D.I.Y. Illuminated tailcap

Oh boy! That is going to be soooo freakin cool. Thank you PD, you’re just awesome!

Awesome. I’ve got some projects which needed thinner boards… I wonder how thin Oshpark can go now. (the CNQG brass AA light has a very thin driver that I want to replace)

D’oh. I missed the thread for a few weeks and now the boots are all sold out. Most of my boots are 14mm though, so I suppose I can deal with no glow (or green glow) on a few 16mm lights.

That looks like a very nice design. Very easy to tweak to exactly the right level.

It’s definitely not plug-and-play, but at least it’s possible to measure the OTC and tweak the firmware accordingly. That option is only available for firmwares which include source code, though, so it’ll probably be trickier to get guppydrv to work.

It’s quite likely a heat issue as you said. Especially if the driver was one of Manker’s blf-a6 drivers or otherwise didn’t use nice X7R capacitors. You may be able to at least work around it a little by tapping extra fast when it’s hot… or by upgrading the driver hardware. I haven’t seen nearly as much heat sensitivity on RMM’s drivers.

Oh wow. That’s an interesting idea. I suppose you’d have to put the LEDs as close to the inner edge as possible though, to allow the waterproof seal to stay intact. The boot really needs flat surfaces on both sides. The 14mm boot I measured has only about 11mm as its inner diameter. Maybe 11.5mm. It seems pretty tight, but I’d definitely be interested in one which might fit.

The switches I checked appear to need a 7mm hole. So, that only leaves 2mm on either side between the hole and the inside of the boot.

If the attiny works, I’d imagine it probably needs to check voltage, blink appropriately, then go to sleep with a watchdog running. Repeat forever. Something like that, anyway. And pay extra care to go into the lowest-power mode possible.

BTW, on an earlier project, I discovered that the attiny can provide a high and a low power level on its output pins, and it can do this even while it’s asleep. So, it could potentially have a very dim steady glow which blinks brighter periodically. This is how the Ferrero Rocher driver works. IIRC, I measured it at 0.33mA while fully asleep or 0.36mA with the low glow, and it might be able to go lower. IIRC the high glow was more like 1.23mA.

0.8mm is their only thin option. It comes with 2oz (double thick) copper. oshpark 0.8mm board specs
Do you think 0.8mm is thin enough for your CNQG brass AA?

Actually, Simon is moving his store so he just marked everything as sold out until he’s back up and running.

I'm really curious how/if this Attiny-directed tailboard is coming off the ground :-)

Meanwhile, I would love a 14mm version of the 'dumb' board: six leds in a circle, one resistor, for use with a stock switch board.

I’m not sure how thin of a board I need. It’s hard to measure under a millimeter. :slight_smile:

Six LEDs seems perhaps a bit much, but I guess if it fits… why not? On a ‘dumb’ model I’d be happy with two.

However, if the attiny approach works, it’d be cool to light the tail different colors for different voltages. See the charge state at a glance. With RGB it could do something like…

  • > 4.0V: blue
  • 3.8V - 4.0V: cyan
  • 3.5V - 3.8V: green
  • 3.2V - 3.5V: yellow/orange
  • < 3.2V: red

Perhaps blink out messages in magenta or white. Or it could just cycle continuously through the rainbow. It might be a nice effect for the first 1-2 seconds after the main light turns off. :slight_smile:

With different colors, it might be desirable to have a different resistor (or pot?) for each color, to help balance the overall brightness of each.

Another option is to have four LEDs where they’re all lit while full, and they drop off one at a time as voltage gets lower.

(I’m assuming the attiny can use 4 pins to turn LEDs on and off, so up to 4 independent power channels)

The six leds is because the leds are now much closer under the silicon cover and will be individually visible as little lights. And six looks nice that way :-)

If it’s attiny plus six LEDs, I’d suggest three independent power channels, each connecting to opposite sides of the board. And probably an extra resistor per channel somewhere, if it can fit…

This should be do-able, but not for 14mm boots unless we use a smaller mcu. Should I use pin 3 or pin2 for the third output?

We are already using pin3 to turn on and off leds so makes sense to keep with that.

ok, that’s what I was thinking too. On Pins 5 and 6, I figured we could use pwm to set brightness. Anybody know if that’s just wishful thinking or not? That’s why I didn’t put limiting resistors on Rev6. Also, do we need the input capacitor and reverse-polarity diode in the tailcap? If not I’ll take them out. I already added easy-to-bridge pads to bypass the diode.

Tonight I’ll do a board for 14mm boots that’s a mix of Rev5 and Rev5.1, and a more complicated version of Rev6. I thought you people would be tired of new versions by now… :zipper_mouth_face:

I layout all my circuits essentially by eye, so please check to make sure I didn’t make a mistake. Routing the traces on this one literally gave me a headache.

I wouldn’t count on PWM on any of the pins, because the MCU needs to be asleep most of the time to lower the power usage. It can only do PWM while it’s awake. However, it can at least continue supplying power while it’s asleep. Mark the pin for output and set it high… or if it’s just right it can also be set low for a much lower brightness, but that might take some serious tweaking. I don’t think it’s supposed to work that way.

It probably still needs a resistor on each channel, to avoid using the MCU’s full power capacity. It can send a lot more power than desired for this sort of thing. IIRC, comfychair posted details about this in the Roche F6 thread: Roche F6 hacking

If I make code for a tailcap, I figure I’ll use the Ferrero Rocher driver for most of the development. It’s similar enough.

I don’t really know, but won’t the tail board be getting reverse voltage every time the main light is on? Or does its power get fully cut then?

The only thing which worries me on the board layout is that some traces (R2 to pin7) might be a little close to the inner circle. If it needs to be filed, those traces could be damaged. Also, comparing an attiny to my existing tail washers, it looks like it’ll barely fit if at all.

I think you may be overthinking what is happening in these tailcaps. The current flow doesn’t reverse or anything, the tailcap is just at a different point in the circuit so the cell tube is the effective ‘pos’ instead of ‘neg’. The reason the tailcap light turns off when the light is on is simply that the main switch is the path of least resistance, so all the flow bypasses the tailcap circuit.

Anyways, I ditched the diode, so anybody building these will need to be careful of polarity when they are assembling/testing.

I don’t like the clearance around the middle either, but I couldn’t see any other way. I tried to make it a tiny bit better with this one. If nobody sees any other issues, here is the Osh link.

…………………………………….

And here is the Rev5 board for 14mm tail boots. The led’s are all within 11mm radius of the center. 3 pairs, each pair set by its own resistor.

I think it looks a little like Captain America’s shield.

OP updated with new versions. We are accumulating quite the list…

Loving all the new revisions, scared to buy and try any because tomorrow you'll have a better one!

I'm doing good with the mtn17ddm/guppy drv and BlF A6 firmware and lighted tailcaps. Since I switched to 10k pots things are close to normal. After the lights get heated up, not necessarily hot just warm, or after an extended period of runtime (tailstanding on medium for 20-30 minutes) the modes do weird things until the light cools off. I gather there isn't a good fix for that yet? If I'm wrong please let me know! i believe the C1 cap is a 10uF X7R. I'm putting together a fet driven xml2 light to test and see if it has the same problem or if it's just triples.

I am not surprised, the triple that I build it on copper in an S2 and it draws very hard. I suspect that the OTC is heating up and the values read by the ADC are all over the place. I may convert it to a Qlite driver and see what happens. Hey Dog, all these revs… I want the one that will give me multiple LEDS and only replaces the washer… Which one do I want. Standard diameter, not 14mm.

Looks like I need this one…
PD68 Illuminated Switch - 19mm Rev5.1b

Dam, I do not have any 0603 resistors.

Yeah I suppose I went a bit overboard with that 5.1 board… I could just stretch out the smaller version so you can use 0805, or you could probably actually fit 0805 on the little pads if you hand-solder

You really should look at the MMU MCU, it’s super tiny, ideal for this application. I use it on my Texas Poker driver. 10mm driver board. The downside is that you have to mount it on a flash board then remove it and put it on the driver. Can’t flash it on the driver for changes.

You could add programing pads. Just pads for pogo pins if nothing else.