The Pressure Sensors:

The fingers of the left hand of a trumpet player are mostly inactive. They are not trained and skilled but they are usable for controlling actions while the ‘normal’ playing is done with the right hand.

When you hold the trumpet in the ‘correct’ way two of the fingers are wrapped around the third valve. The index finger and the middle finger. Those two fingers cannot move much (the hand holds the trumpet and should stabilize it) but they can assert some pressure. A logical choice after trying out some other options was to use pressure sensors.

The pressure sensors are mounted with tape on a plastic holder that wraps around the third valve.

Originally commercial FSR’s were used as pressure sensors. They have a few disadvantages though. The resistance changes over time and the sensing range declines. The sensors are not very robust and are likely to break after some time (I had to replace both pressuresensors after about a year). They have an awkward form that makes them hard to handle in a crammed space like I use them in. They come in certain sizes and forms which was not particularly handy for me using them on the third valve of the trumpet. The haptic feedback has to be added by putting rubbery substance on top but that decreases the sensitivity if to much is used. The sensors have an awkward feeling because they are attached to a rigid plastic shape that does not fit totally around the valve.

So I decided to replace them with fabric pressure sensors. They are built following a principle from the Kobakant team: http://www.kobakant.at. The instructions

are here:

The new pressure sensors still have some of the old disadvan-

tages. But the following issues are solved: Much better haptic

feedback, a sensitive area that you can shape and size

following your own insights, attachment around the valve

with a nice feel because of the neoprene. Originally I used

linqstat or velostat as the pressure sensitive material. It seems

though that piëzoelectric non woven fabric from

www.eeonyx.com is much more stable than these materials

Originally I used a voltage devider to use each pressure sensor

in series with a resistor of 10kΩ. Pressing the pressure sensor

down wil result in resistance varying between ∞ and 3.5kΩ.

The fabric pressure sensors have a different range which

differs per sensor and can change a little because of the way

the fabric is attached. So I got the idea to make a small

amplification circuit with two potmeters and an Opamp so

I can tweak the range according to my own wished tresholds and sensitivities. This scheme is from ‘FSR an Overview of the technology’  from Tech Storm Inc.

When the FSR is not touched we have the ‘no touch value’ and we can use potmeter R3 to make Vout = 0V. We can even make a treshold so that a small touch won’t have any effect. After choosing R3 we can use a potmeter replacing R1 en R2 to determine the amplification factor to determine which pressure will give Vout = V+