My failed attempt to design and build a stringless electronic guitar

Hi all,

In the 1970s I designed and built a stringless electronic guitar.
While I had no problem designing and building the mechanical parts of the guitar, I was overwhelmed by the electronic part of the design.
I had purchased all the electronic components of a Schober Theater Organ with thirty voices and tried to control them from the body of an acoustic Guild sized guitar.
So, the electronics inside the guitar amounted to a massive collection of mostly electronic switches.
When I ran into problems with the electronic circuits, I could not afford to hire an engineer to help me, and the instrument never got finished.

If there are any members who would like to see pictures or learn more, I would continue this thread.
Since I don't want to finish the instrument and I don't want it to end up in a land-fill, I would like some advice as to how to proceed.

The choices are: Donate or sell it to a museum,
Donate or sell it to a young individual with electronic and guitar knowledge, to finish it,

I am now involved with tangential tone arm design with minor, manageable electronics.

Sincerely,

Ralf
 
  • Like
Reactions: 1 users
Even I was thinking of such a violin, as I find it hard to shift strings all the time. Since I need only one string at a time, it'd be really nice if some electronics could detect the string position of my finger(s) and play that one automatically. I've however, never been deeply into it, as I strongly feel that it may take a lot of my time.

Could you please share more about the fundas of sound production without using strings / plucking ? Today, such a thing would be based on capacitive touch-sensing but since we didn't have a lot of that in the 1970s, so your technique might have been very different.
'
Thanks.
 
Yes, but I'm looking for something that could sense touch on a fingerboard and automatically generate the appropriate sound for the finger position. Note that the instrument would still be played manually, only that there wouldn't be any actual 'strings'.

I wrote here thinking I could get some clues from the OP's 'stringless guitar'. Thanks anyway.
 
  • Like
Reactions: 1 user
Mood had a "ribbon" controller, a stretched conductive ribbon that could be touched along the length of I assume a long thick film resistor. I assume it simply produced a voltage, which then drove the frequency of a VCO. I've never seen anyone take 4-6 of these, arrange them across a guitar fretboard and make a guitar inspired controller that way.

Of course, you still need the something to pluck part, so that the VCO frequency's tone is gated by some kind of right hand action. I suppose that could even be 6 keys like a piano keyboard, if you wanted to really depart from "strumming".
 
  • Like
Reactions: 1 user
Sounds like an Arduino project, a modern version of the multi-tone music programs we did with 8-bit micros in the 80's. You just have to design a touch interface for the "strings" and do the rest in software. There are tiny ~Arduino boards today that will fit in the existing space in an electric guitar. You may want to provide a midi output, which is just a serial port at 31,250 bps async. Those multi-tone music programs were wave table DDS with a different step value for each note. You will have to program the ~sustain etc. Todays, uPs are much faster so you should be able to get great sound and they have built in DACs.
 
Mood had a "ribbon" controller, a stretched conductive ribbon that could be touched along the length of I assume a long thick film resistor. I assume it simply produced a voltage, which then drove the frequency of a VCO. I've never seen anyone take 4-6 of these, arrange them across a guitar fretboard and make a guitar inspired controller that way.
The Moog ribbon controller uses a long "force sensitive resistor" strip. It is like a pot that you work with your finger. There are several kinds. The type needed for a ribbon controller is usually called a linear soft pot. It has no contact between the "wiper" and the resistive element until it is touched somewhere along the strip at which the wiper contacts the resistive track. In a music application there will be a voltage gradient along the strip. The 'wiper" is connected to some simple electronics which will provide a variable pitch control voltage dependent on where the strip is pressed and an on-off "gate" signal. The pitch voltage is stored in a "sample and hold" circuit when the force (finger) is removed from the strip so that the pitch information is held constant. The gate signal however disappears. The control voltage (CV) and gate signals can be used with any of the popular analog synthesizer circuits on the market today.

https://store.spectrasymbol.com/collections/standard-softpot-potentiometers

There is a neat musical application of one by the marble machine guy in a device he calls the Modulin. Modular synthesizer meets violin:


Several years ago I saw a video by a guy who put three of these next to each other on an unfretted guitar neck. They were wired up to a box full of electronics resulting in a very playable stringless guitar. I searched youtube this morning but could not find it. It looked a lot like a prototype version of the Volt-Axxe.

I bought three of the Spectra Symbol strips after seeing that video, played with them for a while, then lost interest. One got used a couple years ago after watching another video that I can no longer find. My crude version incorporates a soft pot strip on a wood 2 X 2 wrapped in a towel connected to some DIY electronics. This generates drum sounds when whacked with drum sticks that change pitch as you move along the wood stick.


Sounds like an Arduino project, a modern version of the multi-tone music programs we did with 8-bit micros in the 80's. You just have to design a touch interface for the "strings" and do the rest in software. There are tiny ~Arduino boards today that will fit in the existing space in an electric guitar. You may want to provide a midi output, which is just a serial port at 31,250 bps async. Those multi-tone music programs were wave table DDS with a different step value for each note. You will have to program the ~sustain etc. Todays, uPs are much faster so you should be able to get great sound and they have built in DACs.
Look at the Arduino compatible Teensy modules from PJRC. The 4.X versions are powerful 32 bit chips supported by an audio library that makes DIY music synthesizers and MIDI controllers almost easy. I currently have three working synthesizers based on the T3.6 and T4.1 modules.

https://www.pjrc.com/
 
  • Like
Reactions: 1 users
Netlist:

Thank you, Netlist, the Volt-Axxe looks really close to what I imagined. Also, since I was thinking violin, one ribbon (returning X and Y) would easily tell me what string I am on, and play that one appropriately. This should work because each string is the fifth of the previous one, making all of them mathematically related to each other. However, as mentioned by @steveu, I would prefer digital (with good resolution) as that would be more convenient, flexible and more appropriate for the era in which we're now.

Nevertheless, it would still be nice to know the things the OP did in the 1970s.

jjasniew and Tubelab_com:

Thank you, that video shows exactly what I imagined though I think capacitive touchscreen and pressure sensing could also possibly be used. In my opinion, the sensing technology needs to be decided to:

a) withstand debris like dust, skin cells, sweat etc.
b) withstand application of oil, water etc.
c) achieve a long lifetime.

Several years ago I saw a video by a guy who put three of these ..... searched youtube this morning but could not find it.
That sounds like Intellectual Property (IP) issues.
 
No, it was a case of foggy memory. It took some Googling to find the information on the type of force sensitive resistors needed to make a ribbon controller. The kind you get at Adafruit do not work. Once I locked the word "softpot" back into my brain a YouTube search for "softpot guitar" returned it as the first hit. "Softpot" means something else to those of us who have designed hardware for 60+ years. A softpot is a resistor string inside a chip whose value or tap point are chosen by SPI or I2C command.

This is the one that I remember. All of the build details are in the links.


In a slightly different vein, I have one of the few AxeOrgans ever made. Built in the mid 70's it uses typical guitar hardware except for the fretboard which is actually a PC board. The frets are plated up nickel over copper on the front and gapped for individual string contact. The controller / organ used TTL logic as a suitable microcontroller had not yet been invented. When the strings contact one or more frets some logic chips decide which notes to play on a typical top octave generator type organ. The guitar strings remain fully functional so the instrument actually has two separate outputs. The organ circuitry has a pedal such that the typical back and forth rocking controls volume while side to side motion activated vibrato or reverb. Sometime in the 80's I hacked the circuitry to turn the organ into a string ensemble. The PC board / fretboard was patented back in the 70's but it is long expired now. I played this thing through the 70's and early 80's when one of the TO-3 regulators failed putting 12 volts into the 5 volt logic board. It's been in its case ever since and the pedal was lost in a move.

My latest attempt at a wild guitar synth uses a standard fretboard with a connection for each fret and 6 conventional strings that are electrically isolated from each other. Note decoding is done in a Teensy module. I'm still working on getting a VCO to track pitch bends on a string.

Hint: those tiny little dynamic speakers that make the beeping sound in a PC will work in reverse as a single string guitar pickup. The piezoelectric ones are useless and the dynamic ones work better if you take them apart and remove the diaphragm.
 

Attachments

  • P4010681.JPG
    P4010681.JPG
    368.5 KB · Views: 39
  • P4010682.JPG
    P4010682.JPG
    340.6 KB · Views: 38
  • P4010683.JPG
    P4010683.JPG
    408.5 KB · Views: 41
  • P4010684.JPG
    P4010684.JPG
    263.1 KB · Views: 39
Last edited:
  • Like
Reactions: 1 user
This is the one that I remember. All of the build details are in the links.

That was wonderful, thank you ! It probably also means that there's no IP issue related to this.

You just have to design a touch interface for the "strings" and do the rest in software....You will have to program the ~sustain etc.
I guess the following would be how one would go about next, in order to mimic the calibration of a string instrument.

https://en.wikipedia.org/wiki/String_vibration#Frequency_of_the_wave
https://en.wikipedia.org/wiki/String_vibration#Real-world_example

The frequency could then be directly generated using trignometric functions, at a decent sampling rate. The effects processing could also be done outside, like with regular instruments / vocals.

(y)
 
Last edited:
I like the ribbon synth in that video! I assume he's triggering with the 3 pads, down by the "bridge"?

One problem is how to get nuance, like that available on a real guitar. I see he can do glissando, i assume vibrato using the touch board. The triggering part; how to get different tone or harmonic content depending on how / where you "pluck" or "strike" some physical feature? How to make it go louder depending on how hard? How to cut the sound short via a "mute" gesture? It's almost like the note selection methodology is easy compared to a nuanced, expressive instrument.
Hint: those tiny little dynamic speakers that make the beeping sound in a PC will work in reverse as a single string guitar pickup. The piezoelectric ones are useless and the dynamic ones work better if you take them apart and remove the diaphragm.
Just picked a pair of DAEX9CT-4s, for dynamic pickup use. I've found the dynamic pickup sound is like going from an xtal to mag phono cartridge! Almost small enough to fit between strings in a row at the saddle. Staggered would do it. I wasnt thinking non-peizo hexaphonic pickup for nylon string guitar when I got just two, but maybe...how to couple to the string; felt pad? I've found such devices make great soundboard pickups, but I havent tried this particular one in that position - yet.

A somewhat larger model, very similar to the DAEX19OLP-4, sounds wonderful - that E string bass is as clear and deep as blue ocean water. There. All the secrets are out - for those lucky enough to have stumbled on this thread - you'll throw your 3-lolipop peizo pieces OS in the garbage can after hearing one of these, properly placed.
 
All of the build details are in the links.
Unfortunately, I'm getting a 404 on the github link. But we still have wikipedia (see below), the great grandpa of everything and everyone !!
In a slightly different vein ... Note decoding is done in a Teensy module. I'm still working on getting a VCO to track pitch bends on a string.
As opposed to the ribbon, your logic circuitry is picking up from the frets, making it difficult to obtain information on events that happen between frets. Nevertheless, the advantage in your case is that the originally intended operation of the instrument is still not compromised by the added setup.
The piezoelectric ones are useless ..
I find them useful as simple inexpensive pickups for 'acoustic' occasions. I've tried one that I 'salvaged' from an old buzzer on the top plate of a violin. They need EQ but they definitely work.

One problem is how to get nuance, like that available on a real guitar. I see he can do glissando, i assume vibrato using the touch board. The triggering part; how to get different tone or harmonic content depending on how / where you "pluck" or "strike" some physical feature? How to make it go louder depending on how hard? How to cut the sound short via a "mute" gesture? It's almost like the note selection methodology is easy compared to a nuanced, expressive instrument.
If I'm right that's where we need the right hand or even the foot.

You just have to design a touch interface for the "strings" and do the rest in software.

Proposed algorithm for "strings" with simple vibrations:

Eliminating all the unnecessary and boring stuff like tension, mass, square root etc., the string frequency (F) is inversely proportional to its vibrating length (L). For an open string note, this is the entire length from the nut to the bridge. For others, it is the distance from the finger to the bridge.

(High-school) mathematically,

F ∝ 1/ L which implies

F0 / F1 = L1 / L0 (suffix 0 means open string)

Taking the Wikipedia example (2nd link in #14) of a Jackson Professional Soloist XL electric guitar, the first guitar string should be 25.625 inches, which is the open-string length L0, with a frequency 330Hz (E4). Now, let us try to predict the finger position (distance) required for the tone B5 on the same string using the above equation.

B5, being the perfect-fifth for E4 (330Hz), makes 1.5 times the frequency, equal to 1.5 * 330 = 495 Hz.

F0 / F1 = L1 / L0 implies

330 / 495 = L1 / 25.625, solving which gives:

L1 = 17 inches, which is from the finger to the bridge (vibrating length) for B5. The nut to finger distance would simply be 25.625 - 17 = 8.625 inches.

Thus, a finger placed at a distance of 8.625 inches (from the nut) on the E4 (first) string of a guitar should make a sound that corresponds to the note B5. Can any of you guitarists please confirm the same ? Thanks.
 
Last edited:
Hi all,

It looks like I should have been a little clearer in my first post.
The instrument exists.

I have 22 photographs to post and in the next two days I'll take a trip into town and have them scanned.

I promise that you will be amazed even though I did not finish the instrument.

Sincerely,

Ralf
 
  • Like
Reactions: 1 user
Cool, it will be interesting to see what you did.

There have been many stringless guitars over the years varying from Guitar Hero "toys" to some playable instruments. I even tried an "electric neck" type instrument about 20 years ago that simply used 6 SMD phototransistors in 6 different holes behind each fret. Ambient light was used as the light source to keep each transistor conducting. As with the Axe Organ some CMOS multiplexer chips determined which hole was covered with a finger. It's pretty obvious now why this approach was pretty useless in a real playing situation, but the original concept started out on my old "stringlab" which consisted of a fretboard on a 2 X 4 with some tuners, a bridge and some pickups. It was created for photoelectric and other pickup experiments using optointerrupters and other things like the PC beepers. I even made some pickups with magnets, wire and nails. Ambient light isn't what it used to be either. Today we have LED lighting that is modulated both by the line frequency and harmonics, and by whatever randomly drifting ultrasonic switching rate the SMPS uses in each light source. Let multiple modern ambient light sources into any optical setup and trouble will find you.

There have been experiments going on for years using laser diodes for the source. I found this old one many years ago:

The latest iteration is here. Note that StyroPyro known for, and far better than me at blowing stuff up, so the "guitar" could be interesting.


For me, I prefer real strings on a real (strings and frets on a neck) guitar. My playing skill precludes the use of a fretless instrument. I have tried them, and I only make noise not music. I need the quantization provided by the frets. Once you make this choice, the fancy math is not needed as a Look Up Table will suffice.

As to being able to deal with "tricks" like bends, hammer ons and pull offs, one needs to analyze the output of a single string with a pickup. This requires a hex pickup to get a sense of each individual string. There have been several DIY methods, but I am experimenting with two different commercial setups. Graphtech makes piezoelectric string saddles and I had their setup on a test guitar. Roland has been selling guitar synths for 20+ years that use a hex pickup. I also had one of them on my test guitar. Getting the amplitude information from each string is pretty easy with a couple opamp stages.

Tracking the frequency of each string is not quite as simple since each string can span up to two octaves. Experiments with CMOS PLL chips reveal the main issue. A tracking PLL will revert to an idle frequency in the absence of an input. It will take a finite time to lock onto the freshly picked note during which its output must be muted. Low frequency transients from some heavy picking on the strings will push the PLL lower in frequency making the acquisition time random and occasionally long enough for the mute time to be audible.

A wired neck or other means of determining the approximate frequency of the note BEFORE it is played can be used to give the PLL a starting point. Experiments with an all analog approach did not work out, so I switched to a mixed approach using a digital PLL in a Teensy microcontroller. I have not worked on this in a while though.

The original stringlab was robbed for its useful parts and discarded when I had to move everything I owned 1200 miles because my 41 year engineering career ended. I even sold my guitar amps and some guitars. Stringlab II is partially assembled, but not finished. I put my broken Stratocaster back together and have purchased a couple cheap Chinese guitars to experiment with. I built a small 4 tube guitar amp, and have two more under construction. Then it will be time to play again.
 
When I say a "touch" interface, I don't mean the waveform. That comes from a wave table, and you probably want a number of options. No, I'm talking about the plucking attack and decay rates, and the "keyboard debounce" that decides when a "string" has been plucked, and maybe what is called "velocity" in an electronic keyboard. The frequency of a note will be hard coded in software, probably a "tempered scale" according to the well-defined standard scales. The physical means a normal guitar uses to achieve those notes is irrelevant to an electronic instrument.
https://www.researchgate.net/public...527595313972/Frequencies-of-musical-notes.png
 
  • Thank You
Reactions: 1 user