in the firmware to match the new board layout. I can never remember from which perspective those are supposed to be interpreted.
As always, there are still some issues with Rev2:
- When it is turned off but plugged in to USB the display still runs. It appears to be a case of power being pushed through the clamping diodes on the AVR. It’s either the USB interface (which is bus powered) or the charge indicator (which is foolishly connected directly to the bus power. Regardless of the cause, it means that when plugged in to USB it should either be switched on or have the display disabled in firmware.
- The buttons are ridiculously small, so much so as to be completely unusable. Future revisions will replace with a single, larger button.
- As mentioned before, the strap tabs are on the wrong way around and obscure the switches and USB connection.
Pushing forward with Rev3 (which should be renamed as it shares very little with the previous designs) in the meantime.
We are experiencing some delays getting Rev2 working. It passes all of its self tests with the exception of the display. Communication between the two boards appears to be the same as with Rev1 but fails to produce an image.
Pleased to have my logic analyzer around. Beats the hell out of chicken guts or tea leaves for figuring out what’s gone wrong with inter-chip communication. There may be subtleties to timing that I have missed.
Rev2 boards are back...
…and lookin’ pretty sweet (will post pictures shortly). I put the tabs on the display board the wrong way around, so the USB connection and switches will be obscured by the strap. I think we’ll assemble two anyway, for fear of never getting anything done.
Still having trouble sourcing the MAX6960s for any reasonable price, which is a shame. I’ve not been able to produce quite the same display quality with Rev3’s ISR-driven approach. Specifically, balancing the red and green is proving very difficult.
