Week 9 - Output devices

This week I decided to design a board that could control two DC motors. I've decided to focus my final project on the creation of a robot capable of locomoting across a cuboct structure using continuous rotation joints. For this, either standard DC motors or Brushless DC (BLDC) motors will be most suitable.

Designing the board

Using Kokopelli, I started out by replicating Neil's DC motor control board which uses the A4953 Full-Bridge DMOS PWM Motor Driver chip. I then extended the design by adding a second chip and the associated capacitors and connectors. I connected the input's of the second A4953 to PA7 and PB2, as they are only being used as straight forward outputs.

Soldering pad on the bottom of the H-Bridge

The A4953 has a ground pad on the underside of the chip. To attach this I applied solder paste to the pad and placed the chip on the board. I then soldered the pins to the board and came back the ground pad with the heat gun to melt the solder paste. I was a little worried about damaging the chip so I'm not sure whether I actually metled the paste however, measuring continuity with the multimeter indicated that a connection had been made.

Board not working

After stuffing the board I managed to connect the power the wrong way around and it started smoking. As a precaution I changed both H-bridge chips and the regulator and used a red Sharpie to mark the V+ pins. I altered the code to account for the second H-bridge and also the fact that I was using a pin on port B. I uploaded it OK but when I connected a DC motor nothing happened. The ATTiny seemed to be working OK as when I measured the voltage of the output pins PA2 and PA3, their voltage was moving up and down as instructed by the code. When measuring the voltage across the output of the H-bridge chips I didn't detect any changes.

Increasing the voltage above VBB min of 8V

I looked at the output of the ATTiny on an oscilliscope and saw that the inputs to the A4953 were moving between 5.3V and 0V which exceeds the minimum range required for the H-bridge chip of 0.8 - 2V in the data sheet. At this point I thought it might be something to do with my redesign, therefore as a sanity check I fabricated and stuffed the initial design using the trace and cutout .png supplied by Neil. Then Pedro told me that VBB of the H-bridge needs to be at least 8V. After increasing the voltage from the power supply the motor started moving. And when I went back to my initial redesigned board, it worked too!

Lessons learned

• Read the data sheet for new chips and check what the operation voltage is and whether any external resistors or capacitors are required.
• If you're trying to remove a capacitor or resistor and you have two soldering irons then you can melt both pads simultaneously and change the component without having to use a heat gun.
• Label the V+ and GND lines clearly
• Still not quite sure how to attach pads on the bottom of chips

Notes from class feedback

• You can reflow solder paste using a range of methods includinga heat gun, reflow oven or frying pan.
• Wireless networks - On FTDI side you have Tx and Rx lines, get multiple ID'ed processors to listen to the Tx line and talk back through the Rx line. RS
• Stepper motors have 4 or 6 leads, two will be unconnected and the other two are connected??, apply a voltage and see what happens to figure out you should connect it into your circuit. You can short all the leads and try to turn the stepper it should be harder to turn, this is a result of back EMF generated and gives a good indication of output torque of the motor.
• Add LEDs onto power line and data line as standard for debugging
• For low noise boards, bring all grounds to a point rather than busing them to avoid currents flowing past each other. This is relevant when you're powering logic and motors on the same board.