loading
Picture of Electric Wurlitzharmonica
wurlitzer.png
This week I attended Music Hack Day at the Drexel Excite Center.  There were many incredible projects there to investigate, as well as some seriously funky live music and some cool folks, all coming together for a 24-hour musical hackathon. The creative atmosphere was pretty inspiring, and out of this was born a weird new hybrid musical instrument -- the Electric Wurlitzharmonica, or the Wur-monica, for short.

With the Wur-monica, I have found a novel way to amplify a harmonica.  The Wur-monica does not have a microphone or pickup of any kind inside. Instead, the Wur-monica relies on the electro-acoustic properties of the instrument itself as a signal source. In fact, the Wur-monica is based on the same elegant principle as my beloved Wurlitzer 200A electric piano (hence the name). 

How it works:
The original Wurlitzer Electric Pianos of the 1960s were gorgeous machines with a really interesting way of generating sound: instead of strings, the piano had large reeds struck by hammers (see picture). The reeds produced a signal using an ingenious electrostatic principle: the reeds were charged up to hundreds of volts, so that they formed a capacitor with the adjacent ground plate.  When the reeds vibrated, this capacitance changed, pushing electrons back and forth to magically produce ethereal honks, buzzes, crunches, and chimes. If you think about it, a condenser microphone also works in much the same way (except with parallel plates instead of vibrating reeds).  In a sense, each Wurlitzer piano reed is its own condenser microphone.

The Wur-monica works using the same principle as the Wurlitzer piano. Each brass reed inside the Wur-monica is mounted on a brass plate, but does not touch it.  This creates a very small capacitance between the reeds and the plate.  When a reed vibrates, this capacitance changes, creating a signal that can be amplified.  Again, no mics, pickups, or piezos are needed -- wires are connected directly to the reeds, so that the reeds themselves produce their own electrical signal.

The main advantage of this technique is that it is immune to feedback from the speakers, because the device does not pick up airborn sound, but rather the raw vibrations of the metal parts. Of course, I'm also hoping to maybe capture a little of that crunchy, magical, soulful Wurlitzer sound in a pocket-sized instrument.

This project is only a proof of concept, so I didn't spend much time refining the circuitry, the method for mounting the reeds, or the intonation, for that matter. In fact, I only bothered to mount seven of the reeds, and one of them is still kind of wonky and shorts out the circuit if I blow too hard (you can hear it shorting in the video). Still, after only a few hours of hacking, I was pretty pleased with how the prototype looks and sounds. 

Please check it out and leave me some feedback!  And if you like classic devices updated with new electronics, check out my USB Typewriter instructable, too. And please vote for this project in the Pocket Sized Contest!  Thanks!

First Test:


Final Test (just before getting dizzy and passing out):

 
Remove these adsRemove these ads by Signing Up

Step 1: Parts Needed

Picture of Parts Needed
IMG_0806.JPG
IMG_0815B.jpg
You will need:

Perf board
Insulated wire
Magnet wire
LM358 op-amp
Resistors (1MΩ, 2MΩ, 22kΩ, 10kΩ, and 4.7kΩ)
Potentiometer (100kΩ)
Capacitors (22nF and 1uF)
1/8 inch instrument jack
Power supply that produces both +48V and +12V (I used a Behringer Phantom Power Supply)
Harmonica (I used a Hohner "Bluesband")
An amp for testing (I used a computer speaker from RadioShack)
Tools and stuff

You will not need any magnets, microphones, piezos, guitar pickups, or anything like that to do this hack! That is what is so cool about this project: the harmonica will become its own microphone!

Step 2: Open up the Harmonica

Picture of Open up the Harmonica
IMG_0826.JPG
Open the harmonica and separate all the parts, saving all the screws for later. Carefully remove the two reed plates, noting their orientation.

Step 3: Remove the reeds

Picture of Remove the reeds
IMG_0821.JPG
The Wurl-monica works by detecting a change in voltage between the reeds and the plate. If any reed should ever make electrical contact with the plate, the voltage between the reeds and the plate will be zero, and no signal will be produced by the Wurl-monica. 

However, notice how the reed plates are designed -- each reed is riveted into the plate with a tiny steel nail.  This means that each reed is already electrically shorted to the plate!  That is bad news for our design. Therefore, we have to find a way to mount the reeds in such a way that they don't make electrical contact with the plate.

First off, we need to gently remove each reed.  I found the best way is to grind down the head of each little nail, then gently rotate each reed until it loosens and falls off.  Make sure you don't damage the reeds while doing this. Once the reed is off, grind down what is left of the nail so the surface is flush.

Step 4: Shave down each reed

Picture of Shave down each reed
Again, if the reeds ever make electrical contact with the plate, the instrument will not be able to produce sound. But notice that each reed is a very tight fit into the rectangular hole it sits in -- when it vibrates, it is sure to touch the walls of the hole. To prevent this from happening, I decided to shave down the reeds to make them slightly narrower, so they would fit in their rectangular slots without ever rubbing up against the walls.

I did this by holding them on their side against a Dremel wheel. I narrowed the long part of each one by about 10%.  This will probably mess up the tuning of the instrument, but I'm not so interested in that right now.

Step 5: Attach reeds and wires to plate

Picture of Attach reeds and wires to plate
I actually did this step out of order -- I glued each reed down first and then tried to solder magnet wires to them.  That might have worked if I had used high-temp epoxy, but I was in a hurry so I used Superglue gel, which melted under the heat of the soldering iron.  I got it to work using some finesse, but what I should have done was solder the wires first, and then glue the reeds down after the solder cooled. Ah well.  However you decide to do it, you need to attach one wire going connecting each of the reeds together, and another single wire that is soldered to the plate. Keep these wires as short as possible, to avoid picking up interference.

While perfoming this step, make sure you use a thick enough layer of glue that the reed is not directly touching the plate.  While the glue is drying, use a multimeter to check that they are not touching.  It helps if you grind the two services flat before gluing them together. 

The hardest part is making sure the reed stays lined up perfectly inside the rectangular hole while the glue dries. Remember -- it shouldn't ever touch the sides of the hole or else the sound will cut out.

Step 6: Build preamplifier circuit

Picture of Build preamplifier circuit
IMG_0877.JPG
The pre-amp circuit I designed is based in part on the schematic for my Wurlitzer 200A.  It is also based on a simple electret mic pre-amp schematic I found here.

The opamps used are LM385 but you can probably use any single-rail opamp.  They are supplied by the 12V rail.  The 48V rail is only used to charge up the capacitor formed by the reeds and the plate.

How the circuit works:
The 48V supply loads electrons into the capacitor formed by the reeds and plate.  When this capacitance changes, the amount of charge the capacitor can store at that voltage changes (C=Q/V), and so the electrons have to move.  They do so through the 22nF cap and into the input of the first opamp stage, which is a buffer amplifier.  The buffer is designed to have a very high input impedance, so that even a small trickle of current from the reeds can produce a good voltage. The signal then goes into a second opamp stage, which is an amplifier with a variable gain set by the 100kohm trim pot.   Its a very simple circuit, and probably not the best, but hey it worked first time so I don't complain.

Step 7: Reassemble and Rock Out

Picture of Reassemble and Rock Out
Even without reattaching any of the screws or the steel covers, you should be able to test the Wurl-monica by sandwiching the two reed plates onto the plastic mouthpiece with your hands. When you are happy with how it is working, permanently close the case back up with the screws and affix the amplifier circuit to the top of the harmonica with putty or Sugru.

The setup is like this: Right now, a Behringer Phantom power supply provides the 48V and 12V rails to the circuit, while a dinky Radio Shack computer speaker outputs the sound. But there is no reason the circuit could not be battery powered with a tiny built-in speaker.  I just didn't have the time or the parts to do all of that during the 24-hour Hackathon.

In this video you can hear that the sound cuts out occasionally -- that is because one of the reeds is not seated right and sometimes shorts out.  Otherwise the thing works surprisingly well.  It can produce some hum on some settings but there is a certain sweet spot on the trim pot dial where the amp is nice and quiet. The sound is not so different from a regular harmonica, although somehow more "robotic" and metallic, even when played on a clean setting. 

One disappointing thing I found out is that you can't do wah effects with your hands -- this is an acoustic effect and does not really change how the reeds vibrate, so the signal is unchanged. Well, maybe that is a good thing for some purposes but it would be nice to hear my invention with cool wah effects -- hey maybe I should plug it into a wah pedal!

Alas, another challenge I faced while trying this out is that, when exhaling repeatedly into the "blow reeds," your breath creates condensation on the reads which can short them out or change the capacitance. Therefore this idea is only practical for amplifying the "draw" reeds (the ones you inhale to play).  I have yet to think of a workaround for that problem, so maybe a different reed instrument like a harmonium or a melodica would be more suitable to this hack. Maybe I will try that at the next Music Hack Day, which I hope will be soon!

If you like this instructable, please vote for it in the Pocket Sized Contest. Thanks!

First Test:


Final Test (just before getting dizzy and passing out):

MrRogerBarry.7 months ago

Is your 1uf cap polarized ?

What wattage or you using for the resistors ?

nancyjohns1 year ago

this is awesome but you cant really hold it right with all the stuff on the top. I play the harmonica and this would be cool to have :)

wow, this is the most beautiful thing.

Thankyou so much

J-Five1 year ago
Cool, also I couldn't even pronounce the name it.
OneHarp2 years ago
This is really really really really really awesome. I am a harmonica player and an electronics tinkerer, and I would love to beta test this for you. I can think of two suggestions right now: 1) you'll get better results using a better quality harmonica. The reeds are better (stiffer) and you won't need to reshape them to keep them in their slot. I'd recommend a Hohner Special 20 harmonica. 2) There is an easier, less damaging way to remove the reeds: file down the nub of a set of grometing pliers until it's the same diameter as the rivets holding the reeds on. Then, you can just pop the rivets out. I do this all the time, as the reeds on harmonicas eventually fail due to stress fracturing, and need to be replaced. I can show you a picture of the tool I use (made from a cheap set of grometing pliers from harbor frieght).

I have a question, though: is the amplified note the same as the one the reed is playing? i.e., does the capacitance change of the vibrating reed have any relationship to it's vibration frequency?

Seriously, I want to beta test this for you. Check my page for my instructables relating to harmonica, and links to my youtube so you can hear my amped harmonica playing. I think this has the possibility of revolutionizing the harmonica world. To my knowledge, there are only three other attempts at doing something like this: one using small piezo pick ups, one using laser pickups, and one using magnetic pickups. All of them are bulky and expensive, and your solution here looks to be simple, cheap, and effective. I'm down to help, so let me know!
jackzylkin (author)  OneHarp2 years ago
Thanks -- I did think about the plastic screws and spacers, but I didn't see an easy way to remove the pins -- thanks for the suggestions! However the real drawback of this design is that condensation forms on the reed plate and shorts out the voltages. This only occurs when you blow. I believe heating the plates to above body temperature may solve this problem -- but may cause others.
OneHarp OneHarp2 years ago
Also, instead of superglue, it's probably better to drill the holes out, tap for 0-80 screws, and attach them using plastic screws. Put some sort of insulating material under the reeds where they attach to the comb (probably nail polish or dried superglue would work). Then you, won't have the issue of the reed not lining up once the glue has dried.
OneHarp2 years ago
Hi there again... I was thinking about this some more overnight, and did some reading up on the principle at work here, and on different electrostatic organs. Reading about the Hohner Pianet (https://en.wikipedia.org/wiki/Pianet) and taking off from the comment made by rimar2000, below, I suggest that a cool thing to try out would be to mount a strip of copper to the inside of the coverplate, above the reeds crossing them all perpendicularly at their tips, and use that as the charged capacitor. You could use the superglue trick to isolate it from the metal of the coverplate (or actually you could use a cheap Piedmont Blues harp, which has plastic covers). This seems to be how they do it in the Pianet's from the 60's that apparently had a great tone...
I may give this a try when I have some free time next week, as I actually have a small powered preamp I use for a Sony electret condenser mic, which may work for this. Those type of mics uses the same capacitance principle as this pickup, but I'm not sure if there's enough voltage in my power supply (it uses a 9volt battery) to power this sufficiently... I also have several "junk" harps laying about, and some extra copper reed-plate material that I could use for the charged plate... I'll let you know what I come up with!
degomag2 years ago
Genius! We will have to invent a new genere of music for this!
rimar20002 years ago
Great idea!

But I think you don't need to disassemble each reed. You could have all them as ground, and a common bus, isolated from the plate, as captor.
jackzylkin (author)  rimar20002 years ago
That may well be true. It would increase the capacitive load on the reeds, but probably not enough to weaken the signal. Great idea-- let me try that next!
jackzylkin (author)  jackzylkin2 years ago
It may also be that the plate might "shield" the reeds from influencing the captor, since in general the plate will be much closer to the captor than the vibrating reeds. I wonder what the waveform would be if there were such a captor on both sides of the plate.
really interesting invention man! congratulations!
jackzylkin (author)  knarrenheinz2 years ago
Thanks!
caitlinsdad2 years ago
Is there any shock potential here? Kissing a mike and getting jolted is no fun.
jackzylkin (author)  caitlinsdad2 years ago
The 2Mohm resistor should prevent a painful shock from the power supply. And the capacitance of the reed bar is probably in the picofarads, so it will not shock you.