www.megadrum.info

Anyone for electrostatic drum sensors

This is my mag' powered test unit.

This thing is running an eight zone head plus rim. It uses the ATtiny45 so needs a separate USB input for each drum. I have two passive and one crazy dual personality op-amp PZT conditioner, but the cost and simplicity of the magnetic version is a bonus

What I think is really need is a serial chained connector for the whole kit: any loony comments on using CAN (TOYOTA owners aside) or Ethernet for hammering the plastic please!

Very very nice Ken!!

But, ( there is always a but) wich program like SD 2.0 or BFD(2) will give me the room to program nine zones on one pad?
Or are the eight zones going to be translated to cc messages for positional sensing?

r0bbie wrote:But, ( there is always a but) ...

The standard seems to be a CC 16 followed by the note on.
I wanted this toy as a trigger to http://www.falstad.com/circosc/. Very bottom up and probably a waste of time

But - then again <this> is an interesting fact:
'The mute patch on the bayan is made from 'a mixture of gum, soot, and iron filings'.

Now Tabla have <positional sensing> *and* the pitch can be modulated by using the heal of your hand

In general PZTs are too lively ('too hot') for e-drums and this produces unwanted resonance and ringing into the system so I've been playing with some alternatives.

+ Electrostatic IC's:
Got bored because I couldn't get enough sensitivity out of it over the required deflection of 25mm

+ Laser optics (The Bewitchin' Pool):
Really accurate and predictable results from this but not actually mounted in a drum. This is my head and software reference.

+ Using the head as a restive element:
Weave silver coated nylon thread through your muffle head and away you go.
I have not tried concentric patterns yet but I might sew LEDs into my shirt...

Hall effect (HES):
This was my preferred system. The output is totally configurable removing the need for velocity curves and such but they have the huge disadvantage of sensitivity to lateral movement.
This complicates repeatability of DIY construction.

Linear Variable Differential Transformer (LVDT):
Verrry good - very repeatable. Could be made more sensitive than PZT and HES with no lateral problems. True linear output in a compact unit.

Optical LED mouse hack:
I still have to try this but I believe these chips can be interrogated @ 100ns or so?
Dangle a suitable runner directly connected to the head across a mouse sensor. This sensor could be made very small with no PZT-like adjustments needed.

Be sure - megadrum is the *FINEST* trigger you can use.

Any-road-up; I think I have sorted my multiplexing issue by using a jitter free sample period.

My method is inspired from <http://www.jandspromotions.com/philips2005/Winners/AR1746.htm>.
This design essentially works out the frequency component using a (tuned) traditional method, his deviations are from a known 'universal' constant.

In my application, real time slope definition is possible for drum applications in < 16 samples @ 40 9bit samples/ms.

So by combining slope detection and an adaptive threshold I hope to do away with lots of user input.
The adaptive threshold is intended to work with inputs at @ 1.1v!
Maybe I have have 6 (lower frequency) rim inputs mux'ed into one channel

Back in the clean room - what's wrong with them apples...

Looks Ok but it's still magnetic ...

Is it a drum pad ??!!

Is it a bird?
Is it an airplane?
Care to share some details?