Vielen Dank für den netten Kommentar!

What goes in, is probably a very fine, and warm, sine. That stays that way and directly drives your headphones. Possibly the amplifier clips, so it pick ups higher harmonics that color the sound a bit harsh. You might check this on a scope. I haven't tested it (yet) but the top-to-top voltage of the magnet-output is probably only a few millivolts. Perhaps you could try a good amplifier with a 'phono-' (or even mic-)input?

@@arnoud-whizzbizz I guess you are right! I checked the output of the instrument with a scope some weeks ago and only changed to the combination amplifier/speaker one or two days before Maker Faire Ruhr. I will check it now and let you know!

The output without headphones or amplifier is about 1 V, with headphones it is about 200 mV. You are right, the amplifier is the problem. It doesn't seem to be simple clipping though. My son soldered it at a fischertechnik Modelshow at Münster. I will take a closer look at it later.

@@arnoud-whizzbizz Ok, the LM386 amplifier contained a 10 µF capacitor from pin 1 to pin 8. This causes an amplification of 200. Removing this capacitor gives an amplification of 20. Still not headphone sound, but much better. Only the lowest note is now less loud than the others.

@@faszinationfischertechnik2528 Better! The thing with the whole LM386 is, is that it only kind a 'defeats' its input. So (apart from a bad signal-noise ratio) it might clip its input. A better amplifier would have a buffering input to accept real low amplitude inputs without distortion and feeds that to the real 'amplifier'-stage whose amplification could be regulated. Another BIG difference between your headphones and the output signal of this simple amplifier is that the latter is not symmetrical. It should have a DC-offset to output the sine without distortion in the first place. Worst case if that you're only listening to half the sine (although I don't think so if you looked at it at a scope, or not?) Remember that your input signal is symmetrical. Best would be a real symmetrical amplifier, but you can't feed that on a single battery.

Thanks! If time permits, I want to make a pentatonic instrument with keyboard. By pressing the keys you will bring the electromagnets closer to the tone wheels, so that you can play louder or softer and you can play chords. This also eliminates the clicks from the contacts. The clicks are not too bad, but the sound of the tone wheels with headphones or a decent, big enough speaker is so good, that I wish to get rid of them. To get to a pentatonic scale you just have to add a third axle. On this axle there are the same tone wheels as on the faster axle here, but it turns 3/2 as fast. Thus, no new tone wheel design is neccessary.

@@faszinationfischertechnik2528 Perhaps less 'clicks' if you replace the contacts for reedcontacts and score the song with magnets? :-) Best would probably be an easy mechanical method of plugging in on the sine when it crosses zero, then you'll don't have the higher harmonics of cutting in the phase of it. Mechanically that would mean you switch just between magnets on the inducting wheels 🤔

@@arnoud-whizzbizz Good idea! One could use the magnets from the dynamics "Kugelbahnen". However, the strength of these magnets vary quite a lot, so I expect that it could be difficult to get a constant length to the notes.

@@faszinationfischertechnik2528 Best would be to play the strength of the induction that's been picked up directly (as you suggest), by moving the induction coils (ft electromagnets). Much better idea than switching sines (mechanically) with any mechanical contact.That would be great!