MegaDrum Velocity Sensitivity

[alexmcintosh]

Dmitri;
my original conclusion, was the same as your conclusion. And after several experiments, the best result I got was to simply connect a 33k Ohm resistor in series. This worked better than adding an aluminum plate or a mouse pad (which tended to be to mushy & give unstable results)
After these experiments you've suggested I'm perplexed that a 27mm & a 35mm piezo that put out 20V & 30V peak to peak (respectively) would not peak out the TMI (set to 1023 HighLevel). And both generate the same midi number of 74. It makes me think there's something limiting the input.

Sincerely.. Alex

[dmitri]

Dmitri;
my original conclusion, was the same as your conclusion. And after several experiments, the best result I got was to simply connect a 33k Ohm resistor in series. This worked better than adding an aluminum plate or a mouse pad (which tended to be to mushy & give unstable results)
After these experiments you've suggested I'm perplexed that a 27mm & a 35mm piezo that put out 20V & 30V peak to peak (respectively) would not peak out the TMI (set to 1023 HighLevel). And both generate the same midi number of 74.

The absolute voltage in this case doesn't matter. It's the relative voltage produced by piezo between medium and hard hits. If you want to make sure of this, take an oscilloscope and measure signal from the piezo when doing medium and hard hits - you will see that they produce signals with the same voltage.

It makes me think there's something limiting the input.

I don't understand how you came to this conclusion.

[alexmcintosh]

Dmitri;
I came to the conclusion since two different Piezos both connected directly to the ribbon cable generate the exact same midi value 74. They are two different Piezos with exactly the same value maxed out. And, that just seems to odd that something that can generate 30 V cannot Max out the HighLevel, I would think it could.
I did attach the Piezos to a sound card oscilloscope. I noticed some clipping on the waves for stronger hits. But each hit seemed to have it's own corresponding wave, as the harder hits had more clipping. I don't know if this clipping is the Max level you mentioned or if it's a fault of using a sound card oscilloscope.

Thanks.. Alex

[dmitri]

The only other reason I can imagine is if you modified the Curve you use to have a flat horizontal shelf at level 74.
Sound cards will clip signals higher then ~5V.

[alexmcintosh]

Dmitri;
is there a way for me to (safely) max out the input of the Megadrum? Something like a 5V battery and connect - (minus) leg to pin 15/16 & + (positive) leg to pin 33? Therefor I could test to see if the TMI can read a maximum signal.
I've looked at all of the curves (via MegadrumManager MDM) and all go up to 127. (Again, the best solution is to add a series resistor of 33k Ohm, which will then gives me a nice linear response from soft to medium to hard hits. Which indicates if I reduce input signal to the TMI then I get a great linear response)
Thanks.. Alex.

[dmitri]

Tapping on a bare piezo easily maxes out MegaDrum inputs with the default Gain setting.

[alexmcintosh]

Dmitri;
but it does not easily max out, as I've showed. It easily goes to 74 of 123 on a HighLevel of 1023. So, what would cause a maxed out signal to correspond to midi 74? Could there be an ADC conversion with a bit missing? (or perhaps a bad solder joint causing a missing bit?) It doesn't seem to correspond directly to binary but 63 (binary 111111, OR 3fh ) is close to 74? I'm guessing here, but what do you think?

[dmitri]

Dmitri;
but it does not easily max out, as I've showed. It easily goes to 74 of 123 on a HighLevel of 1023. So, what would cause a maxed out signal to correspond to midi 74?

Incorrect Curve, Compression level.
Signal voltage maxes out around 2-3V when loaded on the input.
Incorrect wiring.

Could there be an ADC conversion with a bit missing?

If it were the case then you would not have been one experiencing this.

(or perhaps a bad solder joint causing a missing bit?)

If it were the case then there would be almost zero chance of this happening on all inputs.

It doesn't seem to correspond directly to binary but 63 (binary 111111, OR 3fh ) is close to 74? I'm guessing here, but what do you think?

I don't have any more ideas besides which I already suggested.
If you suspect the main board being fault, PM to arrange sending it back to me for testing.

[airflamesred]

74 does seem an odd figure. No help, I'm afraid.

[dmitri]

If you haven't missed anything from describing settings/wiring/testing procedure/piezos used and etc, I do really advise you to send it back to me along with one of your piezos for testing.