nRF52840 clock only ticks when a debug cable is connected

[lindemer]

I have a Thingy:91 connected to a Garmin LIDAR Lite v4 (built on the nRF52840) via four GPIO cables. The code running on the Garmin reads one GPIO pin to signal a measurement and toggles another one when the measurement is complete. The two devices communicate exactly as expected, but ONLY when the debug cable is still connected to the Garmin.

I've used a voltmeter to verify that the Thingy:91 continues to signal for a new measurement, but after the debug cable is removed, the Garmin sends one more pulse and then stops. When I plug the debug cable back in, JLinkRTTLogger shows that the program has picked up right where it left off. I've tried starting the devices without any debug cables plugged in, but this also causes the Garmin to stop after a single measurement has occurred.

Garmin has not provided a hardware schematic showing how the internal nRF52840 is connected to the LIDAR sensor, and their driver library is pre-compiled. Is there any explanation for why the clock would stop without an external debug probe connected?

[Jared]

Hi,

Just to make sure this isn't a matter of how the sensor is supplied: 

but ONLY when the debug cable is still connected to the Garmin.

Could you elaborate a bit more (maybe provide a sketch) on how they are connected? I'm not sure which debugger interface on the LIDAR sensor that you're referring to. I see from the datasheet that it has a I2C interface along with VCC, GND and Mode control pin. Maybe I'm looking at the wrong datasheet?

How is the Thingy:91 supplied?Is the LIDAR sensor supplied by the Thingy:91? The debugger interface is connected to the 1V8 supply net of the nRF9160. Is the Thingy:91 running without the debugger connected, if so what do you measure on the supply of the LIDAR then.

regards

Jared 

[lindemer]

Hi Jared,

Please excuse the delay. The datasheet you're referring to is for the previous generation of the device. The new one is available here. A 10-pin JLink cable has been connected as shown at the top of page 8, and the four other exposed pins are connected to the SPARE1-SPARE4 pins on the Thingy:91. (The boards are also connected to a shared ground.)

We are interfacing with the LIDAR sensor with the method described at the top of page 3. (Two GPIOs are used to trigger a measurement and see if one is ready, and the other two are I2C for reading the distance.)

The problem is that this system works only when the JLink cable is connected to the LIDAR sensor (not the Thingy:91) AND JlinkRTTLogger is running. We can confirm this by measuring the voltage on the GPIOs and seeing that the LIDAR stops toggling the "monitor" GPIO pin when the logger is stopped or the JLink cable is disconnected.

Hope this helps.

[Jared]

Hi,

I assume that you're still supplying the device after you've disconnected the J-link?

lindemer said:

The problem is that this system works only when the JLink cable is connected to the LIDAR sensor (not the Thingy:91) AND JlinkRTTLogger is running. We can confirm this by measuring the voltage on the GPIOs and seeing that the LIDAR stops toggling the "monitor" GPIO pin when the logger is stopped or the JLink cable is disconnected.

I think this depends on the application running on the LIDAR. There is nothing in the nRF52840 HW which makes it dependent on the J-link and RTT logger for it to run an application. As we don't have any insight on the application running on the LIDAR, I think this question should be directed to Garmin.

regards

Jared