💬 EFEKTA nRF5-C Motion sensor

  • Great work, this looks interesting.

    How did you solder the nrf5? Can you also put in a schematic in PDF?

  • @omemanti Thank you for your support and interest in this project. You also have great projects on nrf5. Soldering is very simple, now I can say it for sure as there is experience. A little flux, a little solder paste, this mixture is applied to the place for the chip on the PCB. Install the chip, clamp with tweezers-clip, adjust the position. Heat with a blow dryer. After soldering with a blow dryer, apply flux on four sides, conduct soldering iron on each side. It's easier than soldering mega328.
    As for the scheme, it will of course be added, as well as the sketch code, the project is open. But give me time..

  • Hardware Contributor

    another nice board 😉
    but I see same "little" routing mistake as your other motion board.
    if it's your 1st rf boards, I would recommend to keep nordic layout the same (they provides gerbers too), maybe read some docs on rf design, to get best range, against detuning and interference with some others devices on different 2.4ghz protocols (even if ceramic antenna and meandered antenna are more prone to detuning once enclosed etc).
    It could be interesting if you also provides pi network footprints close to antenna, in case your customers have tools and would like to tune their board and be closer to CE/FCC.. pi network parts are usually important for ceramic antennas as you can't tune them by tweaking their length, so it relies on additional matching network. whereas pcb antenna can be tuned by tweaking its length and/or using additional parts.

  • @scalz Of course I will continue to do tests, especially radio frequency tests. While I lean on the built-in balun(see Key features - On-chip balun (single-ended RF)). This is my first experience of developing a device without a ready module. I study RF design docs ..The length of the printed antenna track was calculated, it is not hand-drawn. While everything suits me, I am happy with the result, but, of course, I understand that there is always something to improve, especially in the topic of radio frequencies. And I hope that my research will give impetus to other developers to our theme MySensors and similar devices with better RF design will be more and more.

  • Hardware Contributor

    results can vary from environments (on desktop or in real environment, presence of others 2.4g protocols, less range with others nrf24/nrf5 tuned differently etc). easier to check with a VNA (a true one). but if it's only for your personal use and range is enough for you, great.

    if you take a look in nordic's forum you will get explanations how parts should be routed close to the antenna pin (like the pic I showed you from datasheet). even if onchip balun.. (it won't prevent harmonics)
    I meant width of the rf feedline trace. you don't care about the length (what calc did you do for length?). well, actually, you must care, when long enough, you definitely need the right route width, for the impedance matching to 50ohms.
    From the matching network(inductor+capacitor which should be as close as possible to the ic on your board..) to the antenna or to pi network, for a 1.6mm 2layers pcb and 10mils isolation, calc for route width should be approx 60mils..

    about bom, imho, I prefer to order crystals, 3.9nh inductor and small cog.np0 0402 capacitor from known sources, so I'm sure they are not out of specs. especially the crystals for reliable freq&communication

  • @berkseo great, I'm going to try it soon, I have 5 nrf52832 chips laying around (got them with the nrf52832-DK). Would be a shame if I didn't use them 🙂