Rak811 Tracker Board

I dont use GPS module (not soldered).
I removed power LED, software disabled LED1 & LED2. Now current consumption of board is:
Hold the RESET button - 39.5 mA;
Wait mode - 6.5 mA;
Transmit of packet - 68.8 mA;
Two short strange consumption after transmit - 22.2 mA.
Now drop of voltage is 70 mV per one day (18650). I would very much like to get values in microamperes in sleep mode.

I did as you wrote, but I can not put the device into sleep mode.
Are you able to share your code? Or working example, please.

With best regards.

Hrm. Does this CPU have any non-volatile storage in it?

I think it has eeprom starting at 0x08080000.
With mbed you can access it like this.

#include “mbed.h”
#include “stm32l1xx_flash.h”

FLASH_Status writeEEPROMByte(uint32_t address, uint8_t data) {
FLASH_Status status = FLASH_COMPLETE;
address = address + 0x08080000;
DATA_EEPROM_Unlock(); //Unprotect the EEPROM to allow writing
status = DATA_EEPROM_ProgramByte(address, data);
DATA_EEPROM_Lock(); // Reprotect the EEPROM
return status;
}

uint8_t readEEPROMByte(uint32_t address) {
uint8_t tmp = 0;
address = address + 0x08080000;
tmp = (__IO uint32_t)address;

return tmp;

}

Should be similar with cmsis,
#include “stm32l1xx_hal.h”

1 Like

correct, any question send email to ken.yu@rakwireless.com for coordinate.

6 Likes

Should I expect this to work with the Tracker? https://www.aliexpress.com/item/STOTON-high-gain-SMA-connector-28dB-BEI-DOU-GPS-Antenna-Ceramic-Patch-Built-in-GPS-Active/32816386209.html?spm=a2g0s.9042311.0.0.3da24c4dFq5qNG

It seems not to work.

This one works:

Look what I found on AliExpress
http://s.aliexpress.com/ZbiYBfYv

After seriously read the discussion by this thread and rak forum, we collect the suggestion and decide to optimize the rak811 tracker design as following:

  1. The R1/C1 pin of GPS power switch chip U5 needs to be connected to a pull-up circuit。
  2. Power indicator LED1 light removed. Reserve charge indicator, and two GPIO control status indicators.
  3. The EXTINT0 and ANTON pin of the GPS module needs to be connected to GPIO to facilitate wake-up in low power mode.
  4. CP2102 chip removed, add a serial adapter board accessories。
  5. Add GPIO to detect the power supply type of the device, battery-powered or USB-powered. It can be detected using the power chip PG pin and CHG pin.
  6. Directs one UART interface for users to expand to other devices。

Any more suggestions to RAK Wireless?

9 Likes

Arduino Compatible ?

Thanks Kenuy!!

Yes, that does work. I have one already. I’m looking for something with a shorter cable, so it’s not so bulky.

Hello, have you test with an another antenna ? I search an small active/passive antenna compatible RAK811 Tracker but I read about chip MAX-7Q is only with passive antenna. I’m not sure. I find only antenna with a long cable.

I’ve been getting excellent results with the passive GPS antenna that comes with the board. However, for LoRa, I have replaced the antenna with another one twice the size, giving me better antenna gain.

–Konrad

I have the same issue. Before getting a RAK811 I was playing with TheThingsUno + Ublox GPS. The RAK811 miss a lot of places that with the other setting I was able to connect. Did you solve the problem? I guess the problem comes with the LoRa antenna…

For those interested in the next gen of the RAK811 … https://www.rakwireless.com/en/beta-wistrio

I got mine RAK811-'s in two weeks ago from AliX not knowing there was a new ‘generation’ coming. But no hard feelings because my two units work excellent together with Cayenne.
It took only two days to get it working with all help I could find on the internet.

The new generation seems to be much more power efficient but be aware that it takes time to get a stable / reliable product both hardware/software. In that sense I do not have good experiences with the SODAQ tracker. I am looking forward to buy some discounted RAK811 trackers for my project.

Hi Konrad,

My experiences with antenna that comes with the board and directly attached to the SMA is indeed good.
My larger antenna with cable 4 meters and magnetic foot does not perform better on the workbench. (cable loss)
The larger antenna will only benefit when the antenne height / freedom of sight is utilized.
THe transmit and receive antenna’s must be in the same plane/direction to make use of the antenna “gain” If this is not under control than it can make it worse.
Marcel / PE1MLT

  1. The WisTrio still uses the RAK811 LoRa module
    -> Rather than upgrading the board to a newer u-blox GPS, a Quectel GPS is used
    -> A Quectel L76 GPS module replaces the u-blox Max-7Q GPS module
    -> The Quectel module has more efficient low power sleep states than the Max-7Q
    -> The command set between u-blox and Quectel GPS modules are very similar so minimal changes to the firmware in the RAK811 module were required
    -> RAK811 firmware should rather mature since this is a leveraged product
  2. A BME680 sensor was added to the board
    -> Adds a couple euros to the BOM but provides OK environmental data
    -> The BME680’s Air quality measurement is not very robust and can be easily misinterpreted.
    -> Personally, I prefer the BME280 - less power usage compared to the BME680,
    -> For both the BME280 and BME680, the temp and humidity sensors are very accurate
  3. The RAK811 Tracker board is no longer avail on their AliExpress storefront -> only 3rd party sellers on AliExpress are selling it now
  4. RakWireless appears to be putting more weight on their NBIoT products than their LoRa products and it will be interesting to see how they move forward

An specified error of max +/-1 degree C in the 0-65 degree range is very accurate? Well it depends on the application of course. The Si7021 with max +/- 0.4 degree C error over -10 to 85 degree C is more accurate.

1 Like

If you want accuracy, might as well go with the SHT31 with +/-0.2C accuracy (typical) over 0C to 85C - or the Si7021. BME280 is +/-0.5C at 25C and increases to +/-1C over 0C to 65C. For power consumption, the ranking from lowest to highest of average current consumption is SHT31 (1.7uA), Si7021 (2.3uA) and BME280 (3.7uA) where a temp/RH sample is taken every sec. In my experience in batch testing of sensors, Si7021 has a more accurate mean but larger standard deviation as compared to the BME280. In our products, the BME280 temp is LPF and typically gets better than +/-0.5C accuracy - which is acceptable in our application (higher importance is placed on RH and pressure accuracy than temp accuracy). The following link is to an interesting report but the sampling sets were small http://www.kandrsmith.org/RJS/Misc/Hygrometers/calib_many.html

1 Like

Marcel,
Definitely coax losses at 915MHz, antenna height over average terrain, and proper polarization (vertical?) are all important to getting a LoRaWAN system working.
73 DE WA4OSH, Konrad

Need to be careful with the temp and humidity calculation library you use. The BME series has at least 9 internal calibration constants which are used to calculate temperature and humidity from the raw temp and humidity reading. If the library calculations are simplified then the error can be a few percent, particularly in the RH reading. I connect the sensors to a Raspberry Pi to get an accurate reading and then a Lora Node and compare the readings.