News and research

Blog

Year two recap

30 December 2024 <- Previous

Input Labs went live in late November 2022, since this date is just a few weeks from the end of the year, this time we are making a double report Input Labs year 2 recap and Year 2024 recap.

Things that went nicely

4 firmware (and Ctrl app) updates

We released a total of 4 firmware updates and 4 Ctrl app updates (they usually go together), some of the new features include:

16-bit mouse values.
Configurable thumbstick anti-deadzone (to compensate for game's hardcoded deadzone).
Double press mappings.
Home button can be used for Steam Deck shortcuts.
New algorithm with much better touch sensor detection.
Unlocked 4 custom profile slots in the Ctrl app for personalized profiles.
Option to load / save / copy / reset profiles.
Ctrl app advanced settings (gyro calibration tweaks, swap gyros, longer calibration).
Ctrl app additional logs (auto-diagnose, touch sensor logs).

Migration from Blender to Build123d

We finally decided on a better CAD tool to replace Blender, that tool is build123d, a Python library to declare CAD as code and uses OpenCASCADE geometry engine internally (the same than FreeCAD uses).

As for today we already ported:

DPad, Select and ABXY buttons.
L1/R1 triggers.
Scrollwheel assemble.
Battery cover.

Alpakka kits in the shop

During summer we introduced the concept of Complete DIY kits in the shop, these are configurators that help users adding all the relevant parts into the cart, depending on how they want to approach their DIY project. This reduced the guesswork, and was well received by the community.

Community

Input Labs' most awesome success is our community, always willing to help each other, sharing mods, cool designs, profiles, knowledge and fun. We are now 1200+ members in Discord, and the community published a total of 43 mods. Keep being wonderful, folks 🤍.

Patents

Since we want to be fully transparent with the community, we explained our strategy regarding patents. In a nutshell, we submitted two of our innovations, in both European and US patent offices.

Many wireless prototype iterations

We created a total of:

10 module prototypes for wireless communication and power management (later consolidated as Marmota module).
2 Alpakka carrier board prototypes compatible with such modules.
4 breakout boards to perform hardware integration tests.

Alpakka 1.0 announcement and pre-order

All these efforts on the wireless development eventually crystallized into the the Alpakka 1.0 reference design. Including the new modular approach with the Marmota module.

Things that could have gone better

The development of the wireless Alpakka, in which we hit many research dead-ends, and it consumed most of our time during the whole year.

In early 2024 we put a lot of effort trying to make the Raspberry Pico W to meet our requirements, the upgrade made a lot of sense since the Alpakka v0 uses the normal (non-W) Pico. The first laboratory-condition tests were very promising, but later end-to-end testing revealed serious issues in both in the BT and WLAN approaches. After submitting bug reports to the appropriate channels, we could do nothing else but abandoning the idea of using the Pico W.

In May 2024 we started experimenting with our own antenna solution, using a Nordic NRF-24 as transceiver and a chip antenna. Again, the preliminary tests were very promising, but later designs with more complex PCBs had diminished antenna performance. We put a lot of effort trying to fix it and failing, and even with the support from experts in our community we were not able to make it work, and gave up on the chip antenna approach.

In retrospective, we should have abandoned the chip antenna approach much earlier, but at least not all the R&D was wasted, as we learnt how to create our own Raspberry RP2040 board with battery power management, which we can use in further developments.

In October 2024 we started testing a Espressif ESP8684-MINI-1 as a communication module (while still using the RP2040 as main MCU). This is a somewhat convoluted solution, but still better than the alternatives. End-to-end prototypes were very satisfactory, and we were confident that this approach is the best one to move forward.

Looking back, is easy to see why it took us so much time to figure out the wireless Alpakka, we didn't develop just one solution, but almost three, it just happened that the first two failed in the later stage of development. On the bright side, we learnt a lot, we came stronger, and we finally have a solid solution we are happy with.

Things we hope to achieve in the 3rd year of Input Labs / 2025

Primary goals:

Complete the Alpakka 1.0 reference design, and start shipping pre-orders.
Put together the Kapybara workgroup, and start working on Kapybara prototypes.

Secondary goals:

Keep updating the firmware and Ctrl app with some of the most requested features.
Finalize the port of 3D models into Build123d.
Secure additional funding via licensing or external investment.
Increase presence in social media and journalism outlets.

Closing notes

Thanks to everyone supporting this project, the dream of an open source controller ecosystem is possible thanks to you!.

Have a nice new year and solstice celebrations! 🤍

- Michael and Marcos