D.I.Y. Illuminated Tailcap - gChart Editions

The HM1160 came from AliExpress. The capacitors, resistors, and LEDs I already had on hand from past projects.

thanks gchart :smiley:

is it possible to get the rest of the conponants on aliexpress

resistors and leds and caps etc, i wouldnt mind makeing a few of these

you think they would also work on convoy c8 and s2s


Can you give the BOM for this v1.2?
Thanks gchart.


  • HM1160 - I’ve only found it on AliExpress, can sometimes also find it by it’s chip label “60AA”
  • PFET, size SC-70 / SOT-323 (such as BSS209PWH6327XTSA1)
  • 1uF 0603 or 0805 capacitor
  • A resistor of your choice (I generally use 10k - 30k Ohm) in 0603 or 0805 (for setting the brightness)
  • 4x LEDs of your choice, 0805

Don’t forget a bleeder for your driver, something around 680-840 Ohm

Oh yeah… If this is going in something bigger than a Tool (a Convoy S2 / C8 perhaps), you could use this single sided board to make things easier. It’s squared off because I’m cheap :stuck_out_tongue: (the square inches you pay are based on the smallest rectangle that can fit around your design). I have dinner of these on hand but haven’t assembled them yet. I expect them to work fine as they’re electrically the same, i just have had a moment to put one together.


Additional information regarding the HM1160 chip.

I’ve been curious about the exact cutoff voltages. And sometimes I noticed a slight difference using my power supply vs a battery. So I managed to dig up an extended datasheet in Chinese, which I ran through Google Translate. From it, I was able to pull the official cutoff specifications. They are as follows:

Volt Range VD1 VD2 VD3 VD4
3.87-4.2 Bright Bright Bright Bright
3.7-3.87 Bright Bright Bright Extinguish
3.55-3.7 Bright Bright Extinguish Extinguish
3.4-3.55 Bright Extinguish Extinguish Extinguish
3.4 or less Extinguish Extinguish Extinguish Extinguish

So… the “official” cutoff levels are: 3.87v, 3.70v, 3.55v, and 3.40v. You might be thinking, as I did, “hey, 3.40v seems really high to be the lowest cutoff! I’ve been taught that lithium-ion batteries are good down 2.8v.” And that’s not wrong. But it’s too easy to think of remaining capacity as being linear with cell voltage. But that’s not the case. HKJ has a great write up about this. Here’s an image from that article:

I used his battery comparison tables and came up with similar figures for some of my most used batteries:
||Shockli 1000mAh 14500|Samsung 35E|Samsung 30Q|Average|
|3.87 v|76|68|69|71|
|3.70 v|58|48|45|50|
|3.55 v|24|29|35|29|
|3.40 v|6|16|21|14|

Armed with this knowledge, you could say about the HM1160:

  • 4 LEDs on: 100% - 71% remaining
  • 3 LEDs on: 71% - 50% remaining
  • 2 LEDs on: 50% - 29% remaining
  • 1 LED on: 29% - 14% remaining
  • No LEDs on: under 14% remaining

Doesn’t sound so bad now, huh?

Circling back around… some of these cutoff values are different from what I initially proposed using my basic variable voltage source. I think the difference lies in that the HM1160 behaves slightly differently when a consistently-connected source is dropping in voltage vs when a source is first applied. That is, when crossing the 3.40v threshold from above by ramping the voltage supply down, the last LED wouldn’t extinguish until around 3.2 - 3.3v. But if I disconnected the tailcap, set the supply to just under 3.4v and reconnected the tailcap, the LEDs wouldn’t light up. I imagine there must be a bit of a buffer zone to keep LEDs for flopping back and forth when the measured voltage is right around a cutoff.

There ya have it… the HM1160.

Wow nice finding gchart! Thank you for providing this info. I think I’m gonna stick with v1.0 first, it’s difficult to find the PFET here…
Regarding to the small opening for switch, can I enlarge the center hole on the board or should I sanding the switch?

I ordered two batches of v1.0 and one was too tight and one for just fine. Manufacturing tolerances I guess. I was able to carefully drill the boards out just fine. 7mm I think, you’d want to measure.

Nice practical reference charts on practical battery capacity - I guess gChart stand for “Great Charts”.

I was a big fan of how your HM1160 tailcap performed on the video. But it’s understandable, there’s a difference in how the circuit behaves using a stepped down DC power supply versus actual batteries. I guess the ideal chip was too good to be true :frowning:

Personally I think the 3&4 LED cut-offs ranges are fine but the 1&2 cut-off ranges are a little too early. The earlier video showing cut-off/extinguishing of all 4- LEDs at 2.9-3V was ideal because it simulates Protected Battery cut-offs. Cutting-off at 3.4V even if there’s only 6-21% remaining is a little early (1 LED was ideal.)

Just a thought - would it be possible to slap a capacitor into the circuit somewhere to help get the LED excited enough to light up on battery power, thus getting closer to it’s performance as on the video using a DC PS?

Hi.Child can you buy the whole lot? PCB and components together.Thank you.

The observed performance difference on the power supply was actually caused by the difference in rising edge detect (flashlight turning off, tailcap turning on) versus falling edge detection (tailcap on, battery running down). The spec’ed cutoffs for rising and falling can be seen at the end of this translated datasheet.

Unfortunately a capacitor cannot change the system voltage, it will only mimic the battery voltage. And there’s already a capacitor in place.

In short, the HM1160 is unfortunately a “take it or leave it” scenario. It’s handy, but not flexible. For flexibility, you’d need something like one of my smart tailcaps which is running an MCU and can be programmed with any voltage cutoffs as desired. Or… I’m currently finalizing a tailcap that is similar to LoneOcean’s tailcap (in concept and functionality, not design). It’ll use 4 red/blue bi-color LEDs and a couple low-power voltage reset monitors to switch from blue to red to off at whichever cutoffs are desired (in 0.1v increments). I hope to have it ready in the next few weeks.

Yes, as long as you’re comfortable installing the tailcap and ensuring that your driver has an appropriate bleeder resistor. And with the understanding that the HM1160 provides preset cutoff voltages and those cannot be changed. I ask $3 per assembled board. Domestic shipping for $1. International depends… I’m legally required to use package shipping which is usually $13, but I have snuck these through in standard envelopes (at your own risk) for like $2.

What is the purpose of the capacitor in the circuit?

Practically all IC’s like to have capacitors around them to help smooth out any small voltage fluctuations. You could probably get by without, but they’re so cheap, you might as well follow the manufacturer’s recommendations (as is done in this circuit, which is taken right from the datasheet).

I see, thanks. :slight_smile:

I don’t own the light but I’m curious….would it be feasible to do a smart tailcap for FW3A?

I think this one has been asked before. Anyway, I don’t think it’s feasible. The issue is pretty fundamental to the way the tailcaps work.
When a normal, power-interrupting clicky switch is off, the lighted tail is an alternate current path. The full loop is Battery Positive across a bleeder somehow to ground, through the lighted tail, to battery negative.
With the FW3A, there’s no power-interrupting switch. So no gap to bridge with the tail there. Could you make a lighted tailswitch board? Not really that either, because you’d need an additional (a third) current path to the switch to power the LEDs.

I don’t think a smart tailcap is possible for the FW3A without a lot of effort. You need a good constant power source to measure.

The main electric path in the tailcap (from what I can tell, that’s the negative end of the battery to the spring to the rim of the PCB to the main body tube) in uninterrupted. That means that the path of least resistance will always be that direct path, so you can’t put a smart tailcap in parallel - no current will flow through it.

The second electric path is possible where the e-switch connects. I imagine the e-switch is connected to a pin of the MCU that is set to “input pull-up” with means it’ll be positive with a ~30k Ohm internal resistor connected. Pushing the e-switch connects it to ground momentarily, which the MCU recognizes. I have before (in the D25 headlamp) connected an LED in parallel to the e-switch. That allows for a “dumb” illuminated e-switch. It didn’t work with green LEDs, but worked with blue. I’m thinking because of vF differences. I’d have to do some testing to see if that’s a viable option for an illuminated tailcap.

Doing a “smart” aux board under the TIR (like what Lexel has done) would be much more feasible and predictable.

Edit: Scallywag beat me to the punch, although I included a few different details. Thanks for chiming in (and confirming my initial thoughts).

My first FW3A-series light should be any day now (I think it just passed through customs) so I can poke around more later, but I don’t expect any surprises.

Cool - didn’t know that could work! I was assuming you’d have to grab yet a third current path down there to power the LEDs (ground, of course, being quite freely available on that end of things)

Well… without more testing, I don’t know how dependable it is. I ended up getting rid of my D25 so I can’t test on that one. The different behavior depending on the LED color makes me cautious about this and how it would operate long-term and with different battery voltages.