Two years ago, I designed my own motor driver breakout board using two L298N motor driver ICs and had them manufactured through BatchPCB. Here it is, finally finished.
L298N motor driver, designed and printed two years ago, finally finished.
The KiCAD files are on my github repo here.
tl;dr: I made this motor driver for my kiwi drive robot. Just go look at the pictures at the bottom of this post. There really isn’t much to circuit board design other than common electrical sense (which can be obtained through self-study) and knowing how to get it made (pay a service like BatchPCB or etch it yourself). If you want to read a rambling story, read on.
I designed the board two years ago when I started working on my kiwi drive robot as my electronics class project. I spent a month figuring out how to use KiCAD, learning to read datasheets, discovering Digi-Key, and researching the pros and cons of the various ways of getting a circuit board printed.
Motor driver schematic in KiCAD.
I eventually decided to pay around $20 to get a board printed through BatchPCB. After a month or so, they sent me two boards (though I had only paid for one!). Unfortunately, I used the wrong footprint size for the diodes I had ordered, so they did not all fit on the board. I spent a day or two trying to use the board without the diodes, and for reasons unclear to me now, I decided the board did not work. I ended up buying a simple line-following robot kit instead and finished that for a grade.
The parts lay in a drawer for almost two years, until one Monday last February.
I tried once again to troubleshoot my motor driver and found, to my surprise, that there was no troubleshooting to do! It made my day (to say the least) to discover that my first-ever circuit board worked, after all. Since I had already cut the chassis and bolted aluminum hubs to my omniwheels in that electronics class two years ago, I had a working kiwi drive platform within a few hours.
Initially, I powered the robot with four AA alkaline batteries. I noticed that when the robot (thus the motors) abruptly changed directions, the Arduino on board would reset. Since at this point I had not yet found a solution to the oversized diodes (or undersized board) and had left them off, I guessed that the motors were producing voltage spikes or drops that caused the board to reset.
In fact, however, the motors were actually pulling so much current (when changing directions) that the alkaline batteries were simply failing to meet the current demand, effecting the resets. This problem was mitigated initially by using a 2S lithium polymer battery and later by using a used MacBook Pro battery to power the robot.
Finally, I made the shottky diodes fit by filing them all down from 0.1090 inches in width to 0.0950 inches so they would fit on the 0.1 inch spacing on the board. Occasionally, I filed them too much and exposed a bit of the copper inside, but that did not seem to affect the diodes’ performance.
The board should be capable of delivering up to 2 A of continuous current each to four motors at 35 V, given adequate cooling. The heat sinks that are currently bolted to the motor drivers are probably insufficient for dissipating the kind of heat that will generate, but they are enough for the kiwi drive robot.
More pictures:
