NXP i.MX RT1170 High End Processor Evaluation Kit - Review

Table of contents

RoadTest: NXP i.MX RT1170 High End Processor Evaluation Kit

Author: dimiterk

Creation date:

Evaluation Type: Development Boards & Tools

Did you receive all parts the manufacturer stated would be included in the package?: True

What other parts do you consider comparable to this product?: RPI zero version 1 would be comparable in speed to CortexM7.

What were the biggest problems encountered?: 1. MIPI Camera SDK test firmware was not functional. Screen gets activated but there is no camera image. 2. Not much documentation on the camera 3. LCD screen came scratched!!!

Detailed Review:

This road-test we'll take a look at the MIMXRT1170-EVK, courtesy of Element 14 and NXP. Many thanks to Element14 and for providing this oportunity to test this platform.
The package included two boxes. The first box contains the MIMXRT1170-EVK which is a development board with a 1GHz MCU platform. It also includes a power adapter, a MIPI OV5640 camera and a Micro USB cable.
The second box includes a 8 inch TFT display with a capacitive touchscreen.
1. Introduction
The RT1170 is a dual core MCU that couples a Cortex Core M7 MPU running at 1Ghz  with an M4 core running at 400MHz. The dual MPU IC comes in a 289 MAPBGA package so assembly by hand may be tricky.
The main reason behind this roadtest was to test the MIPI DSI display. This is a 5.5 inch capacitive touch display that uses 2 MIPI lanes.
I was a bit surprised, since after unboxing the display the first thing I noticed were some scratches on the bottom corner side. I'm not sure if this was part of quality testing.
Even though the display uses 2 MIPI lanes plus a clock lane, it uses a different 40pin FPC cable in comparison with the venerable 15pin connector found on RPI.

 

 

 

 

{gallery} Unboxing

image
image

 

 

2. Peripherals hardware

The EVB has a number of peripherals :

  • 4 digital I2S microphones
  • analog mic
  • QSPI
  • 2x 1Gb Ethernet connector
  • SD card holder
  • M2 connector
  • JTAG Programmer
  • SDRAM
  • NOR Flash
  • NAND Flash
  • CAN Bus
  • 6DOF (Accelerometer and magnetometer)

 

 

 

 

 

 

image

 

{gallery} TFT display
image
image

 

 

 

 

The MCU is supported by a number of tool-chains however I selected the MCUExpresso. This is a heavily customized version of Eclipse with a number of supporting middleware and libraries.

The SDK can be installed either by customizing and downloading it from the NXP site or within the Espresso IDE .

To connect the display pull the FPC connector and insert the display FPC cable.

 

image

 

 

image

 

Make sure to select the EVMIMXRT1170 evaluation board.

 

 

 

{gallery} SDK
image
image
image
image
image
image

 

 

 

 

After installing the SDK, the next step was to try the examples that came with it.

 

{gallery} Apps
image
image
image
image
image
image

 

 

 

 

Compass (Magnetometer ) App

 

 

{gallery} Other Apps
image
image

 

 

 

 

The main focus was testing the display. A couple of middle ware graphic stacks were tested.

 

 

 

 

{gallery} Display
image
image

image

image

image

 

 

TFT display tests.

 

 

The MCU Expresso comes pre-loaded with at least 3 GFX (Graphical display stacks)

 

The next step was to test the camera. This is an OV5640 camera that is configured in MIPI mode.

Testing all 4 examples in both YUV and RB mode did not work.

 

The display gets activates , however the camera does not output any image.

Due to this , it was not possible to test the machine learning examples.

 

 

{gallery}TFT display tests
image

image

 

 

 

 

 

Testing the TFT display.

 

 

 

The Good

1. A number of opensource middleware for the basic implementations.

2. Graphic display stacks.

3. Board includes Arduino connectors.

4. A number of high speed interfaces plus camera and display interfaces for HMI applications.

 

The Bad

1. All camera demo apps do not work in either RGB or YUV mode.

2. WIFI chipset on the backside is not populated.

3. SPIDIF input/output jacks are not populated.

4. Cannot test TensorFlow lite examples due to non-functional camera.

5. Cannot use the MIPI display with RPI since it uses a different pitch connector.

Anonymous