We are using the SEN-13582 line follower array for a line following robot as part of a robotics competition. Recently we have had issues with the array randomly turning off as if the RESET for the I2C chip on the board. This would cause the LEDs to turn off/reset and derail the robot.
The micro-controller we are using with the array is the Teensy 3.5. The array is powered using a 2 cell lipo battery (7.4V) that is regulated to 5V using a LM2596 buck converter. The occurrence is seemingly random and copious amounts of debugging have proved to be fruitless.
Please see to this issue as soon as possible as the competition is happening in a week. Thank you!!
Before we dig into troubleshooting this issue in-depth, can you possibly give a quick recap of the troubleshooting steps you have gone through so we don’t go over something you have already tried? If possible, can you please take a few photos of the Line Follower Array and how you have it connected to the Teensy 3.5 and/or your robot?
Now, to get a better idea of the issue, does the entire circuit reset or is it just the Line Follower Array? Does adjusting the brightness potentiometer help at all? Similarly, does enabling or disabling strobing change anything? Finally, does the issue show up regardless of the battery’s charge level?
So we have tried to use different batteries, different connectors, different buck converters, as well as turning both strobing on and off. The line follower array is the only part of the circuit that resets, and when it resets, it seems to fall into a reset loop where the it permanently is not able to give serial values. However, unplugging and re-plugging the 5V power wires on the line follower array would make it ok before it does the reset again. Adjusting the potentiometer does not seem to help, and we are also able to test a narrow range on that due to the competition surface. The issue shows up regardless of battery charge level, and probing the 5V rail shows no noise.
The yellow and blue wires in the red MTA connectors are connected to SDA and SCL on the Teensy 3.5. The two 5V wires are tied together, as are the two GND wires. Those are put into a screw terminal that is on a 5V rail.
Thanks for that information and taking those photos. Everything looks connected properly and you have already tested pretty much everything I would recommend off the bat so I am not really sure what would be causing this issue.
Are you possibly able to pull the Line Follower Array off your robot and test it with another microcontroller circuit just to see if you can replicate the problem? That might help identify where the issue is coming from. Also, if you add debug statements are you able to identify when the Array locks up? Is it the same point every time or does it appear to be random? Similarly, is the time between when the robot is powered on and the Array locks up consistent or does it happen at random intervals?
So it occurs mainly randomly, using both the Teensy and the Arduino MEGA.
However, we do notice a more frequent occurrence of the shutdown when the robot has to do a hairpin turn on the course. That seems to be the only observable pattern. Other shutoff times are completely random and spontaneous.
The symptoms you’re describing sound a lot like you have an intermittent loose wire somewhere that disconnects very briefly causing I2C communication to fail. the connections on the green board look good from what I can see, but I can’t really see what’s going on where you’re connecting to the line follower. You might try re-doing that connection and see if things clear up.
The only other thing I can think of would be some electrical noise or a power dip that might be coming from your motor drive circuit. You could try a separate power supply for the logic and motors and if that clears the issue, it’s probably being caused by the motors. Have you tired disconnecting the motors from their power source and manually ‘driving’ the robot around by hand? If the sensor doesn’t fail then, it’s likely motor noise or a power issue.
We’ve not seen this issue come up before with this sensor and that is leading us to thing it’s something unique to your setup.
We have tried different kinds of connectors on the line following array with the same result. We agree that it is likely to be a noise issue for power, since we see in the schematic that the reset pin on the I2C chip is tied to power. Our motors are over 10 inches away from the array and the processor, and it runs on a separate 15V line controlled using an H-Bridge that is decoupled from the rest of the circuit using opto-couplers.
Would bridging the 5V power and GND of the line follower array with an additional capacitor be useful, and if so what value should we use?
Would bridging the 5V power and GND of the line follower array with an additional capacitor be useful, and if so what value should we use?
That might be helpful. It’s hard to determine what size cap you need but I’d stick a 0.1uF and a 100uF cap across the power leads for the sensor as close to the board as possible to act as a filter. You may need to experiment with the value of the 100uF cap for best results.
