This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

OPA134: Device that vibrates the strings of a guitar

Part Number: OPA134
Other Parts Discussed in Thread: LM386, , ULN2004A, ULN2002A, ULN2003A, LM4871

I am creating a device that vibrates the 6 strings of a guitar. Let me explain what I have working, then let me explain why I need something better.   From the output of the guitar's magnetic pickup, I "Yed" the guitar's signal to a OPA134, arranged as a non inverting op amp with gain. The output of the OPA134 then went to a LM386 set to the maximum gain of 200.  The LM386 was originally designed to drive a 8 ohm speaker.  The output of the LM386  is what drives the "things" that vibrates the metal guitar strings.  Ok, what are those "things" that vibrate the guitar's metal strings?  The "things" are 490-CEM-1203(42) magnetic buzzers.  I removed the top half of the buzzers, the part that made the noise.  This left only the bottom half of the magnetic buzzer, which contained the magnet, with copper wire wrapped around the magnet.  If you run current through copper wire wrapped around a magnet, you create a electromagnet. By now you should be getting the point.  Meaning the guitar's sine wave signal activates the tiny electromagnets laying close under the guitar's steel strings.

It seems to work best when there are two of these tiny magnetic electromagnets under each string, especially the smaller strings.  Each of the tiny magnetic electromagnets are 42 ohms.  I selected 42 ohm values for a reason.  That reason was to try and match out the ohm value of the output of the LM386, which is 8 ohms.  I connected the 42 ohm tiny electromagnets all in parallel.  This means 4 of the 42 ohm tiny electromagnets in parallel had a ohm value of 10.5 ohms.  6 of them in parallel had a ohm value of  7 ohms.  This means we are already below the 8 ohm design of the LM386.  Therefore, if we put 12 of the tiny 42 ohm electromagnets on the guitar, in parallel, we have a total ohm value of 3.5 Ohms.  The LM386 was not designed to drive a 3.5 ohm load. 

Possible solutions: Hooking up two of the LM386 power amps, each driving half of the load. That seemed like the easy way out.  I then tried another approach for the electromagnet part of the device.  A guitar's pickup is wound with approximately 40 AWG magnet wire, to ohm values ranging from 10K ohm to 20K ohms.  I removed the original magnet wire from a guitar's pickup, then wound new 32 AWG magnet wire around the magnets, until I got a 8 ohm reading.  This did work, and did vibrate the strings, but did not vibrate the strings as strong as I would like.  The distance the guitar strings are away from the electromagnet field determines how strong the magnetic field needs to be.   To me this meant the LM386 was not putting out enough current, to create the magnetic field vibration I wanted.  This led me to think I needed a amp more powerful than the LM386, and that would require searching.  The other problem involved with finding a new and more powerful amp, is finding one with suppression diodes.  Any time a magnetic coil collapses current can flow in reverse.  So a new amp would have to be a audio amp designed to drive a speaker.  A small audio amplifier with one input, and one or two outputs.

Then I started thinking outside the box.  Meaning there should be a totally different and better way to vibrate the strings. I thought of using a transformer to match the ohm values,  but that idea seemed to draw too much current.  Maybe I am wrong about too much current.  I also thought of creating a stronger single electromagnet, but that also required more current.  Then I happened upon a idea that seemed good.  I read data sheets on TI's Darlington Transistor Arrays.  If I could only bias the 7 inputs of the Darlington Transistor Arrays from the output of the OPA134, this would provide everything I thought I needed. Then I find that the ULN2002A has a 7 volt zener on the input, and was designed for P-MOS switches.  The other two similar arrays ULN2003A and ULN2004A  are designed as switches for TTL and CMOS devices. 

At this point, I am dead in the water, and it is not a good feeling!  I would like some suggestions on how I can better accomplish controlled string vibration?  Thanks Keith Hilton

  • Maybe all I need is one powerful Darlington Transistor biased as audio, following the OPA134.

  • Hi, Keith,

    Thank you for providing all the details of your application.

    Definitely, you would require of an amplifier with more current capabilities than the LM386. We would recommend a similar device such the LM4871. This device may support loads down to 3 ohms. So, it could be a better replacement to handle multiple loads.

    Please take a look at it and let us know if you have additional questions or comments.

    Best regards,
    Luis Fernando Rodríguez S.
  • Thank you Luis. I will check out the LM4871. I read the data sheet and it looks like it might be a solution.
  • Hi Keith,

    I think it's a good decision to use a single coil pickup to generate the magnetic field. But it could be that the Alnico magnets are saturating at rather low magnetic fields, so that they could stop working at rather low stimulation currents. Maybe it is helpful to replace the Alnicos by soft iron cylinders showing less saturation? Only a guess...

    Kai
  • Kai, thanks for your ideas. The pickup I wound to 8 ohms using 32 AWG magnet wire had 6 magnetic cylinders. The magnet wire was wrapped around the vertical row of cylinders, not around each individual cylinder. On the other hand, the little buzzer magnets had magnet wire wrapped around each individual cylinder. The cylinder's pickup I wound had much bigger magnetic cylinders than the tiny little buzzer magnets. The problem with individual cylinders of any kind is that they have to be lined up with the strings. If there was one solid bar magnet across all strings, the the need to line up the magnet on individual strings would not exist.
    After some failed experiments using flat bar Alnico magnets, I realized there is a lot to know about the magnetic field, and north and south poles created by a particular magnet. I suspected using really strong magnets, or magnet, could hinder the natural vibration of a guitar string. Meaning, even if the vibrating device was turned off, a strong magnet would still exert either a strong pull or strong push on strings. I have thought about using soft iron. Soft iron could change the magnetic field, depending where it was used with a magnet or magnets. Soft iron, without a magnet, with magnet wire wrapped around it, does create a magnet. The more wraps of magnet wire, the stronger the magnet. Also, the more current run through the wires, the stronger the magnet. Alnico magnets are super strong. In the failed experiments using a bar Alnico, I wondered if the magnet was so strong, is was changed little by the wrapping of copper wire I did, and the current from the LM386. To vibrate strings, there has to be a change in the magnetic field. I got the suspicion the Alnico was so strong there was no change in the magnetic field. Seemed to me the Alnico was so powerful that it would take a unrealistic amount of magnet wire, and a unrealistic amount of current to alter the magnetic field put out by the Alnico. I wonder if soft iron and magnet wire alone is the answer. Of course that would totally depend on how many ohms of magnet wire would be needed to create the needed push and pull. The amount of magnet wire is limited by how low a amplifier can drive ohms, and how much current the amplifier can produce.