Danny Griffin HTMAA 2023

Week 09 :: Output Devices

How do we send haptic feedback?

1. Add an output device to a microcontroller board you've designed.

And program it to do something.

This week, I am designing a board to connect an ESP32C3 microcontroller to a 10mm vibrating coin. The coin is a DC motor, which I will try to connect to a GPIO pin.

To understand how to connect the vibration motor, I found this example helpful: Reference circuit.

Revisiting the board that I designed during Electronics Production week, I am removing extraneous elements so that I can focus on incorporating the DC motor.

In this iteration of my board, I'm trying to space the traces farther apart to give myself more room to work.

was having significant trouble connecting anything to the board - and as I kept heating the same locations, the copper layer began to delaminate. In the end, I resorted to heavyhanded soldering to try to connect things together. In retrospect, this board had many faults with the traces, and it would have been simpler to re-mill a clean board.

Eventually, I was able to use my wired connections to begin programming and testing the vibration motor.

Something is wrong. With the microcontroller plugged in, the vibration motor is powered. This should not be the case, and it means I've made some mistake in surface mounting certain elements - possibly the mosfet, or the diode, or some of the resistors?

Looking at the board at this point, everything is a mess so it's hard to begin diagnosing the problem. I began probing around the board with a voltmeter to see if any components were causing issues.

I took a guess that perhaps I misunderstood the circuit, and the anode of the diode was the wrong way around. I reversed my diode, plugged the microcontroller back in, and immediately, the diode heated up, turned bright orange, and started smoking. Reversing the diode was the wrong move, and this was the point where I decided this board was no longer worth salvaging.

Niklas Hagemann introduced me to these sockets, which could be used to test the board without soldering the microcontroller itself.

Excerpt from the datasheet for these sockets, which show how the little feet at the bottom stagger in different directions for the footprint.

Working more systematically this time, I remilled a board with cleaner traces.

Stuffing the board was slightly different this time, as I used sockets rather than surface mounting where I could.

I learned that it's helpful to isolate the vibration farther away from the board, because the vibrations tend to dislodge delicate connections and make troubleshooting difficult.

In the end, I had heated and replaced the elements in this board so many times that the traces began to detach from the board.

A short summary of issues I ran into:
I should have used a Schottky Diode, rather than a Zener Diode.
At different points, I read different information about where resistors should be located, and what resistance level they should provide. Any resistor or capacitor that I connected to the ground pin would cause the mosfet to fail.
This circuit was designed for an N-Mosfet, but I wasn't completely sure what specification was important.

Finally, after systematically replacing every component in the board, I was able to confirm the correct components for my intended use.

2. Group Assignment: measure the power consumption of an output device.

Working withNasibe Nur Dundar Arifoglu, we tried to learn a bit more about the performance anf requirements of the motor I was working on.

Out of curiosity, we tried using the voltage generator to see if the motor would vibrate more strongly with correspondingly higher power.

The verdict: yes, a 10 volt vibration is much stronger than a 2.9 volt vibration. This exercise was actually helpful to realize that 3.3 volts is plenty for powering this type of motor.