the Beagle Bone Black

the BBB pins used to manage motors

cr_002 - the excavator.     Introduction. cr_002 is a whim. What? Is a whim? Yes. My son had a broken toy, an excavator toy, which when working, was managed by a wired controller. But before starting this challenge the wired controller was lost. So the toy was as you see in the picture below.  The challenge was: transform it bringing it to the year 2016. That is, repair the toy’s mechanic, install on it a CPU-board able to manage all peripherals, install a custom electronic board acting as middle-layer between CPU-board and motors, manage the toy remotely by installing  a wi-fi unit to control it over air; install a unique power unit (battery) able to send energy to all electronic devices on the excavator. Frankly speaking the toy plastic material was very good and the several modifies haven’t damaged it. When I decided to start this relaxing hobbyistic activity, I thought: “Simple thing,  I read on Internet several people did a similar thing, simple…" There is an italian proverb saying “from saying to doing there’s the sea in the middle”. I think there will be a similar sentence in every place of the world. So I’m going to describe what I made and what made me slow (obstacles) during this extremely simpe hobbyistic realization. Step 1-Mechanic. The toy has been doned from other child to my son, so the excavator has already past a hard period of its life in other hands. The mechanic was good but damaged. Even if we are talking about a toy, I have to admit is one of the best toys I’ve ever seen. The possible movement were (are) forward, backward, left, right, up and down for the bucket. The bucket was perfect as mechanic and as electric motor. It has been very simply to repair the left-right mechanic, the motor was ok. The main transmission was damaged for several reasons. Even if it was a toy the transmission between wheels and motor was realized by means of differential transmission. Frankly speaking I wasn’t aware how the motor transmits the movement to the wheels in our cars until I opend this toy, and I also learnt go-karts do not have this type of transimission and for that, they have several difficulties to face the curves. The differential transmission makes a different rotation speed for the right and left wheel. It acts in a way that the internal wheel  turns slower than the external wheel and so, the car can face better the curve. Maybe, for several toy incident, the gears related to the differential transmission lost their tolerance, so I repaired it with a new metal gear. The left wheel lost the tolerance with its axis so I used resyne glue to repair it. Below the differential gears and the plastic chunck detached from the left wheel. But, what I was struggling with, it was the gears between motor and wheel axis. From the right picture you can see a strange white box. It contains all gears to transmit rotation from the dc electric motor to the differential transmission of the main axis. Among the gears there was a shaft mounting two plastic gears, one getting the movement from the motor and the other transmitting the rotation to another gear. And the thinnest of the two lost its tolerance with the axis, so when the motor was rotating at high RPM it wasn’t still able to transmit the movement to the axis. But, the hard problem has been the fact, after repaired it, it kept on working for just some second after the completion of the whole excavator. Actually it broke 3 times. Why ? First the tolerance between axis and gear was alreay lost for the previous use of the toy. Second, my personal requirement has been making use of one battery as central energy unit; but I chose a cheap battery, payd about 14 euros, extremely heavy; the battery weight was about 2 Kg, so it was a tremendous increase of offort for the little gear. Third, I increased the emf at the motor pins, adding some Volts to the nominal voltage, so all gears were more stressed. After several tests the axis of the two gears broke in two chunk. Having listened to an advice of a friend of mine, I substituted the gear axis with a chunk of solder electrod (about 2 mm thick). From an electronic point of view I made use of a PWM to power the motor. Indeed the pulse width modulation is great way to change the rotation speed of a motor without loosing torque. In general if you want to put down the speed of a motor you can use a resistor, but doing it, you’re changing the elecromotive force applied to the motor but you’re also changing some of the main parameters of the motor: current, torque. But if in one second (1 sec. is an example) you give power to your motor for 40% of the time and you do nothing in the remaining 60 %, you’re making use of the pulse width modulation, a genial way to reduce motor speed without losing the torque. So replacing the gear shaft and buying an already-done PWM board, I figured out. Actually the pwm has been monuted between the power source and all motors, because the source power is 12 Volt and the motors have all a lower nominal voltage.         Step 2-The heart: BBB. According to the requirements, I need of a CPU board able to manage the motors and to receive commands through a wifi connection. I thought  the easy way to do that was making use of a cpu programmable board, and because I already knew the Beagle Bone Black I opted for that device. The BBB is a single computer boards with a SITARA CPU (1 Ghz) running some Linux distros and providing a few general purpose pins. The board computer does what you want it do (in theory). And how did the BBB come in handy? The board had to manage information coming from a remote client, using a wifi connection whose access point be the BBB board. The BBB had also elaborate the information and set some pins. So in other words, the board was switched on, it had several processes running on it and one of them was in charge of managing a wifi antenna and receiving information from a client connected to the board; the information was processed, and according to the read message(s) some pins were set. If you’re thinking it is(was) as simple as you read these rows you’re wrong. The BBB doesn’t provide a chip for wifi connection so you have to buy your wifi usb dongle and make it running on the BBB. I chose a not expensive wifi dongle usb: the TP LINK TL-WN725N (with Realtek chipset). If you surf the Internet, you can find several articles about the creation of a Wi Fi Access Point using the Beagle Bone Black. But going deeper and trying to transform what you read into a working device is another story. I wasted  much time trying to realise an access point following the instructions found on the web; to realise an AP reliable and working you have to find out the right combination of hardware, software and correct configuration. I lost a couple of weekend to make an AP with a  BBB Rev.C the TL-WN725N and the right Linux distro, and the topic is so long I decided to describe the step by step tutorial in another article on this site:  Make an access point with the BBB from scratch. Read the article carefully to implement your own AP.  So let’s move on the other requirement aspect. The BBB is now ready to accept wifi connection from a client, assign it an IP address (by DHCP) and receive everything over the air. Now we need of some agent continuosly waiting for a message, process in a predetermined manner the information (we can extract from the message) and change the value of some pin. There are several ways to do that on a linux machine, one of that is writing a C program. In the following lines I’m going to describe the cr_002 server side C program. The whole program, and its client side written in Objective C for iOS deveices, can be downloaded from the download section of this article.      Step 3 -The cr_002 C program. We are in the 2016 and do you write a C program? Yes it is about 50 years old, but for linux and electronic devices it’s the number ONE. Like each C program cr_002 begins with a main. The main has an infinite loop (see videogames programming) and within it, we make use of the UNIX fork. The fork generates  a child process and at the end of it (either it terminates in a right way or in a bad way), the while + the wait() instructions creates it again. The child process call the bbb_open(), the process_command() and the bbb_close();         while (true)     {        if( pid_command == -1 )         {            pid_command = fork();                         if(pid_command > 0)              {                bbb_open();                 printf(APPLICATION_NAME", launching process_command\n" );                 process_command();                 bbb_close();             }                     }                 pid = wait(&status);                 if( pid == pid_command )         {            printf(APPLICATION_NAME", process_command aborted\n");             pid_command = -1;        }            } Let’s examine the bbb_open() and the bbb_close() functions; before that you need a little explanation about the BBB pins setting. If you see within the  global.h file you’ll see the following lines //---------------------------------------------------------------//  from operating system////      move down    = echo 1 > gpio44/value;echo 1 > gpio45/value;//      move up      = echo 0 > gpio44/value;echo 1 > gpio45/value;//      move right   = echo 1 > gpio47/value;echo 1 > gpio26/value;//      move left    = echo 0 > gpio47/value;echo 1 > gpio26/value;//      move forward = echo 1 > gpio27/value;echo 1 > gpio46/value;//      move back    = echo 0 > gpio27/value;echo 1 > gpio46/value;//////      register pin_0 -> gpio45 -> bbb_P8_11 pin -> GPIO1_13 -> 32+13 = 45  OFF-ON of UP-DOWN//      register pin_1 -> gpio44 -> bbb_P8_12 pin -> GPIO1_12 -> 32+12 = 44  UP-DOWN//      register pin_2 -> gpio26 -> bbb_P8_14 pin -> GPIO0_26 -> 0 +26 = 26  OFF-ON of RIGHT-LEFT//      register pin_3 -> gpio47 -> bbb_P8_15 pin -> GPIO1_15 -> 32+15 = 47  RIGHT-LEFT//      register pin_4 -> gpio46 -> bbb_P8_16 pin -> GPIO1_14 -> 32+14 = 46  OFF-ON of FORWARD-BACKWARD//      register pin_5 -> gpio27 -> bbb_P8_17 pin -> GPIO0_27 -> 0 +27 = 27  FORWARD-BACKWARD//////      motor wire colors////      bbb_P8_12 - gray//      bbb_P8_11 - white////      bbb_P8_14 - yellow//      bbb_P8_15 - brown////      bbb_P8_16 - black//      bbb_P8_17 - black////--------------------------------------------------------------- #define COMMAND_LENGTH 2 typedef struct _commands_struct {    unsigned char command[COMMAND_LENGTH]; } commands_struct; //bbb_P8_11, GPIO1_13#define BBB_P8_11_BUCKET_ON_OFF 45 //bbb_P8_12, GPIO1_12#define BBB_P8_12_BUCKET_UP_DOWN 44 //bbb_P8_14, GPIO0_26#define BBB_P8_14_TURN_ON_OFF 26 //bbb_P8_15, GPIO1_15#define BBB_P8_15_TURN_LEFT_RIGHT 47 //bbb_P8_16, GPIO1_14#define BBB_P8_16_MOVE_ON_OFF 46 //bbb_P8_17, GPIO0_27#define BBB_P8_17_MOVE_FORWARD_BACKWARD 27 When I decided to manage the three motors of the toy, I opted for a custom electronic board (we are going to see it in the next section) that needed of six pins of the BBB board. Why? Because for each of the three motors, one pin is to manage ON/OFF status the other is for the rotation direction. So I fixed the pins to use, reading the port P8 and P9 features (I advice you the site and/or the book of Derek Molloy http://derekmolloy.ie) and after that, I wrote them in the global.h file. As you see they are the pins 11, 12, 14, 15, 16 and 17 of the P8 port. To use a pin, with the BBB, remember that the symbol P9_XX or P8_YY is just a human notation; for example P9_15 corresponds GPIO1_16;  the physical value to address the pin is  [if gte mso 9]> 0 0 1 68 388 extech 3 1 455 14.0 Normal 0 false false false EN-US JA X-NONE