One of the harder things about doing conference badges is getting everyone involved so we had two ideas for a badge, an im-me clone with a blackberry keyboard (neat) has more hacker cred, but less overall appeal. But people love the blink, so we figured lets grow a base platform, same CPU, same LED's and the Blink people love.
Add on a mini prototyping area so people can add their own components, something that has been popular in the past, and make the Wifi work ( ran out of time last year to get it working before conference) . I literally just got it working two minutes before i started this post. Two weeks to go.
5x4 (why not 5x5 charlie) WS2812B's
Mini breakout area
PSOC4 IO's broken out
ESP8266 (working this time) with the GPIO's added to the PSOC4 so they can interact, and the ESP8266 can be reset
And we'll put together a bunch of mini projects that can be easily added, speaker, switches, IR, photosensor, fli
Picked the right diode it seems, I was worried since my board with two diodes wasn't doing well, but forgot we weren't using the logic level converter, added that and dropped the two diodes to one and works up to 7V , 6.4V is the highest with the
Running in WiFi mode. Inside a ER20 tool holder box.. Music is quiet in the background but you can hear it later in the video.
Continuing to build the stencil holder and start pick and placing. Built 8 badges by hand so far.
6 days to go..
Unfortunately we are low on ESP8266's since we gave them away like candy last year, and couldn't get them delivered in time.
I had another crazy work year, so i was super behind on the Badge this year, i finished the eagle design a couple of weeks ago, and the PCB's arrived on Friday from DHL.
I always run into the issue of scale when doing boards in eagle, and some of the logos didn't turn out as well as i wanted. In fact some of them (ours) are so subtle as to be called into question of existing.. So next year, lots more print outs., and possibly a larger badge
Grabbed up all the spare components from last year, only missing the SMA diode( reverse protection and voltage drop) and the logic level inverter.
I bought this battery charger from CVS or some such in San Francisco when I was at GTC, after a few days all it did was shine the LED's continually so I let it run out and decided to take a look inside, after opening the case carefully (threw it against the concrete floor) opened it to reveal this, oddly i have that same exact board from an aliexpress binge last year, but i digress.
cool, as expected the 18650 and a charging circuit, so i carefully binned it , since who knew the state. Later I realised i know where I saw an SMA diode looking like an RP, so grabbed it back out of the bin , removed the diode and back into ewaste it went.
Next was the level shifter, i thought i had a bunch lying around but nope, I had one board from last year ( i never actually got a full working badge as i swapped mine out with someone who destroyed theirs, which is now mine) it had a CPU, one LED and as luck would have it a level shifter, so pulled that off.
"My" badge from last year... IC1 is what we want
Great now the build out begins.
sorting out components
add some tools
clear a 10cmx10cm space in the garage
build the minimum components to get it to work, or so we think.
Wait, that board from last year i ended up with it, because it didn't work.. check level shifter. yep in but no out... OK find another, don't i have a level shifter on my old Xmas lights, yes, same package , yep! install it, nothing...
Ok stop think, the problem with the old Xmas lights was they ran 400khz vs 800khz, why, slew in the level shift, because transistors and caps, not an IC.. so what was that IC, wait that is actually the 3.3V VREG.. ok damn.
CPU's working, its doing the right thing. These boards can be daisy chained together, so lets run one from a Xmas light controller board.
Success! OK lets try running everything at 5V vs 3.3V so no level shifter and jumper it.
We did find out what the problem was, so more on that later, capacitance was our first issue.
Tools grow rapidly to accommodate debugging
Also we found an extra 2000 LEDs in the back office, that is just how we roll around here. 2000 WS2812Bs under a table in a box.
Next step getting the ESP8266 to work, i flipped around the connector compared to the XMAS lights i had pulled up. (in my defence there was like 10 things going on as i was doing this) and discovered i'd made the connector so the WiFi would be at the back pointing out. So that had to be desoldered ( solder sucker added to tool table) and put back, did that, still nothing. Ahh right added the extra GPIO's so put them to HighZ and we also added RESET, so made sure they were OK, still nothing , power LED is on now and the blue light is blinking. So it appeared to be working.
OK Scope out check all the connections, seems to be OK but not working and ESP has been verified. Go back to schematic, this is Sunday morning so a lot less going on, and i think wait a second isn't that the very first rev of the xmas board that had the wifi going off the edge, yep, sigh so i had gotten it the right way around, desoldered the 4x2 idc header again and reinstalled, hey look its trying to connect to the WiFi , reset loop. Voltage issues, i suspect noisy as well since the ESP hates noisy power and i'm running it off the bench, the LEDs are working mostly but wrong colours and bad data. Scope on the power supply shows it super noisy (...