{"id":4166,"date":"2020-09-22T13:16:30","date_gmt":"2020-09-22T12:16:30","guid":{"rendered":"http:\/\/dronesonen.usn.no\/?p=4166"},"modified":"2020-10-04T20:43:08","modified_gmt":"2020-10-04T19:43:08","slug":"third-sprint-puck-buddies","status":"publish","type":"post","link":"https:\/\/dronesonen.usn.no\/?p=4166","title":{"rendered":"Third sprint – Puck Buddies"},"content":{"rendered":"\n

Nicholas<\/h2>\n\n\n\n

Electrical:<\/strong><\/p>\n\n\n\n

\"\"
Figure\u00a01<\/em>\u00a0Arduino pin assignment\u00a0<\/em> <\/figcaption><\/figure>\n\n\n\n

All the parts we use to test are subject to change before the final implementation. They do however work in the same logical way so the software should stay more or less the same even if the parts are changed. <\/p>\n\n\n\n

Current\u00a0BOM:\u00a0<\/strong><\/p>\n\n\n\n

  • Arduino Uno\u00a0<\/li>
  • Potmeter\u00a0(50K\u03a9)\u00a0 <\/li>
  • BYJ-48 stepper motor and driver board\u00a0<\/li>
  • Simple button\u00a0<\/li>
  • 10K\u03a9, used as a pull-down resistor (Only pullup internally)\u00a0<\/li>
  • 3 LEDs\u00a0<\/li>
  • Relay module\u00a0<\/li>
  • 250\u03a9 resistors (to protect LEDs)\u00a0 <\/li><\/ul>\n\n\n\n

    Started to connect all the electrical components together to test a base version of the software.  <\/p>\n\n\n\n

    \"\"
    Figure\u00a02<\/em>\u00a0Basic test setup to test the Arduino code\u00a0<\/em> <\/figcaption><\/figure>\n\n\n\n
    \"\"
    Figure\u00a03<\/em>\u00a0General arrangement for components in box\u00a0<\/em> <\/figcaption><\/figure>\n\n\n\n

    However, to make it easier to transport around we decided we wanted to make a box containing all the components. We could then use this box to test all the features of the software while waiting for the real machine to be constructed. I\u2019ll come back next week with more schematics and drawings as well as more detailed reasoning for the different components we ended up using. <\/p>\n\n\n\n

    Lots of <\/strong>debugging the Arduino code. <\/strong> <\/p>\n\n\n\n

    Anders and I have spent a lot of time this week debugging the Arduino code and adding functionality. I\u2019m now going to use the next few paragraphs to highlight some of the issues we faced and how we ended up solving them. <\/p>\n\n\n\n

    Get stepper to work<\/strong>:<\/strong> <\/p>\n\n\n\n

    First of was to get the stepper to work. As most are probably aware of, steppers is based on relative positioning which means that the stepper itself doesn\u2019t know where it is. Only the \u00b5Controller that sends the steps knows where it is supposed to be. Due to this we needed to add a homing\/calibration function that would set the stepper at position 0 when the switch was activated. The limit switch is attached to an interrupt pin (pin 2) to make sure that the 0 point is registered as accurately as possible.  <\/p>\n\n\n\n

    Another issues was that we needed to convert between degrees and steps used for the stepper. For testing purposes we are using a  BYJ48 stepper motor. If we look at converting degrees to steps here we will notice a problem. The total number of steps for this stepper is 2048 (more or less, some searching points to the gear ratio being different between different manufacturers). That means that 1\u00b0 =  2048 \/ 360\u00b0  => 1\u00b0 = 5.6888 steps. Since a stepper only can move whole steps we would lose 0.6888 steps for each degree moved. If the stepper is asked to move from 0\u00b0 to 90\u00b0 that would mean it would miss a total of 61.2 steps or roughly 10\u00b0. Due to this we added an extra function that tries to add the extra steps missing based on the degree target. While still not entirely accurate it\u2019s considerably better than the original solution.  <\/p>\n\n\n\n

    One could of course ask the question why we wouldn\u2019t simply send steps over the serial connection instead of the and eliminate this whole issue. Our reasoning is that by creating this abstraction the user, either a person or an program, doesn\u2019t need to know what kind of stepper and gearing that is used. The \u00b5Controller will be programmed according to parts used and then it will simply convert the incoming degrees to steps for the stepper that is used. If we would have sent steps both the client code, currently Python, and the \u00b5Controller would have to be updated.  <\/p>\n\n\n\n

    Other stepper implementation:<\/strong> <\/p>\n\n\n\n

    We deiced to use the standard stepper library as it\u2019s easy to use and works out of the box. However, as is described here [1] the Stepper.step() function is blocking. As we would need the Ardunio to perform several tasks more or less at the same time we couldn\u2019t send all the needed steps at the same time. Instead we placed the Stepper.step() inside a while loop that continually increments it by 1. By doing this we are able to read new serial data and shift the target for the stepper without it needing to reach its original target. <\/p>\n\n\n\n

    Serial.parseInt<\/strong>() problems:<\/strong> <\/p>\n\n\n\n

    So, moving on to our next problem. Slow response from the serial read function in the Arduino code. While doing some testing I noticed that the stepper used an unusually long time changing direction after new coordinates were given. Due to it being late I was tired I decided to just turn a LED on and off with a simple line of code in-between. Using this method I noticed that the LED was staying lit for far too long when Serial.parseInt() was called. So, I went back to the documentation to try and find out how the function worked and why it would be so slow. The answer to the first question is that the Serial.parseInt() function reads the input until either a non-number data is registered or until the timeout reaches 0. The way I implemented the serial communication means that I do not send any \u2018\\n\u2019 (newline) after each message. Due to this the first criteria for the Serial.parseInt() is not met and a such it will wait until the timeout reaches 0. By default the serial timeout is set to 1000ms, and this explained the long delays I noticed. By using Serial.setTimeout() and setting the value to 10ms I greatly improved the response time of the stepper.  <\/p>\n\n\n\n

    Summary:<\/strong> <\/p>\n\n\n\n

    We have done a lot of work this week on the software for the project and we have now established the necessary groundwork. This means that the all the core software components now are in place and we can start to create new branches and add features. This is a huge milestone as it allows us to more independently work on different components. We can now start to optimize the Ardunio code, or look at how to implement OpenCV and YoloV3 on a RaspberryPi or another small computer. I also need to start working on the electrical schematics and wiring diagrams.  <\/p>\n\n\n\n

    Referanser<\/strong>\u00a0<\/p>\n\n\n\n
    [1]  <\/td>\u00abArduino,\u00bb [Internett]. Available: https:\/\/www.arduino.cc\/en\/Reference\/StepperStep. <\/td><\/tr><\/tbody><\/table>\n\n\n\n

    Anders<\/h2>\n\n\n\n

    Testing Launching Mechanism<\/strong><\/p>\n\n\n\n

    Thursday last week we tested our firing mechanism prototype, this prototype is scaled down 50%.<\/p>\n\n\n\n