Closed neilh10 closed 2 years ago
From Xbee3 LTE-M/NB-IoT data sheet
Nortons current analysis: https://en.wikipedia.org/wiki/Norton%27s_theorem Xbee pin10/GND to pin 1/BEE_VCC: Decoupled with C6/4.7uF, then close to X6220B331MR/LDO, then close to C21/10UF, + close to FDN340P to V_BATT, + short thick trace to C26/10uF + short thick trace to C27/10uF + C27/10uF which connects with longer thick trace to BQ24704, which provides 5A connection to C24/10uF and C25/10uF that connects close trace to LiPo battery connectors.
nh: it should be noted low ESR capacitors are required to supply the current pulse for all LTE modems when they turn on, and the above Xbee3 specifically identifies a 100uF/16V. The 100uF, low ESR, makes the circular current flows much simpler. Communication devices that don't perform to specification, are likely to appear to have many software problems than a single hardware one.
The details of theXC6220 seems to specifically identify its not good on the frequency response, which is what is needed for an LTE modem current pulses.
From https://www.torexsemi.com/file/xc6220/XC6220.pdf XC6220 series is a highly accurate, low noise, high speed, low dropout, and large current CMOS voltage regulator. The series consists of a voltage reference, an error amplifier, a current limiter, an inrush current prevention circuit and a phase compensation circuit plus a driver transistor. With a 0.2Ω on-resistance driver transistor integrated, the ultra low dropout voltage performance with output currents up to 1A
The XC6220 series has a built-in phase compensation circuit which means that a stable output voltage is achieved even if the IC is used with low ESR capacitors.
Notes on Use (Pg12)
Sufficient power filtering capacitors are included on the EnviroDIY LTE sim7080 cell module, so no additional caps were added to the Mayfly 1.0. But since some users might want to use the Digi brand LTE boards directly on the Mayfly, filtering cap(s) will be added to future versions. Short term tests of the Digi LTE bee directly plugged into the Mayfly with no LTEbee adapter (making appropriate changes to the solder jumpers on the back of the Mayfly 1.0) showed no problems with battery power input.
@s-hicks2, thanks for creating a new cellular modem for the EnviroDIY community.
I agree with @neilh10 that the Mayfly v1 needs to support the Digi Xbee3 LTE-M modems -- both the older ones that we are already using and also the major revision that they are currently developing -- in addition to other future modems. Therefore, I think it is important to add that capacitor chain that @neilh10 displayed above.
The Mayfly 1.1 Xbee socket has a good capacitor decoupling for supporting an Digi Xbee LTE. Many thanks
Cellular modules recommend a low ESR ceramic decoupling capacitor close to the power pins, with the lowest impedance current loop path. For Maylfly 1.0A3 there doesn't appear to be a large decoupling capcitor between BEE_VCC and GND, though perhaps the path through the regulator to the regulator's 10uF and low impedance LiIon battery wroks Can a largish ceramic capacitor (100uf) be applied to Xbee pins. I'll post references to data sheets, but I have limited access this afternoon.