strange input behavior

[zabredfo]

Hello,

I've built the all in one kit v2.8 but unfortunately I have some
problems with the inputs on my megadrum:

- On every input that is connected to an input y0 of a 74HC4851
multiplexer (pins 4, 13, 23, 31) I measure about 40-50 mV (AC). When
I connect a piezo element to these inputs it makes a quiet
sharp, buzzing sound (~ 230 Hz), the VU meter shows a high input
signal on this channel and midi signals are triggered more or less with
a constant rate (but that is to be expected, I guess).

- On all the other inputs there are about 5-10 mV, but I don't know if this is
actually a problem, is there supposed/allowed to be any voltage on the
inputs ? There is a much quieter sound with a much higher frequency.
These inputs pick up the signal from the piezo element pretty good, with
the default settings of the megadrum.

I use the megadrum (firmware 20090211, 16 MHz) with the default
settings and for now with a simple piezo element, since I have no drum. I
don't think that the parameters threshold, gain, etc. have anything to do
with the problem.
It doesn't make much of a difference whether I use USB or a separate
power supply (I suppose 12 V, 2 A is OK...?). I also measured directly
on the pins and of the 40 Pin Connector, so I guess there is no problem
with the way the jacks are connected.

I find it interesting that the output voltage problem is spread over
all multiplexers, and that the multiplexers all behave the same.
And while it is possible that they are damaged by now I guess that this
strange behavior is initially caused by something different.
It seems to me that the AC on the inputs (and possibly it's frequency ?)
depends on the select inputs of the multiplexers. Is that possible ?

I only ordered the PCB and the ICs from Synthex (many thanks !)
so I guess the first error source would be the components I bought,
but I'm quite sure they are the correct ones. I've also searched the
board for shorts but so far I couldn't find anything. But I have only very
limited electronic knowledge and this is my first electronic project, so I
don't really know where to look for the cause of this problem...
Any help is very appreciated!

Thank you very much!

[dmitri]

Hello,

I've built the all in one kit v2.8 but unfortunately I have some
problems with the inputs on my megadrum:

- On every input that is connected to an input y0 of a 74HC4851
multiplexer (pins 4, 13, 23, 31) I measure about 40-50 mV (AC). When
I connect a piezo element to these inputs it makes a quiet
sharp, buzzing sound (~ 230 Hz), the VU meter shows a high input
signal on this channel and midi signals are triggered more or less with
a constant rate (but that is to be expected, I guess).

I'm not sure what pins you are talking about since 4851 has only 16 pins.
On a pedal input and on any input configured as a switch there should positive pulses
which can be registered by a multimeter as a positive bias.

- On all the other inputs there are about 5-10 mV, but I don't know if this is
actually a problem, is there supposed/allowed to be any voltage on the
inputs ? There is a much quieter sound with a much higher frequency.
These inputs pick up the signal from the piezo element pretty good, with
the default settings of the megadrum.

I don't know why you get a buzzing sound. When I connect a piezo to any input configured as 'piezo'
I don't here any buzzing.

I use the megadrum (firmware 20090211, 16 MHz) with the default
settings and for now with a simple piezo element, since I have no drum. I
don't think that the parameters threshold, gain, etc. have anything to do
with the problem.
It doesn't make much of a difference whether I use USB or a separate
power supply (I suppose 12 V, 2 A is OK...?). I also measured directly
on the pins and of the 40 Pin Connector, so I guess there is no problem
with the way the jacks are connected.

I find it interesting that the output voltage problem is spread over
all multiplexers, and that the multiplexers all behave the same.
And while it is possible that they are damaged by now I guess that this
strange behavior is initially caused by something different.
It seems to me that the AC on the inputs (and possibly it's frequency ?)
depends on the select inputs of the multiplexers. Is that possible ?

Voltage on inputs depends only on if the input is a pedal, piezo or switch input.
Even on inputs configured as piezo there could be positive bias due to separation in time of the multiplexer selection
and setting the input as piezo or switch/pedal.

I only ordered the PCB and the ICs from Synthex (many thanks !)
so I guess the first error source would be the components I bought,
but I'm quite sure they are the correct ones. I've also searched the
board for shorts but so far I couldn't find anything. But I have only very
limited electronic knowledge and this is my first electronic project, so I
don't really know where to look for the cause of this problem...
Any help is very appreciated!

Thank you very much!

[zabredfo]

I'm not sure what pins you are talking about since 4851 has only 16 pins.
On a pedal input and on any input configured as a switch there should positive pulses
which can be registered by a multimeter as a positive bias.

Sorry, I should have made clear that I am talking about the pins of
the 40 Pin Connector that connects to the jacks. So basically what
I failed to say was that channels 8, 16 and 24, which are always
set to 'piezo' (right ?) behave somewhat as if they were set to
'switch' in that they have a voltage of ~ 50 mV and a piezo
connected to these inputs buzzes, not very loud but clearly audible.

Probably due to the buzzing, I suppose, these 3 channels need
a higher threshold (~60) and a very low gain (0) so that
they don't trigger midi events at a constant rate.

I just find it noticeable that exactly these channels that are connected
to a y0 ("Lo-Lo-Lo"-selected input) on separate 4851 multiplexers behave
so differently from the rest. I thought that could mean something but
it could be a coincidence.

Now I understand that at least the pedal input (pin 4 on the 40
pin connector), also connected to y0 on a 4851, works as it
should (positive pulses).

I don't know why you get a buzzing sound. When I connect a piezo to any input configured as 'piezo'
I don't here any buzzing.

Well, the other (not 8, 16, 24) channels set to 'piezo' make the piezo buzz
almost inaudible, I have to agree - maybe I heard my tinnitus there,
so I guess I can consider one of my "problems" as solved

Voltage on inputs depends only on if the input is a pedal, piezo or switch input.
Even on inputs configured as piezo there could be positive bias due to separation in time of the multiplexer selection
and setting the input as piezo or switch/pedal.

I am not sure I understand what you mean. Would that explain a positive bias on 'piezo' inputs
even if ALL inputs were set as piezo, so that there wouldn't ever be a
switch between "switch voltage" -> "piezo voltage" on multiplexer selections ? This is my setting,
but I guess whatever causes the channels 8, 16, 24 to have this high voltage might be
considered as a "switch setting" of a channel in this scenario.
Nevertheless, it sounds like some small voltage of few mV on the inputs is not harmful
to pads connected to them, that's good.

In case there isn't any other place where this information can be found (without studying the
schematics and the datasheets ) ...for electronic illiterates like me (but willing to learn),
could you please give some indication what "good" voltage ranges for the
pedal input(hi/lo impedance), a switch and a piezo input might be ?

Dmitri, many thanks for your help (and of course megadrum itself)!

[dmitri]

When a pedal input is being scanned, a weak pull-up resistor is applied to all ADC inputs. This will cause positive pulses on inputs 8,16,24 even if they are set as piezos. This should not effect anything except maybe causing a slight buzzing in the piezos but it really shouldn't cause you to raise the threshold compared to other inputs. I may change the future firmware to apply a weak pull-up only on ADC0/ADC7 when the pedal input is being scanned.

[zabredfo]

If a firmware update might solve that problem, it would be great.
Thanks a lot for the help, dmitri!