A stompbox with built-in spring reverb

If anyone hasn’t noticed yet reading this blog, I and my wife are amateur musicians. We’ve also released a short album that we’ve entirely written and recorded at home. For the most adventurous among you, it’s publicly available here. It features also the Fork-o-Drumbot in one song!

So now we’d like to put up a show made only by the two of us. But we need additional instruments to give us the groove for the live performance. What’s better than a stompbox then? Well, a self-built one!

Spoiler: if you want to know how this works, go on reading!

Spoiler: if you want to know how this works, go on reading!

This is the final result. If you want to know why I think this is a bit different than the other stompbox projects, just keep reading after the video.

/// insert video


I used a piezo pickup. There are many piezo amplifiers designs out there in the internet. Some claim you can simply solder the audio cable to the pickup. That would kill all the bass frequencies because the cable will act like a resistance and the piezo has a significative capacitance = high pass filter.
Discarding op-amp design, which may be great but add complexity, I stumbled into J-FET preamps like this one. So I decided to modify it for my needs.
Here is the final circuit:

Stompbox schematic

Stompbox schematic

It is pretty simple. My target was to keep it simple but still have a good sound and a volume knob (I want to connect it to an amplifier for bass guitar).

I had to buy an N-channel J-FET, so I went for an audio-grade one, a 2SK117. The 1M logaritmic potentiometer will show always the same (high) impedance to the piezo pickup, while reducing the amplitude at the gate of the J-FET. J-FET are great because they are bipolar, you can polarize them both ways. The reason why source is not connected to the lower potential, as per convention, is that I have discovered that I have mounted it that way, and it works great 🙂

I played with resistor values until the output (positive terminal of C2) was almost half the voltage of the 9V supply. D1 and D2 where a late addition, and they work as a compressor (or better, a chopper), it will be better explained later.

It’s an history of a few fails steps.

#1 Order components and mount circuit on the breadboard to find proper values for resistors etc.

#2 The circuit works! Now we need an enclosure. A shielded one. No wood around (or patience of waiting for a laser cut wooden box)… Cardboard!

I had a very stiff cardboard box sitting around:

Cardboard box

Victim cardboard box

I glued tin foil for shielding on the inner part of the opened box

Cardboard box with tin foil

Inner side of box shielded with tin foil

After that, I closed the box again, cut the excess tin foil, cut a hole for gluing the piezo element and hot glued all the components in place. Oh, I have also soldered the circuit on a piece of perfboard.

Finished cardboard box

Finished cardboard box. The battery is hold in place by a magnet, and the blue wire under the hot glue blob in the bottom right corner connects GND to the shield.

#3 Play with it! Craccracriccrecr… (fail 1)
I must say, the shielding worked really well, I was hearing almost no hum from the amplifier. There were a couple of problem though. One has to do with paperboard, and one with tin foil. Some may argue, you should not use cardboard for something you stomp on all the time. Indeed the cardboard was flexing, the tin foil crackling, and all that crackling was picked up by the piezo element and amplified…

#4 Ok let’s start again, let’s build a proper box
There’s a good laser cutting service in my city, so I decided that was the way to go. I designed a box made from 5mm thick MDF. I am not very expert in this.

  • Slightly longer than my foot, for letting me rest the whole foot on it;
  • Sturdy construction, especially in the heel area where most of the weight goes;
  • Fancy drawer for the 9V battery;
  • Easy inspection (top board fixed with screws and nuts in t-slots).

I have drawn the box as I imagined it in SketchUp:

SketchUp model of closed box

SketchUp model of closed box. I tried to make the “heel” side more robust with a few cross-elements, while leaving the “toe” side more free, like a resonant box.

SketchUp model of open box. You can see the 9V battery drawer and the T-slots for the square nuts.

SketchUp model of open box. You can see the 9V battery drawer with the round cut for the magnet to keep the battey in place, the T-slots for the square nuts and the holes for the volume knob and the output jack. The lower part is all press-fit, while the upper board is fixed with screws, for easy removal.

#5 Create the cut layout, send it for cut, wait and… Assemble!
The laser burns away material as it cuts, so it is necessary to design all the “fingers” of the joints slightly bigger. The holes must be slightly smaller as well, or you’ll have to to use glue for the assembling. I documented myself, I even found some tables about the amount of material cut away and I applied the correction. Too much correction (Fail 2). Or better, I didn’t take into account that you don’t need this compensation in the direction of the thickness of the MDF board, as this is constant (it’s not cut by the laser!). The parts didn’t fit, even using a hammer. It’s been a long work of file (I have made so many joints fingers…) and hammer. Eventually I destroyed the battery holder, because it was too weak to resist my impatience! (Fail 3)



Here you can see how it was before closing it. I have shielded almost everything by wrapping it in tin foil.

Open box

Open box

#6 Play with it! (Again)

Closed box. I have added a spiral pattern on the top panel :D

Closed box. I have added a cool spiral pattern on the top panel. I probably have a weird passion for noisy wooden boxes 😀

The box is terribly noisy, the 50Hz hum is very strong. That’s maybe because I haven’t shielded the green wire in the picture? Yes! (Fail 4) 😦

Eventually, I have enhanced the shielding by wrapping everything in tin foil, pots included, but the tin foil is really fragile. At the end I bought a proper shielded cable and sent the noise (almost) away. But still, I was not very satisfied about the sound. It sounded more like a click, instead of the punch of a kick drum! But as I have heard in many videos in Youtube, it’s a common feature of stomp-boxes with piezo pick-up.

#7 Tuning the sound (spring reverb):
The piezo pick-up is a contact microphone. That means that, unlike a microphone in a box, it doesnt pick up the vibration of the air inside the box, rather it picks up the flection of the panel itself. That is a pretty dry sound, because it is not caused by the vibration of the air inside the box volume, neither by waves travelling through the large membrane of a drum. It’s simply the flection of the panel, returning quickly to its original position.
Now, for adding a reverb effect to the sound of the pickup we have basically three ways:

  • a digital reverb (but I don’t want complex circuitry, digital signals nor Arduinos in my stompbox);
  • an analog reverb ic, also called a delay line or a bucket-brigade device (at the beginning I thought it was the most viable solution, later I realized that as I was growing old those devices went out of production and became vintage and expensive!)
  • a spring reverb, an electro-mechanical device used in old-style guitar amplifiers.

A spring reverb is too of a complex beast for my stompbox (I have stolen the image from this project):

Slinkyspring Reverb

There is a power amplifier which drives the drive speaker. The movement of the drive speakers induces vibration in the spring, and this vibration is picked up by other pick-ups, with additional wobbling which sounds like echo (or water, most of the times).
After being disappointed by this situation for a few hours, I realized I already had everything I needed:

  • I didn’t need a power amplifier, because my foot is powerful enough and the signal I need to “process” is already available exactly at the piezo element position;
  • I didn’t need any other pickup for the waves in the spring: the piezo pickup transduces its own traction and compression, so it’s enough to apply the wobbling of a spring on the other side of the element.

This is what I came out with:

My spring reverb

My spring reverb

I have soldered the spring of a pen on the back face of the piezo element. That is a very rigid connection, so it’s perfect for transmitting the vibration. Then I have attached the stompbox to the amplifier and I have started knocking on the other side, to hear what sound it was producing.
The cool part, is that you can tune the sound of the reverb by changing the mechanical parameters of the system (yes, once you have closed the box, it’s done! So it’s better to take some care in this step):

  • Stretching the spring or adding a weight at the end (the small bolt) of it lowers the resonance frequency, thus the pitch of the sound. The other way for increasing the pitch;
  • Adding dampening elements, like inserting pieces of cloth or Kleenex in the spring, shortens the time of the reverberation.

It’s a short and rigid spring, so the sound doesn’t have many feedbacks, but I think it’s perfect for what I need. A longer, softer spring would have allowed longer and more articulated reverb, but it would have also been too sensitive to movements of the box. I am satisfied of the sound, hear the difference with and without reverb:

In this recording there are some taps recorded holding the spring, and some with the spring left free. You can’t hear any difference with laptop loudspeaker,  you need to hear low frequencies with headphones or big speakers, and the effect is very subtle. I have tuned the sound a bit more after this recording and I have to say, with the bass amplifier it sounds really cool!

#8 Tuning the sound (compressor)
Another thing that I have modified after this recording is adding compression to the sound. (Dynamic range) compression is the process of (softly) limiting the amplitude of the sound with the aim of letting the details of the sound audible. This is the waveform of the first five knocks of the audio file above:

Stompbox sound waveform

Stompbox sound waveform

There is a first very loud peak and after that the rest of the sound (the reverb) is very low. So I must keep the volume of the amplifier low enough to not distort, and I won’t hear anything else. But if I “cut” the sound where the two red lines are, I can use more volume thus enjoy more reverb.
Also here, I went for the easy and cheap solution. There are a lot of compression circuits out there, but you can do this also by simply adding two 1n4148 diodes in parallel to the output (see the schematic above), one for limiting the positive peak, one for the negative one. 1n4148 start to conduct around 0,7V, with a soft knee, and then they act like a resistor:

1n4148 characterstic

1n4148 characterstic

That means that after 0,7V of amplitude the sound is gently reduced, allowing me to use more gain on the external amplifier.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s