RPLidar A3M1 incorrect measurements in sunlight

We have equipped several quadrocopters with RPLidar A3M1. It works wonderfully in cloudy skies. As soon as the sun shines, it doesn’t. An enquiry with Slamtec Support revealed that the LIDAR should be operated in “Stability” mode for outdoor use. This is also stated in the technical documentation on page 12. The current driver in Ardupilot does not seem to take this into account. Unfortunately, my programming skills are not good enough to write my own driver for the A3. Is there any other possibility? I might sponsor a working driver if someone would be willing to program one… :slight_smile:

(There are no obstacles around)

(Slamtec Documentation)

Hi @webmotions,
I do not have an RPLidar to help out with the driver, however, I did add a small feature a few months back that adds a customizable “low pass filter” to the output of the sensor. If because of the sunlight, your sensor is giving out false spikes, then the filter can help you (of course this is just a workaround, and doesn’t solve the actual underlying problem)

This is included in the Copter 4.1 beta. There is a slight problem with BendyRuler (auto modes avoidance) in beta 5, which will be fixed in Beta 6. So If you want to flash beta 5 and test out the filter in Simple Avoidance (loiter), or wait till Beta 6 for BendyRuler then that might potentially solve your problem

Also can you attach a log of the flight? Will be helpful to see the kind of noise the sensor is producing…

Hi @rishabsingh3003

Thank you very much! Unfortunately I won’t have access to the device again until Monday and will post the log then.

I can’t deliver a beta version to the customer, so I will try to port your patch back to 4.07.

Ah might not be easy to port the low pass filter… we changed the proximity library from ground above just a few months back and then I applied the filter patch. It’ll be best if you just add a custom filter from scratch of your own and see what might work to 4.0.7. Let me know if you need help doing this

Hello @rishabsingh3003,

Below are 2 log files. The file 2021-06-19 is the recording of the flight in the video. But the PRX recording seems to stop after a short time. You can only see how the obstacle detection was switched on and off via channel 13. Therefore I have added the log 2021-06-17. In this flight the PRX recording takes place over the entire flight. The potential thrust loss story was due to faulty ESCs, which was determined and corrected independently.

Thanks for your efforts!