cancel
Showing results for 
Search instead for 
Did you mean: 

VL53L8CH, VL53L7CH Multi-Sensor Time-of-Flight Fusion

SelimMuhammed
Associate II

Dear ST, For the ToF sensors supported VL53L8CH or VL53L7CH is it possible to make a hardware that depends on 4 x VL53L8CH or 4 x VL53L7CH integrated to increase the number of zones from 8 X 8 to 16 X 16 ?
and if it is possible what would be the limitations in both HW and SW for this system?
also provide me a reference for this implementation if exists.
Thank you.

1 ACCEPTED SOLUTION

Accepted Solutions
Bin
ST Employee

Hi:

A guide for using the artificial intelligence enabler VL53L7CH and VL53L8CH multizone Time-of-Flight (ToF) sensors - User manual

According to UM file, a multi-module implementation using several VL53L7CH or VL53L8CH devices is supported at system level.

Both devices support I²C communication. VL53L8CH also supports SPI.
Section 2.3 states that the I²C address can be changed to avoid conflicts and to facilitate the use of multiple modules for a greater system FoV.
UM3183 also includes reference diagrams for:
Multiple sensors on I²C bus
Multiple sensors on SPI (for VL53L8CH)
Therefore, from a hardware connectivity point of view, using 4 sensors in one system is possible.

However, it is important to clarify that this does not create a native single-device 16×16 zone sensor. Each module still operates independently with its own 8×8 output. A larger effective array can only be achieved by host-side stitching / fusion of the four sensor outputs.

In addition, the resulting FoV extension should not be considered seamless. Based on the module optical structure and orientation behavior described in UM3183, several practical limitations must be taken into account:

    each sensor has its own zone orientation
    the effective image is subject to horizontal and vertical flip
    module mounting orientation must be kept consistent
    optical axes may not be perfectly aligned
    mechanical tolerances may introduce offsets
    adjacent FoVs may overlap or leave gaps
    edge zones may not behave exactly like center zones
For these reasons, a 4-sensor arrangement should be considered as a system-level expanded FoV solution, rather than a perfect mathematically uniform 16×16 grid.

Main limitations
Hardware

On I²C, each device must be assigned a different address, typically by controlling LPn during initialization.
On SPI (VL53L8CH only), each sensor requires an independent NCS line.
Mechanical placement, sensor orientation, FoV overlap, and possible optical interaction between modules must be carefully evaluated.
Software

The ULD does not provide a built-in function to merge 4 sensors into a native 16×16 output.
The host must handle initialization, synchronization, coordinate remapping, calibration, and data fusion.
Using 4 sensors increases bus traffic, memory usage, and host processing load.

Reference in UM3183
Please refer to:

Section 2.3 – Schematics and I²C/SPI configuration
Figure 4 – Multiple sensors on I²C bus
Figure 5 – Multiple sensors on SPI
Section 2.2 – Effective orientation
Section 4.15 – External synchronization pin (VL53L8CH only)

Best regards,
Bin FAN

View solution in original post

2 REPLIES 2
Bin
ST Employee

Hi:

A guide for using the artificial intelligence enabler VL53L7CH and VL53L8CH multizone Time-of-Flight (ToF) sensors - User manual

According to UM file, a multi-module implementation using several VL53L7CH or VL53L8CH devices is supported at system level.

Both devices support I²C communication. VL53L8CH also supports SPI.
Section 2.3 states that the I²C address can be changed to avoid conflicts and to facilitate the use of multiple modules for a greater system FoV.
UM3183 also includes reference diagrams for:
Multiple sensors on I²C bus
Multiple sensors on SPI (for VL53L8CH)
Therefore, from a hardware connectivity point of view, using 4 sensors in one system is possible.

However, it is important to clarify that this does not create a native single-device 16×16 zone sensor. Each module still operates independently with its own 8×8 output. A larger effective array can only be achieved by host-side stitching / fusion of the four sensor outputs.

In addition, the resulting FoV extension should not be considered seamless. Based on the module optical structure and orientation behavior described in UM3183, several practical limitations must be taken into account:

    each sensor has its own zone orientation
    the effective image is subject to horizontal and vertical flip
    module mounting orientation must be kept consistent
    optical axes may not be perfectly aligned
    mechanical tolerances may introduce offsets
    adjacent FoVs may overlap or leave gaps
    edge zones may not behave exactly like center zones
For these reasons, a 4-sensor arrangement should be considered as a system-level expanded FoV solution, rather than a perfect mathematically uniform 16×16 grid.

Main limitations
Hardware

On I²C, each device must be assigned a different address, typically by controlling LPn during initialization.
On SPI (VL53L8CH only), each sensor requires an independent NCS line.
Mechanical placement, sensor orientation, FoV overlap, and possible optical interaction between modules must be carefully evaluated.
Software

The ULD does not provide a built-in function to merge 4 sensors into a native 16×16 output.
The host must handle initialization, synchronization, coordinate remapping, calibration, and data fusion.
Using 4 sensors increases bus traffic, memory usage, and host processing load.

Reference in UM3183
Please refer to:

Section 2.3 – Schematics and I²C/SPI configuration
Figure 4 – Multiple sensors on I²C bus
Figure 5 – Multiple sensors on SPI
Section 2.2 – Effective orientation
Section 4.15 – External synchronization pin (VL53L8CH only)

Best regards,
Bin FAN

Thanks, Bin 
I really appreciate your words that clarifies all approaches for this configuration 

wish you all the best