Reading out data through python not working [installation problems]

Hi,

The ADXL313z is supported in different ecosystems - the plug-and-play, evaluation-only, closed source option using the SDP-B as a controller, and the open source, highly flexible, suitable for evaluation and prototyping NoOS and iio ecosystems.

To use the board with the SDP-B you will need to download the application specific GUI software from an FTP site - ftp://ftp.analog.com/pub/iMEMS_Sensor_Eval/EVAL-ADXLx-SDP%20Eval%20Platform/

See here for more info - https://wiki.analog.com/resources/eval/user-guides/inertial-mems/accelerometers/adxl313/eval-adxl313-sdp, https://wiki.analog.com/resources/eval/user-guides/inertial-mems/accelerometers/adxl313

The SDP-B API is not made available to the public. You will not be able to write your own scripts to automate it.

I do not support the ADXL313z (you can try this forum MEMS Inertial Sensors - EngineerZone (analog.com) if you have more questions on the ADXL313z, or if there is a problem downloading that GUI), but if you want to proceed with Python from Windows/Linux you will need to use a compatible microcontroller ( for example STM32 Nucleo, uri will most likely be a serial COM port presented by the ADXL313 firmware for the STM32)  
https://wiki.analog.com/resources/eval/user-guides/eval-adxl313z/no-os-setup
https://github.com/analogdevicesinc/no-OS/tree/main/projects/eval-adxl313z/src/examples/basic

https://github.com/analogdevicesinc/no-OS/tree/main/projects/eval-adxl313z/src/examples/iio_example

It should also be possible to use Python on a Linux sbc with GPIOs, such as a Raspberry Pi, to communicate directly with the ADXL without a microcontroller - but it would involve more research on how to set it up in your environment. https://github.com/analogdevicesinc/linux/blob/main/drivers/iio/accel/adxl345_i2c.c
I have not used an IIO device natively on the Raspberry Pi - The uri needed may "ip:localhost". This page is a good introduction to using iio on Raspberry Pi - https://wiki.analog.com/university/labs/software/iio_intro_toolbox

Other forums exist specifically for Microcontroller NoOS and Linux software queries - Microcontroller no-OS Drivers - EngineerZone (analog.com) and Linux Software Drivers - EngineerZone (analog.com)

So in summary, there is a Windows GUI for a quick plug-and-play look at the accelerometer's performance using the SDP-B, or there are different resources for more flexible and custom control of the board - with a learning curve, and requiring other hardware.

Hi, thanks for the quick reply

I did figure out how to access that data through that ftp GUI software (which is already very nice). Unfortunately i do need to access the raw accelerometer data for processing and in my case some machine learning.

The other options you provide then need direct pin-connection to the ADX313 accelerometer board (and in this way not use the sdp interface board or eval cb1z board)? Is my understanding correct?
To avoid a next mispurchase, is there a list of boards where the accelerometer data is easily accessible by (ideally) python or arduino? Or at least boards where the guides are a little more clear :) ?

many thanks 

bdck

Hi,

"The other options you provide then need direct pin-connection to the ADX313 accelerometer board (and in this way not use the sdp interface board or eval cb1z board)? Is my understanding correct?" I would encourage you to verify this in the accelerometer forum or the linux or NoOs forum. But yes, for direct access to raw values you will need to put the SDP-B to one side and you will most likely need to wire directly from the microcontroller/RaspberryPi to your ADXL board.

"To avoid a next mispurchase, is there a list of boards where the accelerometer data is easily accessible by (ideally) python or arduino? Or at least boards where the guides are a little more clear :) ?" I understand your frustration/disappointment here. Honestly, I haven't used an ADXL from Python. Do you have a Raspberry Pi? If you do, the most straightforward option may be to burn the Kuiper Linux image to an SD card, jumper wire to your board and try to access from Python. I'm making an assumption here that the linux drivers for the ADXL devices and other iio tools are pre-installed in Kuiper.
Using pyadi-iio in python, you should be able to run this example - https://github.com/analogdevicesinc/pyadi-iio/blob/main/examples/adxl313_example.py
You will need to figure out the uri to pass to the example.

I would encourage you to double check this advice in the Linux software forum!

Hi,

"The other options you provide then need direct pin-connection to the ADX313 accelerometer board (and in this way not use the sdp interface board or eval cb1z board)? Is my understanding correct?" I would encourage you to verify this in the accelerometer forum or the linux or NoOs forum. But yes, for direct access to raw values you will need to put the SDP-B to one side and you will most likely need to wire directly from the microcontroller/RaspberryPi to your ADXL board.

"To avoid a next mispurchase, is there a list of boards where the accelerometer data is easily accessible by (ideally) python or arduino? Or at least boards where the guides are a little more clear :) ?" I understand your frustration/disappointment here. Honestly, I haven't used an ADXL from Python. Do you have a Raspberry Pi? If you do, the most straightforward option may be to burn the Kuiper Linux image to an SD card, jumper wire to your board and try to access from Python. I'm making an assumption here that the linux drivers for the ADXL devices and other iio tools are pre-installed in Kuiper.
Using pyadi-iio in python, you should be able to run this example - https://github.com/analogdevicesinc/pyadi-iio/blob/main/examples/adxl313_example.py
You will need to figure out the uri to pass to the example.

I would encourage you to double check this advice in the Linux software forum!