Halfway hihat pedal positions suppress hits of rolls

[anttipi]

FWIW, I also measured the Atmega with the front panel's polarity switch in the other position and the readings are 2.6 V and 3.8 V.

So the higher of the voltage readings dropped a bit when the polarity was switched but even then the Atmega is outputting 1.6 V more than when the exact same cable and sensor are attached to the ARM.

I wish I could make some sense of these numbers. I can only surmise the low voltage is causing interference/instability on the ARM. I suppose it can't be a firmware issue of CC reduction handling on the ARM because others are evidently not experiencing this. So something must be wrong within the ARM's hardware. All I can tell is nothing is rattling or doesn't look loose inside but that's the limit of my "expertise".

[dmitri]

As I suggested, you can send a sensor to me for testing. Another thing which I could suggest is to power the sensor separately from 5V (e.g. directly from USB) and feed only the signal output from the sensor to MegaDrum's pedal input and see if it makes a difference.

[anttipi]

Another thing which I could suggest is to power the sensor separately from 5V (e.g. directly from USB) and feed only the signal output from the sensor to MegaDrum's pedal input and see if it makes a difference.

Would it be safe to solder the power from a second USB cable onto the sensor, i.e. so that the sensor is powered by Megadrum AND a second USB cable simultaneously? Megadrum and the extra USB cable would both be connected to two different USB ports on the same PC. Also, in this case, does it matter which USB cable the ground is soldered onto? Or would this kind of configuration fry up my PC and/or Megadrum?

The max input voltage for the sensor is 10.5V anyway.

[dmitri]

Don't see why you cannot do it from two USB ports provided you wire it correctly. Just make sure with a multimeter before connecting to USB ports that ground and +5V are correctly wired. Also as a precaution you can add 50-100 Ohm resistor on +5V wire to the sensor.

[anttipi]

Don't see why you cannot do it from two USB ports provided you wire it correctly.

When soldering the 2nd USB cable directly to the hall sensor to provide extra power, is it sufficient to only solder the power wire to the hall sensor's power pin? I.e. can the ground be left disconnected on this extra USB cable? Obviously, the hi-hat controller cable from Megadrum is already soldered to the hall sensor's ground. In other words, two wires would simultaneously provide power but only the controller cable coming from Megadrum would be also connected to ground and signal.

(The two USB cables - the one powering Megadrum and the one going directly to the hall sensor - would be connected to two different ports on the same PC).

[dmitri]

Yes, you can connect only the power wire.

[anttipi]

OK, I soldered the extra power wire. The hall sensor should be now getting around 8 volts.

Results:
1. The reading on Megadrum's LCD is still jumping up and down regardless of the CC reduction setting, please watch here: https://youtu.be/mvhHKWhDv4I
2. The max level reading on Megadrum's LCD went up by about 250 to ~750, i.e. the range increased with the extra voltage.
3. At CC reduction level 0, Megadrum Manager's raw midi data still shows constant flow of CCs when holding the pedal position steady.
4. However, already at CC reduction level 1, the CCs come to an almost complete stop when holding the pedal steady. CC reduction 2 seems enough, whereas before even level 3 didn't eliminate it. (New Algorithm still causes CC leakage at CC reduction level 2, so I have to leave it to disabled.)

I have to test if I still get some erratic behavior from Megadum (very occasional missed or delayed hits, seems to only happen when you've played a little longer). I'm worried about the jumpiness I'm seeing on the LCD screen, I suspect it's eating up Megadrum's processing resources but I can't say for sure.

[dmitri]

OK, I soldered the extra power wire. The hall sensor should be now getting around 8 volts.

Results:
1. The reading on Megadrum's LCD is still jumping up and down regardless of the CC reduction setting, please watch here: https://youtu.be/mvhHKWhDv4I
...
I have to test if I still get some erratic behavior from Megadum (very occasional missed or delayed hits, seems to only happen when you've played a little longer). I'm worried about the jumpiness I'm seeing on the LCD screen, I suspect it's eating up Megadrum's processing resources but I can't say for sure.

If you mean some Lvl reading changes then it will not "eat up" processing resources.

[anttipi]

Dmitri, just out of curiosity, have you ever tried a hi-hat controller with a hall sensor (for example, a Roland one) with Megadrum?

I'm just kind of surprised I'm apparently the only one having these instability/interference issues. All possible causes external to Megadrum's hardware and/or software have now been ruled out (i.e. plugs, cables, soldering, hall sensors of different brands and sensitivities, voltage supply, external interference, host PC etc.).

On a further note: CC suppression seems to work less efficiently when new pedal algorithm is enabled vs. disabled (i.e. less suppression on identical CC suppression level setting). Could this maybe be tweaked in a future firmware so that the new pedal algorithm would be usable?

[dmitri]

No, I myself never tested it with a Hall sensor. But other did.
I do testing with a variable resistor. I can tweak it but I first need to be able to reproduce it to do any tweaking.