Please see this link for my group's assignment (session 1).
This week, I started making some incremental progress towards my final project. The idea is to build a modular synthesizer out of found electrical components, which teaches the user how to probe for different kinds of signals. A speaker will provide auditory feedback, which will enable the user to hear the different kind of output signals, while a screen will show them visually what the different signals look like. My goal for this week was to design a PCB using a microcontroller to generate sound.
As a starting point, I wanted to fix my board from Week 6. This board simply feeds the output of thee SAMD21's DAC through a variable low pass filter. For some reason, I was not able to demonstrate a frequency response. I showed the board to Neil in class, and he was able to identify the issue rather quickly. I had grounded the third leg of my potentiometer, turning it into a voltage divider as opposed to a variable resistor. Cut the trace marked by the white arrow fixed this problem. The screenshot of my logic analyzer shows the input signal generated by the DAC on channel 0 and the output of the low pass filter on channel 1.
In order for this signal to be playable through a speaker, I would need to amplify it first. There are many different ways to amplify audio signals, as discussed in class. I decided to check on the inventory in the electronics lab to see which of the chips were available. The first one I spotted was the LM4871M Boomer, a 3W Audio Power Amplifier. Taking a quick look at the datasheet showed this to be a rather straightforward circuit, so I decided to go with it!
I duplicated the typical application circuit above, making sure to utilize the RC values we had available. For instance, since we didn't have any 0.39 uF capacitors on hand, I placed 4 0.1 uF capacitors in parallel to achieve the same effect.
I am getting more comfortable, moving through the workflow from schematic, to PCB layout, and through the Carvera milling process. That being said, one issue I had with my board is that the carvra did not mill out the through holes in my design. I believe this is because I did not export those features properly. I plan to work through this issue on my next board. For this week, I decided to make this board work through a bit of arts and crafts, using an exacto blade to make larger pads to which I could solder my input and output leads.
Alas, unfortunately, the board did not work at first. I probed around using an analog/digital logic analyzer. It appeared as though my input signal was making it through to the output pins of the amplifier IC, but the output did not look right (see below). Channel 0 is the output of the DAC, Channel 1 is the output of the RC circuit, Channels 2 and 3 are the LM4871M's two outputs. I was working with TA Alan on this during Friday's office hours when Neil popped by. He suggested that I use an H-bridge instead, following this example. from the HTMAA website.
At first, I did not see a schematic or footprint posted for this example, so I did my best to reverse engineer things based on: 1. a screenshot from the example video, 2. the the TB67H451 Datasheet and 3. the example code.
It was only after I went all the way through the milling process that TA Anthony pointed out to me that the foot print of the board in the example is shown under the H-bridge section of the output devices page. Alas, this was good practice with reverse engineering! The only major difference between my board and the example board is that I used D0 and D1 for my outputs where as the example uses D6 and D7. I was hoping that after updating the output pins in the code, this would simply work! Silly me...
It took quite a bit of time, but ultimately, we figured out the issue. I still need to upload a few more images and videos. But, huge shout out to Anthony for spending HOURS with me. I learned a lot and had so much fun!