Project Report. Object Following Robot

Transcription

1 Project Report Object Following Robot Group Members: 1. Haad Yaqub Rathore ( ) 2. Muhammad Umar Javed ( ) 3. Huzaifa Arif ( ) 4. Hunza Zainab ( ) Project Objective & Introduction: We intended to make an object following robot that follows the object that is closest to it to its right, left and the front. We tried to do this by using sound. Our sonar transceivers sent and received sound waves in ultra-frequency range (so that the sensors do not detect human voice or voice audible to human beings). The wave that was the highest in amplitude after reflecting back from the object indicated the direction of the object and we made our robot move towards it. Project Specifications: LM324 quad Op-Amp Op-Amp 6 A number of capacitors (from 1nF to 5nF) 10 SQ 40R for SONAR reception 5 SQ 40T for SONAR transmission 5 LP555 (timer used to generate TTL of 40KHz) 2 74HC138 (decoder) 1 1 OR gate IC 1 1 Inverter IC 1 A number of resistors At least 5 working PCBs H bridge LM298 1 Germanium diodes

2 Burnt: 3 x 741 Op-Amps. 2 x lm324 quad Op-Amps sensors' and transmitters' legs borke. 1 x LM298 (the legs borke because its really hard to insert this IC in the breadboard) Project Deliverables: Generated an AC signal using 555 timer and successfully transmitted, received, amplified it. Once the signal is amplified, a comparison was done to check whether it is only noise or a signal from an object. After all three transceivers were set and received different amplitude signals, we compared the three signals and decided where the object is using a decoder. Once it was known where the object is, we made our motors work to turn the robot in to that direction. Project Modules and Design Overview: Our project can be divided into 3 unique circuits. 555 Timer circuit (Transmitter circuit). Receiver circuit. Decider circuit. A 555 timer IC was used to create a TTL wave of 40kHz that the transmitter would transmit. The circuit used is as follows:

3 We used two formulae found on the internet to calculate our the capacitor and resistor values to create a TTL of desired frequency. f = 1/(0.693 x C x (RA + 2 x RB)) And we also tried to maintain the delay in the circuit at 50% to create a perfect TTL using the following formula D = t1/t = (RA + RB) / (RA + 2RB). Next comes the circuit that received, amplified and rectified the signal. This 40KHz sinusoidal wave that the receiver received (the receiver always received a pur sinusoidal wave) was first amplified by a factor of 20. The op-amp used here was 741. The values of resistors were chosen so that the gain was 20. Also this is a non-inverting amplifier. Next we rectified the resulting signal. This was a halfwave rectifier with a large enough capacitor to convert the signal to DC. We also added a low pass filter that eliminated any higher frequency noise. Next the signal was amplified again (this was done because the voltage of the signal was low due to the drop taken by the diodes). Both of these amplifications used the 741 Op-Amp. Next we used the LM324 Op- Amp that compared this resulting DC signal to a reference DC signal of 0.2 volts (we set this exactness using a potentiometer (gave us flexibility in our ciruit). This PCB at the very end gave either of the two outputs: 0 or 5 volts. It must be mentioned here that in between the these amplifier circuits we did have at least two high pass filters which were necessary to remove the DC offset that crept into the circuit for no reason. (we still cant figure out why this DC offset came but it was clearly there and was tampering with the proper decsion making process).

4 The last step was creating a circuit that could decide where the object is and also make a move according to it. For that we used a 3 to 8 decoder. It took the output of all three receiver circuits and if there was no object 000 minterm was activated and the object remained stationary. If there was an object in the middle then minterm received was 010, if on the right 001 and left 100. The three outputs and 010 were taken and their logical OR given to the LM298 s logic supply. On the other hand, the right and left minterms were given directly to the H bridge terminal that controlled turning and the forward one was given to the H bridge terminal that controlled forward moving. The logical OR as stated above was the one controlling the first motor (for forward motion) while the the right and the left minterms (001 and 100 respectively) were used to drive the second motor for turning purposes. Also the elegance in this design made sure that the receiver circuit was completely detached from the decision making circuit otherwise we would have face unnecessary voltage level drops as the motors started to power up. Technical Issues Faced and how you resolved them: Lack of batteries (portable and light weight). Heavy batteries caused the body of the robot touch the ground causing turning to be impossible. Selecting a better car body. The body we got was from a toy car. It was not designed withstand weight. Also the decision making and the turning had to be achieved using a decoder. It was not a differential drive (that s what most robots have). Thus the logic we designed was based on the type of the robot we had. There was an unnecessary DC offset every time we switched on the circuit. Because we were working with a comparison based circuit, the DC offset over the sinusoidal signal was an anomaly in our system and had to be eliminated. We finally worked it out the basic first order Low Pass Filter would solve the problem and it did. The sensors were broken and did not function properly. The first day of the project (the day we started working) every sensor was working perfectly fine but by the end of the project they simply weren't. The entire circuit was designed on very precise voltage levels. By the end of the project the sensors gave such low voltage levels that they were simply useless. We then had to demonstrate the

5 working circuit using a sinusoidal source instead of the sensors. Had the store had more working sensors and receivers our demonstration could have been with sensors instead of sinusoidal sources. I would suggest that since it s very hard to work with analogue sensors, using either the HCSR04 IC which basically does everything our circuit was doing in a single IC or getting a few spare sensors for the project. We faced difficulty controlling the gain of op amps using fixed resistance values. We were able to control the gain of the amplifiers using variable resistors instead of fixed one; this gave us more flexibility over experimental errors that are inherent in any analogue system. Our timer circuit worked perfectly fine on a breadboard and when the exact same components were transferred to a PCB, our frequency value which is dependent on resistance and capacitance changed drastically. There were probably some of these two properties available in the copper tracks of the PCB. To overcome this problem we had to hit and try with resistances and capacitances close to our decided values by adding them in parallel and testing our PCB.