vl53l8cx with SPI interface start ranging fails

[lukestroh]

To begin, I am using the vl53l8cx breakout board from Pololu with a Nucleo-F446RE and PlatformIO. I have been having trouble getting the VL53L8CX to start ranging using the SPI2 interface. I get successful responses from the `_is_alive()` and `_init()` methods, but find that I get an error status of 1 from the call to `vl53l8cx_dci_read_data()` within the `_start_ranging()` method. Subsequently, a few lines down, `_start_ranging()` errors out with VL53L8CX_STATUS_ERROR.

I tried following @RJRP 's solution here, but it didn't solve my problem.

I have jumper wires from the Nucleo-F446RE to a breadboard. The wires are possibly too long, ~12cm in length.

For clarity, the lines in question:

/* Read ui range data content and compare if data size is the correct one */

// This gives status = 1
status |= vl53l8cx_dci_read_data(p_dev, (uint8_t*)p_dev->temp_buffer, 0x5440, 12);

// This gives status = 255
(void)memcpy(&tmp, &(p_dev->temp_buffer[0x8]), sizeof(tmp));
if(tmp != p_dev->data_read_size)
{
status |= VL53L8CX_STATUS_ERROR;
}

I have enabled most of the macros in `platform.h` to reduce the size of the information (I left NB_TARGET_DETECTED, DISTANCE_MM, and TARGET_STATUS as active). My SPI baud rate is 1.3125 Mb/s, and I have properly configured CPOL = 1 and CPHA = 1. For further information/configuration setup, here is the link to my project repository: https://github.com/lukestroh/nucleo-f446re-vl53l8cx-spi-platformio

[John E KVAM]

Two things to try... Shorten your SPI wires and try a different speed. 

The init is a large write, and there are firmware handshakes. I find it odd that you got through so much only to fail. You must be pretty close. 

I'm betting a shorter wire will do it. 

After that we can try better wires, twisted pair and some other techniques to get your length back.

- john

[John E KVAM · Accepted solution]

I had one customer claim that his power supply was a touch marginal. When the sensor started, it drew more power and things went south. 

Perhaps you could put a scope on the power input and see if it drops about the time you start the sensor. 

That might explain how you can communicate just fine - until you issue the start command. 

On that Pololu board, you might also just put a scope on the 3v3out pin and see what you get. 

Unfortunately, this is just a guess. But I think it's worth a test.

- john

[lukestroh]

Thanks @John E KVAM, I scoped the 3.3V output and you're right, it was pulling down. I fixed the power supply and everything is working, I appreciate the insights.

[RobotMachines]

This is good to know about the initial power draw. I just went through a weekend of trouble because of my sketchy wiring on a breadboard but was finally able to get the VL53L8CX working in both SPI and I2C to an Arduino Nano RP2040 Connect. Used USB power and 3.3v from the RP2040 for the sensor. So far so good. 

Note to all, keep trying, it works. Can't wait to try them on my robot. Have 20 more on the way.