RAK8211NB - no response from NB-IoT Modem

Posted on
  • Hi there,

    I am trying to send my first AT commands towards the BC95 NB-IoT module using one of the examples. For some reason I only receive Error codes.

    Checked the board.js file (Tx D28 and Rx D29), as part of the script "usart" is being used.
    Loaded the latest firmware version 2v03.

    Any hints?

  • Have you powered the board from a battery? If you just use RAK's dev shield then it only powers 3.3v, which means the modem never gets any power so never powers on

  • Hi,
    I am afraid that I have used indeed only the USB connector of the dev shield (sorry, forgot to mention this). Now you are mentioning this, I remember something into that direction.

    Question, I have both a solar panel and a battery. P2 is the standard connector for the two-wired solar panel connector (~ 5.5V). The battery (~3.5V) is also two-wired (same connector)?

    Based on the required power supply of 3.5 - 18V, I guess the battery can be put on P2 as well unless P2 does not like solar light based electrons :-)

    Thanks Gordon.

  • @etmpvga, to handle the variable output of a solar panel (assuming with no controller or a like) and the need for a steady supply of the mc, you need something in between the panel and the battery... and depending on the device you use for that, it may even have an regulated output on it, 3.3 or 5V. Don't know what mc board you have - whether it has a regulator or not...

    You need to make sure that your battery is not over charged.

    For once cell 1S / 18650 Lithium and solar 6V solar panel and the need of 5V, something like a 3.7V Lithium Li-ion Battery Charger Module Step Up Converter (google/ebay) works.

  • I'm pretty sure the 2 wire connector is for a solar panel, and 3 wire is for LiPo (with a thermistor on the 3rd wire?).

    There's a proper LiPo charge chip on board, so I believe connecting a solar cell is fine.

    You'd have to ask RAK though or look at the circuit diagram as that's only from memory :)

  • Tried to find a circuit diagram... all I found is some interface in documentation... Reading the details about the 2 wire connector P2 for a solar panel and the 3 wire for battery (LiPo - expected) - especially the 3rd pin on the battery connector P3, which say Charge indicator, and all the suggested applications and available tutorials, I guess it is safe to assume that powering works as expected: a Solar panel up to 12V nominal (without load producing something between 17..18V) and and LiPo battery on the battery connector.

    You can verify/this as follows with a 5..9V power source, a 220R resistor, 470R 1/4W+potentiometer (pot) and a Volt - or multi - meter:

    • connect (for now) only GND of a 5V (9V) source (USB / battery) to ground of solar panel connector
    • connect + of battery connector and a wiper of pot with a resistor
    • set wiper of pot to 1/2 (1/3)
    • connect one side (side closer to wiper position) of pot to Ground of Battery
    • connect other side of pot (side further away from wiper position) to + of solar panel connector
    • connect volt meter to + of battery connecter and battery Ground
    • verify above connections
    • connect + of 5V (9V) of power source to + of solar panel connector

    A) The volt meter should something below 4 Volts.

    b) Measure and write down voltages on Charge indicator pin of battery connector against Ground AND then 3.3V / VCC33 of board (VDD_nRF of P1 or VCC33 of P5). If you see no typical 0 or 3.x Volts value, you may need to connect a 10K..47K..100K resistor across the multi meter probes (Charge indicator is marked as an 'O'(output) type of pin... but it all depends on the board's circuitry whether a load is needed to see actual 'output' - it can be logical output, but physical input, like open-collector/drain).

    c) Put volt meter back on battery connector pins.

    d) Turning the pot wiper slowly towards the side connected to the + of the power source should show increasing value on the volt meter.

    e) When at 4Volts, repeat b) - and you should see same values

    f) Put volt meter back on battery connector pins.

    e) Turning the pot wiper slowly even more towards the side connected to the + of the power source should show increasing value on the volt meter beyond 4Volts BUT STOP at 5Volts

    f) Now repeat b) and you should see at least one of the values having 'flipped', like from 0 to 3.x or 3.x t0 0 volts.

    g) Keep the volt meter connected to the Charge indicator and power source the way you notice the flipping of the values.

    h) Turn the pot wiper slowly back - towards the side connected to Ground. At one point you see the value flipping again back to about what you measured in step b).

    The flipping of the value at the Charge indicator and the voltage at the + at the battery pin tells you when charging of the battery stops or begins - is off or on. The flipping may happen at different voltage depending if the input moves from lower to higher or from higher to lower voltage. The flipping point when going upwards in voltage is called cut-off point, and for a single LiPo battery should be at 4.20 +-0.05 to protect the battery from overcharging (some batteries have one built in, but it is still good to stop charging).

    If you have another multi meter, switch it to mA range (0..200...1000) and put it in series with the resistor between pot wiper and + pin of the battery connector. You will then see a change of current flow from neg to positive (value) and vice versa, or from flow (a value) to no flow (0 or practically 0).

    Using a digital multi meter may be a little bit tricky... because if the flowing currents pulse, you see values jumping all over the place. Using an analog (old magnetic technology) multi meter balances / middles the values out (but showing lower values). You can 'fake' that balancing / middling out (to a certain extent) by putting a capacitor between the digital multi meter probes and measure thru a resistor and (schottky) diode in series. After a while the capacitor will be charged close enough to the actual peak value of the voltage showing at the measurement point. Taking into account of the measured the forward bias (voltage drop) of the (Schottky) diode, gives enough accuracy (
    +-0.5..0.1V) . The reason of taking a Schottky vs. a normal diode is that the Schottky diode has a much lower forward bias vs. a normal (Si) diode:02..0.3V vs 0.6..0.7V and is faster in switching between passing and blocking of current flow and you can measure values thru a Schottky down to about 0.3V.

  • Hi, thanks for all input.

    Tried tonight all suggestions, without any luck :-(.
    Used solar cell which produces around 6.0V in combination with the USB.

    Found the possible cause of the problem, check the voltage of pins 45/46 of the BC95B module, which was about 2.7V. While the specifications ask for more....

    Checked also the 8211.json file, this one is referring to the M35 module. Not sure if this might be the problem, the pin definitions of the NRF52 seems to be ok...not sure at the BC95B module.

    Any final ideas are welcome.. :-(

  • And a final mistery, trying to hook on a batery...looking at the pin settings of P2 (from left to right), on PCB you see + -, while P3 (from left to right) shows - T +. And to make it even more confusing , the RAK documentation shows + T - (the other way around....).

    Measuring the pins by checking the resistance, it seems the most right pins of P2 and P3 seem to be connected...

  • Find another ground, for exaple antenna or a polarized capacitor on the neg side and measure against the ground (-) of the battery and cell pins and you should get an instant 0 ohm and it will stay 0. So much for figuring out the ground (-) pins.

  • Hi all, received battery and attached it to P3. GPS is working now. NBIOT module is not working. Looking at the electrical scheme, both should and are powered. Measuring the NBIOT BC95B power, I have around 3.7V.

    Not sure this is sufficient or is there another issue. Checking the same electrical scheme and comparing this with the JSON file for RAK8211, it looks like pins 14 and 15 are swapped. Not sure if this is ok.

    Any last thoughts?!

  • The rx and tx pins are swapped for the bc95 vs the m35 module.

    I'd suggest you update the firmware on the bc95 as well. This presents a challenge in that you need to solder some wires to the bc95 to a usb->serial converter. I use the usb base board for this. There's a windows program from quectel that does the firmware upload. You can also do this 'over the air', but I've not tried this.

  • Post a reply
    • Bold
    • Italics
    • Link
    • Image
    • List
    • Quote
    • code
    • Preview

RAK8211NB - no response from NB-IoT Modem

Posted by Avatar for etmpvga @etmpvga