Adding a 5th input for a Piezo kick pedal…

[QKillerQ]

Hi!

So, I built a kick pedal with a piezo that acts the same as the drum sensors in your circuit. I can plug it into any of the drum pad units and it triggers perfectly well.

I'd LOVE to be able to connect my kick pedal as an additional sensor/trigger so I can play Clone Hero with a bass drum.

I have tried adding an additional input to the XAIO board directly to a GPIO and I was having issues with that. (I wish it had ONE more ADC pin, lol). I also tried coding it so that I could use one of the spare pins on the i2c expansion board as well, But I keep running into issues. It's almost as if there's not a single pin free to properly add this.

I don't need anything crazy I just need something that converts the Piezo after a certain voltage to a digital button press - and it would be even fine just assigning that button press the spacebar or a single button.

If it's a big deal, disregard, but if it's just a few lines of code that could be altered I would appreciate that if it was possible.

Like I said, I tried several things and the button works to the point that continuity at the board will trigger when I press the button so I know the Physical connections are right. It's just something with the software.

Than you in advance!!

If you'd like me to send you what I've tried I can do that as well.

Also, I'm more than willing to help test it out if needed.

[ravinrabbid]

Hi,

sorry, I'm a bit confused where exactly you got stuck.

I have tried adding an additional input to the XAIO board directly to a GPIO and I was having issues with that [...]

So you are using a normal IO (i.e. non ADC) pin? Have you checked that the voltage consistently crosses the threshold voltage for a logical high when hit? How did you connect the sensor, did you use any additional components? How do you try to read the input pin in software?

Like I said, I tried several things and the button works to the point that continuity at the board will trigger when I press the button so I know the Physical connections are right. It's just something with the software.

So you tried with a button instead of a piezo and it works? Then it should not be a software issue.

Cheers

[QKillerQ]

Hi!

Thanks for the quick response. Let me answer your questions:

Pin type: Yes, I'm using GPB7 on the MCP23017 I2C GPIO expander (digital pin), not an ADC pin on the XIAO (although I did try several GPIO pins on the XIAO, even though most pins seem to be used already in the circuit).

Voltage measurement: I haven't measured the actual voltage output from the piezo circuit yet - that's likely the key issue. The piezo circuit works perfectly when connected to the ADC inputs on the existing drum sensors, but I suspect the voltage may not be reaching the logic high threshold (typically ~2.4V) needed for reliable digital detection.

Circuit: I'm using the same protection circuit as the working drum sensors in the hybrid circuit that I built.

• Piezo+ → Diode (cathode toward ground) → GPB7
• Capacitor between pin and GND
• Piezo- → GND

Software: Reading through MCP23017 I2C in the controller processing loop, similar to how the navigation buttons work.

Additional findings:

• Physical continuity tests pass
• Code compiles without errors
• No detection in debug mode or gamepad testing

I suspect the issue is that the piezo circuit's voltage output (designed for ADC) isn't sufficient for reliable digital GPIO detection on the MCP23017. Would you recommend:

1. Adding a comparator circuit (LM393) to convert the analog piezo signal to clean digital?
2. Using an I2C ADC expander (ADS1115) instead for a 5th analog input?
3. Is there any other approach I'm missing?

The kick pedal definitely works - it triggers perfectly when connected to any of the existing drum ADC inputs.

I just remembered, I DID try a microswitch at first - However, it was just a quick splice into one of the navigation buttons (the SHARE button in TATACON mode) for proof of concept, and it worked just fine.

I quickly realized that I now wanted to make a piezo sensor for quicker response time and rapid-fire hits with the pedal. In addition to the nicer pedal, I wanted it so that the trigger wasn't simply spliced an existing navigation buttons.... I wanted to give it its own, dedicated pin as well as make an entry into the firmware to turn the kicker pedal on and off (just because).

If it helps, with the PIEZO connected, I do get continuity whenever I tap the sensor and that continuity comes right at the pin on the board that I'm connected to so I know it's getting there. It's just not flowing past it.

So, I guess that's tells us that the piezo sensor may not be putting enough for the software to read. BUT, like I mentioned above, if I'm getting a continuity at the board and ground when I tap my pedal
Piezo, I would think that's enough to signal the board.

Perhaps there are multiple variables at play here. The piezo sensor works, but I now have something incorrect in the software for the dedicated pin.

Do you see any reason why pin A7 or B7 (both unused) on the expansion board wouldn't be allowed to work if my code changes were correct?

Thanks again for your help!

[ravinrabbid]

That makes thinks a bit more clear, thanks :)

A few unsorted thoughts on this:

• Using any IO Pin (Internal or External) likely not only has the problem of the voltage being too low, but also the pulse length being too short. IOs are read with the same frequency as USB packets are sent (around ~1000Hz), while this is more than enough for manually operated buttons, to detect the peaks on an audio signal (which a piezo produces in the end), this is most likely too slow.
• Keep in mind that the MCP23017 runs on the second core, which I used for low priority tasks. So this could lead to delays/dropped inputs (Maybe try to play with the buttons attached there first, to see if there are any issues).
• Have you tried removing the capacitor from you circuit? The piezo produces a very low power signal, so the capacitor could considerably lower the voltage.
• Using a comparator might still yield the issue of the pulse length being too short, so you would also need some kind of sample-and-hold.
• Using an external ADC is probably the cleanest way to solve this, keep in mind that the ADS1115 only has a sample rate of 860sps, which is lower than the targeted USB packet rate, so you might to use something a bit faster.
• You can also have a look a piezo trigger circuits used for e-drums, but at this point you are already pretty close to what is on the DonConIO.
• Measuring continuity over the piezo doesn't tell you much here. The piezo will produce a voltage when hit which probably somehow falsely triggers the continuity of your multimeter. What you need is the peak voltage (and the pulse lenght), which might be hard to measure without a proper oscilloscope.
• Using the unused pins of the MCP23017 should not be an issue, just connect a simple switch and you should see if your code changes work.

[QKillerQ]

Thanks for all of that random info - it made perfect sense to me.

For more clarity, as a Metal Drummer. My foot really needs the full action of a bass drum kick pedal along with the matching mallet resistance and feedback.

That's why I used to used a Roland KD7 kick pad to play rhythm games. Picture attached:

However, since the KD7 is just a piezo sensor at its core, I wanted to make my own to save money.

That's why I have my heart set on the Piezo route - that would make this prefect perfect. I suppose I could get another KD7 and purchase an aftermarket analog to usb dongle and call it a day. But that would be $200+ just for that pedal and conversion. lol.

I also, since I have so many spare boards and parts - to include another XIAO 2040 lying around, thought about simply building another Doncon2040 XIAO circuit with JUST the Piezo pedal on one of the ADC pins since we know that circuitry works perfectly.

It would have the trigger set to send the SPACE keyboard key - and I would just connect that 2nd board to the computer via USB as a separate controller.

I'm sure I could probably figure out how to connect the 2nd board to the first, main board via UART or something, but that's too much of a hassle for sure.

The question would now be - could your firmware be trimmed down to

A) not require the display
B) not require any navigation buttons.
C) the key designation and threshold could then simply be set in the VS code before flashing - since that's something that won't need to be adjusted on the fly often, if at all, like the drum pads on the main circuit are.

I believe I remember you mentioning that a line of code can be altered/removed to get rid of the navigation keys dependency...

Just thinking.

This is fun!

[ravinrabbid]

Simply building a second controller is a astonishingly simple solution which would have never occurred to me xD

I think for A) and B) you don't need to do anything since those things run on the second core and won't interfere if they lock-up. You can remove

DonCon2040/src/main.cpp

Line 141 in 94e4911

multicore_launch_core1(core1_task);

to be on the safe side.

For C) just make sure to clear the flash, if there is no menu, nothing should ever be saved and the values will always be taken from GlobalConfiguration.h

Other than that you might need to silence the unconnected inputs from wildly triggering, either by grounding them or ignoring them im code.

[QKillerQ]

Got it!!

https://youtube.com/shorts/TfB4RjUUsTk?si=MLzDqS9FjnwCJoEg