LoRa Discovery Kit

[Tesla DeLorean]

Posted on April 05, 2017 at 22:40

LoRa Discovery Kit

STM32L072 and LoRa Discovery Kit featuring Open LoRa module by Murata using STM32L072CZY6TR MCU

http://www.st.com/en/evaluation-tools/b-l072z-lrwan1.html

Features

• Ultra-low-power STM32L0 Series MCUhttp://www.mouser.com/stm32l0, Cortex® -M0+ based with 192Kbytes of Flash memory, 20Kbytes of RAM, 6Kbytes of EEPROM, LCD, crystal-less USB, T-RNG, PCROP
• 157dB maximum link budget
• +20dBm, 100mW constant RF output versus Vsupply
• +14dBm high efficiency PA
• Programmable bit rate up to 300kbps
• High sensitivity: down to -137dBm
• Bullet-proof front end: IIP3 = -12.5dBm
• 89dB blocking immunity
• Low RX current of 10mA, 200nA register retention
• Fully integrated synthesizer with a resolution of 61Hz
• FSK, GFSK, MSK, GMSK, LoRa™ and OOK modulations
• Built-in bit synchronizer for clock recovery
• Sync word recognition
• Preamble detection
• 127dB+ dynamic range RSSI
• Built-in temperature sensor and low battery indicator 1.65V to 3.6V power supply

The AT_Slave example (creates AT modem via ST-LINK VCP) is about 41KB in size

• Attached is original shipping firmware (April 2017) Ping-Pong Demo
• Added LoraDisco_ATSlave_US915.hex a build of the example from the Cube Extensions, USA 915 MHz option selected
• Added LoraDisco_ATSlave_US915_V111.hex a build from the V1.1.1 tree
• Added LoraDisco_ATSlave_US915HYBRID_V111.hex
• a build from the V1.1.1 tree w/US915_HYBRID (17-Aug-2017)
• Added LoraDisco_ATSlave_US915HYBRID_FULL_NOT_TINY_V111.hex.zip a build using Keil native sscanf/printf w/AT+SEND fixed

[Tesla DeLorean]

Posted on April 05, 2017 at 22:45

Murata datasheet for product

http://wireless.murata.com/RFM/data/type_abz.pdf

Pin usage analysis

CMWX1ZZABZ-091 (MURATA LORA)
  1 PA12/USB_DP
  2 PA11/USB_DM
  3 GND
  4 VDD_USB
  5 VDD_MCU
  6 VDD_RF
  7 GND
  8 PB4/DBG_SX1276_DIO2
  9 PC13/DBG_SX1276_DIO3
 10 PA5/DBG_SX1276_DIO4 ?? UNUSED
 11 PA4/DBG_SX1276_DIO5 ?? UNUSED
 12 PB1/DBG_SX1276_DIO1
 13 PB4/DBG_SX1276_DIO0
 14 PB15/SPI2_MOSI
 15 PB14/SPI2_MISO
 16 PB13/SPI2_SCK
 17 PB12/SPI2_NSS
 18 PA10/USART1_RX
 19 PA9/USART1_TX
 20 PA8/MCO
 21 PA5/ADC5/DAC2
 22 PA4/ADC4/DAC1
 23 PA3/ADC3/LPUART1_RX/USART2_RX
 24 PA2/ADC2/LPUART1_TX/USART2_TX
 25 GND
 26 ANT
 27 GND
 28 PA1/DBG_CRF1 (ANT_SWITCH_RX)
 29 PC1/DBG_CRF3 (ANT_SWITCH_TX_BOOST)
 30 PC2/DBG_CRF2 (ANT_SWITCH_TX_RFO)
 31 STSAFE_NRST
 32 VREF+
 33 PA0/WKUP1/UART4_TX
 34 MCU_NRST
 35 PB8/I2C1_SCL
 36 PB9/I2C1_SDA
 37 PB2/LPTIM1_OUT
 38 PB7/LPTIM1_IN2/USART1_RX/I2C1_SDA
 39 PB6/LPTIM1_ETR/USART1_TX/I2C1_SCL
 40 PB5/LPTIM1_IN1
 41 PA13/SWDIO/LPUART1_RX
 42 PA14/SWCLK/LPUART1_TX
 43 BOOT0
 44 GND
 45 PH1/OSC_OUT
 46 PH0/OSC_IN
 47 TCXO_OUT
 48 VDD_TCXO
 49 GND
 ..
 57 GND
 
WLCSP49
  1 A1 VDD_USB
  2 A2 PA15 RASIO_NSS
  3 A3 PB3 RADIO_SCK
  4 A4 PB5 LPTIM1_IN1
  5 A5 BOOT0
  6 A6 PB9 I2C1_SDA
  7 A7 VDD
  8 B1 PA12 USB_DP RADIO_TCXO_VCC ??
  9 B2 PA14 SWCLK
 10 B3 PB4 RADIO_DIO_0
 11 B4 PB6 LPTIM1_ETR
 12 B5 PB8 I2C1_SCL
 13 B6 VDD
 14 B7 PC13 RADIO_DIO_3
 15 C1 PA10 USART1_RX
 16 C2 PA13 SWDIO
 17 C3 PB7 LPTIM1_IN2
 18 C4 PC1 RADIO_ANY_SWITCH_TX_BOOST (CRF3)
 19 C5 PC0 RADIO_RESET
 20 C6 PC14 OSC32_IN
 21 C7 PC15 OSC32_OUT
 22 D1 PA8 MSO
 23 D2 PA11 USB_DM
 24 D3 PB1 RADIO_DIO_1
 25 D4 VSS
 26 D5 NRST
 27 D6 PH0 OSC_IN
 28 D7 PH1 OSC_OUT
 29 E1 PB15 SPI2_MOSI
 30 E2 PA9 USART1_TX
 31 E3 PB2 LPTIM1_OUT
 32 E4 PA1 RADIO_ANT_SWITCH_RX (CRF1)
 33 E5 PA0 WKUP1
 34 E6 VREF+
 35 E7 PC2 RADIO_ANT_SWITCH_TX_RFO (CRF2)
 36 F1 PB14 SPI2_MISO
 37 F2 PB13 SPI2_SCK
 38 F3 PB11 ?? I2C2_SCL
 39 F4 PA7 RADIO_MOSI
 40 F5 PA4 RADIO_DIO_5
 41 F6 PA2 ADC2/LPUART1_TX
 42 F7 VDDA
 43 G1 PB12 SPI2_NSS
 44 G2 VDD
 45 G3 PB10 ?? I2C2_SDA
 46 G4 PB0 RADIO_DIO_2
 47 G5 PA6 RADIO_MISO
 48 G6 PA5 RADIO_DIO_4
 49 G7 PA3 ADC2/LPUART1_TX �?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?�?

Edit: Original post from 5-Apr-2017, fixing the pin table that the forum translation ate.

[Tesla DeLorean]

Posted on April 10, 2017 at 01:39

In order to evaluate the range I built two nodes, one which pings location every second from a GPS shield, and a second that receives that data from a fixed antenna.

[Tesla DeLorean]

Posted on April 11, 2017 at 22:17

So with a relatively low effort test (used good antennas, but didn't clamber up towers or onto the roof), I got fixes out to 3000ft (915m)

This was with the 14dBm (25mW) TX_OUTPUT_POWER setting. With a better antenna and placement on the vehicle the line of sight was 1.77miles (2850m). The next build of the software I'll try the 20dBm (100mW) setting.

The range for the Murata module is 5-20 dBm (3-100mW)

[Tesla DeLorean]

Posted on April 15, 2017 at 03:54

I've had to rewrite the USART interrupt code, thevcom_Print() function implementing the FIFO on the PingPong demo uses some brain-damaged HAL code that BLOCKS!! It blocks other interrupts until the entire pending buffer is cleared, which means other interrupts have to preempt to function, rather than the IRQ Handler just sending the one character the transmitter will take.

[Tesla DeLorean]

Posted on April 18, 2017 at 19:17

The radio interrupt also causes significant blocking and data loss on the USART side, had to ease the radio off the highest priority, allowing the quick/simple USART IRQ Handler to do it's job and leave immediately. KISS

The radio code also seems to have issues with queuing transmission or splitting large packets. The main() loop needs some attention, as the Rx blocks Tx, so when wanting to chain a number of packets for transmission it is best to check the radio is idle rather than assume Tx complete, error and timeout paths will play happily. Needs to be a method for Tx to preempt a pending Rx window.

Breaking data into 64-byte LoRa packets seems to be effective to send transient data (ie send-and-pray, new fresh data arriving daily..)

[Tesla DeLorean]

Posted on May 03, 2017 at 19:33

Mouser is back ordered into July

http://www.mouser.com/ProductDetail/STMicroelectronics/B-L072Z-LRWAN1

DigiKey also

https://www.digikey.com/product-detail/en/stmicroelectronics/B-L072Z-LRWAN1/497-17068-ND/6616000

[Tesla DeLorean]

Posted on July 16, 2017 at 15:55

Mouser is back in stock today (16-June-2017), after shipping pending back-orders they look to have 478 in stock.

[Tesla DeLorean]

Posted on August 03, 2017 at 17:22

Frustrations-de-jour

The TCXO clock require SB13 to be made, the HSE_VALUE in stm32l0xx_hal_conf.h needs to be changed to 32000000

The comments in mlm32l0xx_hw.c are wrong, it doesn't use HSE.

The HAL does not appear to correctly use RCC_HSE_BYPASS, seems to get stuck in spin-loops because it is neither RCC_HSE_ON or RCC_HSE_OFF

[Tesla DeLorean]

Posted on August 04, 2017 at 22:44

The TCXO is on all the time, code still present delaying for on and off, transitions to sleep, and back-to-back reception. Net result dead zones between Rx where Tx from another node are missed.

Think there are systemic issues with decomposition of large data blocks in smaller blocks for transmission.