It wasn’t today, but I finally got time a few weeks ago to take apart the Rii E21 I won in CK’s giveaway. The whole reason I entered the GAW was because I had had the gear to flash firmware for months, but never gotten around to using it. Since this was supposed to be an easy host for that, I signed up, and I won.
The pill was only finger tightened in, but I couldn’t get it to budge with my fingers, so put my cheap HF snap-ring pliers on it (mistake #1), and the first time I turned it, they slipped off and busted the XML LED.
I already had a couple of XPL-HIs on 16mm boards on order from Simon because I wanted to put one in one of these just for kicks.
Before I got around to it, I broke my clavicle last Wednesday laying down my bicycle on rain-slicked asphalt. So probably no soldering for 6 weeks now.
Thanks. Just based on a quick visual inspection it looks ok. But I definitely need a new helmet. The jacket, shorts, and shirt I was wearing were all ruined.
I killed the MELD-X RGBW driver that came in a light I bought from pinkpanda3310. I transplanted it into a new host with a RGBW XM-L and it worked, but unless the head was unscrewed slightly it just lit up all the dies dimly. Left it on my bedside table and the next morning my son got it and inserted the cell the wrong way, although I’m not sure he got the tailcap back on.
Now the driver just lights up all the dies dimly even when bench testing and I’m not sure whether it was a short or reverse polarity that killed it.
I broke the atrocious stock driver in my Wuben E01 when I drilled a hole through it so I could lever it out. It had terrible visible PWM and a double-tap for strobe which meant you couldn’t change modes quickly so I don’t feel guilty at all.
Unfortunately I also ripped a component and PCB trace off the cheap single-mode driver I was filing down to fit. Now I’m left with a non-functional E01, but to be honest I think that’s still a win - the stock driver really is that bad.
If anyone knows a source of tiny multi-mode boost drivers please let me know. I know MTN sells Lumintop Tool drivers, but postage from the US makes it too expensive.
i took the battery out and stupidly thought ” i wonder if its brighter with some samsung 30q’s”
so uhh that was a dumb idea as the led is only lighting up one quarter of the chip and even then its so very dim that its hard to tell its on
it guess i could replace the led (which i think was an xp50)
Maybe not so much a breakage, but a screw up for sure…
I was admiring some boards I got in from Oshpark today (am I the only one that does that?) when I realized a major flaw: I completely forgot to add the vias around the ground ring. Doh! At least the board is big enough that I could manually add a via with a drill and some copper wire.
PS - my first ever board with a crystal oscillator!
What’s the oscillator for? Why do you need a different and/or more precise signal than the MCU makes? Are you making an accurate clock for a sunset/sunrise alarm light?
Close… it’s a time-based night light for our toddler room (and for travel). Sure, I could just plug in a night light, but what’s the fun in that?
I’m repurposing a Sofirn SC31 for nightlight duties. Sunrise/sunset would be fun, but bedtimes are set by the clock and not the sun. Goal is to have it turn on each day at 8pm, then back off 10 hours later. Rinse and repeat. Other UI features single-click on/off override without affecting the timer, and click-hold for ramping up and down to select a brightness. Low voltage warning using the LEDs under the switch.
Originally, I kept the stock driver and replaced the SC31’s PIC with an attiny412 and used the internal 32 kHz oscillator to help with timekeeping (via the Periodic Interrupt Timer “PIT” peripheral). But it was slipping 13+ minutes each day. I could probably have calibrated the timer to be accurate enough, but this serves as a great excuse to try out new things!