Design Analysis of a Petzl ACTIK CORE Headlamp

Hi! It’s been a while. :-)

Since my last post here five years ago I’ve been around a fair amount, played with new toys, built few things. One thing that recently impressed me though is this little mid range Petzl headlamp.

I’ve been using it for trail running in the night for the past two years, including some fairly long adventures. As an embedded enthusiast I got more interested in the design of the little thing, so I took it apart to study it, and found many non obvious things. Let’s try to highlight some of them here!

Main body

The body of the headlamp is the classic Petzl TIKKA. Apparently Petzl is selling lamps based on this design since the 2000s with a few smaller variations on the body and strap design (this one is fairly basic) on higher end models.

It’s a compact and lightweight design and there’s a 30 degree angle between the back that sits flat on the forehead and the lenses plane, which makes it look a bit goofy in my opinion but also seems to place the beam just about straight on line of sight on the highest position. Well done Petzl!

The body tilts downwards on two pivot points on the bottom, and has a ratcheting built into it that gives five positions, pointing almost vertically down on the lowest one.

The electronics sits on a middle piece screwed to the main body and everything is locked by the battery compartment, the lenses are on a separate single mold.

To state the obvious, there are three LEDs on this one, two high power white ones and a low power red one.

Power source

The original TIKKA was running on 3 AAA batteries, which largely dictates the size of the whole unit. Current models run on rechargeable lithium batteries, and the battery is a separate unit that includes both protection and the charging circuit via micro-USB.

I think that the reason why the charging circuit is not part of the main lamp electronic is both to simplify construction (one less port to open and protect from dust/water ingress) and to avoid problems with mixed chemistry charging (preventing the charger from trying to charge non rechargeable batteries). That said, this setup works great in practice, it’s easy to charge batteries in parallel, and it gives the option to run the lamp on non rechargeable batteries for those long over-night events.

The battery compartment is not sealed, but the main electronic part is. In any case I’ve used this under pouring rain plenty of times and never noticed any water ingress, just don’t submerge it.

Power output

The power control for the ACTIK is trivial, there are three MOSFET (the ones marked G31) to control each of the three LEDs. There’s a pair of limiting resistors for the two high power 3535 white LEDs, and apparently a separate one for the red one.

The white LEDs are PWM controlled and have three (white) power levels: a low power mode on a single LED, a medium power (guessing 50% PWM) and a full power one (100% PWM).

The light power drops with the battery voltage following the discharge curve of the Lithium cell. In practice, I found it to be noticeable only when the battery is almost completely out of charge and about to switch to medium power.

Anyway, these are the power numbers from the light at various charge levels, full power mode:

  • Full charge: 4.2V 1.2A 5W
  • Half charge: 3.7V 0.8A 3W
  • Almost flat: 3V .27A 0.8W

Which makes sense with a ~3V drop on the LEDs and 0.7 Ohm on the limiting side. There’s no heatsink and the LEDs seem to dissipate on the battery assembly.

Petzl offers a constant power mode on more high end models (Reactik and Nao), this is a photo from one of those (thanks FCC):

The way that seems to work (guessing from the layout) is with a pair of transistors (the SOT marked 053) driven linearly by a dual op-amp (the ST QFN16 part next to them) for constant current feedback. In practice, it’s like replacing the fixed current limiting resistors with two variable resistors.

The constant current design is neat and I would love to see it on lower end lamps. In Petzl’s defense though that circuit seems to take quite some space on the board, and I haven’t been able to find any existing high power constant current sink on the market.

That said, both designs are horribly inefficient. I realize that’s not much space for a 1A DC/DC power converter, but that’s what I’d love to see on top end models to squeeze a bit more time out of the battery.

The lens system

This is what really, really intrigued me on the design of this lamp.

At a first approximation, it’s a headlamp with two LEDs. Two is better than one, period. Twice the LEDs, twice the power, half the battery life, easy, next paragraph.

Well, actually there seems to be quite a bit more going on there. The beam produced by the lamp is quite special, and the optical magic that seems to be behind this appears to be a clever combination of three (!!) TIR lenses design squeezed in two optics.

First, I don’t know anything about any of this, but these guys at LEDiL do, and wrote a nice guide about TIR lenses applications, so go read that to see the light. (haha, you see what I did here?)

Now let’s see what we have here, starting from the middle light, the one above the ACTIK text:

The lens is an horizontal oval. It produces a fairly uniform wide beam on the horizontal axis, by means of those vertical grooves on the lens surface. The beam looks a little like this:

In practice, it throws some light power on the left and right side of the lens axis.

Now the left one is the really interesting one, as there’s a bit going on here:

The construction of the lens seems to be this: roughly two thirds of it is flat polished, the remaining third is horizontally grooved, but the features are not a symmetric triangle, and more like a sawtooth. This is the resulting beam:

What’s going on here is that the flat part of the lens gives a focused circular beam straight on the lens axis, taking two thirds of the light power, while the horizontal groove is creating a vertical oval with the rest of the light, but the groove design is offsetting the beam towards the bottom.

The resulting beam of the two lights is the following:

Which we can break down in the three sub-beams in this horrible picture:

  • The horizontal oval in blue
  • The spot light in red
  • The vertical oval in green

What does it mean in practice? Looking down the trail in the middle of a forest, the spot beam gives a nice deep point of vision to see where we are going, the vertical beam is showing us the nearby rock that we are about to trip on, while the horizontal beam gives some peripheral vision which helps getting the sense of speed, and also seeing the bear that is about to start chasing us.

To achieve the same with just a wide angle lens, we would have to waste a significant amount of power throwing light where we don’t need it: in the sky, and on the corner of our field of vision.

I love this beam shaped design! Good job Petzl!

Electronics and assembly

The electronics on this is maybe the least interesting part. The LEDs are on a usual MOSFET + resistor setup, the micro running the show seems to be a PIC (12F1572 3x3mm MSOP?) with its own linear regulator. The remaining IC may be a power supervisor or some other support stuff.

The interesting thing to note about the construction is the flat cable connecting the two boards. To maximize the battery space and keep the unit compact, the two boards are mounted on different planes: the lower one is flat and the LED carrier is angled. The way Petzl went for manufacturing though, was to make a single board (single design, single BOM, single assembly line), connect the boards with a flexible cable, and integrating point to snap the two boards apart for final assembly. Clever bit of design for mass production!

That’s it, maybe way too much enthusiasm for a headlamp, but I’ve learned a lot out of it, and felt like sharing. Thanks for reading!


4 Responses to Design Analysis of a Petzl ACTIK CORE Headlamp

  1. wattnotions says:

    Hey man, this was an illuminating read :D The new one you made sounds cool aswell. If you get around to blogging about it I will be interested to see what you did.

    • Hey, glad you liked the post. Yeah these past few months have been a bit of an obsession and I’ve learned a lot about optics design, batteries, fuel gauges but also stuff wasn’t expecting like how the human vision reacts to various light levels and beam shapes.

      One thing I should probably correct on the article above are my comments about power conversion efficiency and power numbers: the measurements I posted here are taken on the bench with a lab power supply and for some reason I ignored the internal resistance of the lithium cell. Well as it turns out was significant, probably because the current demand at full brightness is pretty high compared to the battery capacity. What this means is that on a full battery at full brightness, the battery immediately drops to around 3.6V, which is really close to the LED forward voltage. With a 0.8A consumption and a 0.7 Ohm drop resistor that should about 85% efficiency, which I reckon is very hard to beat with a switching converter on such a low drop.

      And another effect of the internal battery resistance is that since the current drop as it discharges, the voltage drop curve gets slightly flattened, so the light power drop flattens as well and you mostly notice it in the end.

      So, once again, as a friend of mine pointed out to me: all this work to end up with the same design of a product I already have. :-)

      Anyway, the new project… the 25W four beam dual cell remote battery monster I designed has been fun but I never got around to build a box for it and it tend to rain pretty often around here. :-) Plus there were few things I wanted to change and never got around to respin it.

      What I’ve been working on recently though is a replacement control board for the ACTIK to run it with the firmware I did for the one designed from scratch. That adds a better control interface for what I do (how the lamp reacts when you click the button), remote battery level monitoring, remote control (from a Garmin watch via ANT+). If that works out as I hope, it would be a really cool project to publish, but if you are curious about it, the design is already on my github page:

  2. wattnotions says:

    Hey, I missed the notification that you replied, apologies. Ah so there are two seperate projects? I thought you made a replacement board for the actik already but you actually made a totally seperate design and now you are working on a replacement board for the ACTIK pcb too. You have gone deep down the headlamp rabbit hole my friend.

    Interesting point about the internal resistance of the lithium cell affecting the efficiency calculation.

    There is a niche area of interest for consumer products that have open source replacement PCBs for them. Let’s say for example I wanted to do a project involving a headlamp, instead of having to start from scratch I could just come along, by an ACTIK headlamp, get your board made up and flash it with the firmware from your github etc. and jump right in. If I wanted to add something to the PCB it would just be a case of modifying your eagle design files or whatever.

    I think it would actually be a cool thing to have, even the stuff you mentioned, remote battery monitoring and ANT+ etc are some examples of interesting things that could be added. Now I am thinking about headlamp projects….

    Linking it with a garmin watch opens up even more possibilities especially if the watch had gps. You could track your speed and have the lamp flash if you start running too slowly or something. Or better yet have the lamp switch off completely if you don’t run fast enough :D That would put a spring in your step !

    • Yeah it became a bit of a sickness for last year or so, it started with a design from scratch mostly to play out with power converter and optics… 4 beams, dual cell for the power, never got around to do a box, looked something like this: Unreasonably powerful, good for experiment, not very practical but worked great to study the beams and learning how much power you actually need and what controls would you like. Then I’ve redone everything again in a bike helmet format (, mostly out of necessity, and ultimately I’ve figured I like the ACTIK pattern and weight, but I want the telemetry and user interface from mine, so I’ve redone the electronics. This is how it turned out:

      Pretty pleased with the integration, the electronics fits in place of the old one and the cavity is sealed.

      For this project reusing a commercial box is probably the best I could get, it would be really hard to match the level of integration of this one, with optics and battery compartment in a single molded piece.

      The watch integration opens plenty of options, normally when we go for group runs we take one or two breaks where we stop the watch and shut the light off to not blind each other, with this I have the lamp going off automatically when I stop the watch. Then for long runs when you want to maximize the battery life I can have it switching power level depending on whether I’m climbing or going downhill.

      It’d be great to have Petzl making something like that with the code available so one could fully customize it, the electronic in here is actually very similar to what they had in the Reactik+, so I guess modifying that could be an option too.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: