Noisy encoders on Zynthian v1 [SOLVED!]

#1

Ever since I bought my Zynthian kit in June 2017 it has been very noisy on the encoders. I put it aside for some time because of that. Still it intrigues me.

As soon as I turn the encoders while playing a note the audio starts crackling and clicking. When no notes are played everything is fine.

I have tried changing the encoders, changing the location where I play, moving the encoders/screen away from the Pi, trying all different synth engines, even putting a 40106 in between to clear up the debouncing noise (which it did), but still there was plenty of noise on the audio out.

I’m using a v1 setup (Adafruit small screen) RPi3 B+, Hifiberry Dac+. I used the Gorgona Omega image which is configured for v1 kits. Triple checked my wiring and connections. In fact the Zynthian works like a charm except for the noise on the encoders…

Any tips would be more than welcome!

#2

I wonder if the encoder is adjusting a synthesis parameter which may interfere with the output?

#3

Had that problem once, in another rPI project. The culprit was a poorly filtered TFT display. A bunch of capacitors on the TFT power supply lines solved the problem.

YMMV, obvioiusly,

1 Like
#4

Hi @nenzis!

You have to be more specific … what engine and preset are you using? What kind of noise is it? Is it software or hardware related?

Some engines/plugins produce clicks & glitches when modifying some parameters. It’s not a zynthian specific problem but a bad software design. Probably the author didn’t think that changing these parameter while playing could be useful. Or perhaps implementation is much more complex if all parameters can be modified while playing.

Another question is that the encoder itself could produce electronic-noise that can be listened … this happens sometimes and normally can be solved by adding de-bouncing capacitors (if not already added), good grounding, metal case, etc.

Regards,

#5

hi @jofemodo,

As I wrote in my initial message this is happening in all synth engines. see my included sound file. I also wrote that I used CD40106 (Schmitt trigger) to avoid debouncing noise. I could see a difference in my scope, nevertheless the audio output was as noisy as before.
The TFT I’m using is the one from the kit (PiTFT 2.8). Don’t know if it is well filtered.
I used different power supplies, also my lab power supply but this didn’t make a difference.
Running out of clues…

#6

Does the noise appear in audio recordings?

#7

@wyleu, yes, I enclosed a recording in my first post.

#8

Curioius! If the noise is present in the internal recordings, you should been experimenting XRUNS.

0.) Do you see the XRUN symbol in the status area (top-right)?

1.) Did you experiment the noise with every engine? Every preset?

2.) Is the noise present when moving any encoder, independently of the assigned parameter?

3.) Did you correctly soldered the de-bouncing capacitors to the controllers? Could you send a photo?

Regards,

#9

Hi @jofemodo,

Thanks for taking the time to look into it! Much appreciated.

The internal sound is ok. No noise at all. (no xrun symbol either)

I tested every engine, every preset and plugin, but it is everywhere, as soon as I turn or press 1 of the 4 knobs, independently of the assigned parameter. It is as if it is coming from the screen somehow. Even going all the way back to the admin menu, everytime when I press or turn the encoder the noise is there, when a sound is played.

I have desoldered the caps on 1 encoder to see if it would make a difference, but it makes no difference with the other three. I have breadboarded a CD40106 Schmitt trigger IC and hooked it up to one of the encoders. On the oscilloscope there was clearly debouncing noise before, and adding the Schmitt trigger I got rid of the noise. A very nice exercise, but this made no difference in the operation of the encoders as I expected, also was there no change in audio output either…!

Cheers!

IMG_8868|375x500,75%

IMG_8869
IMG_8869.JPG2448×3264 1.61 MB

#10

This is a really good question.

@nenzis, the recording you attached was done on USB pendrive directly from inside Zynthian or has been recorded by an external device?
If it’s the former, that noise can’t be an electrical interference, because there is no “vulnerable” analog path from synth engines to the file.
If it’s the later… it’s 99% an electrical problem, and I bet my 2 cents that the problem is non related to encoders cabling… if you have a 'scope place one probe on the audio output and other on various places solicited when rotating encoders, on supply rails (AC coupling), on SPI data signals, rotate knobs and search for correlation between signals and noise.

#11

The capacitors seems OK :wink:

If the noise is “analog” noise, i would try to determine the “real” source of the problem. @Axeman has pointed very well the path to follow … if you have an oscilloscope . If you have not, then you should try with a different display. A generic HDMI monitor will serve to check if SPI is the source of problems.

Regards,

#12

Ahhh! Very important too … check that your power source is enough. When system is under-voltage, “strange things” occurs more frequently.

1 Like
#13

Allright, Thanks all for the feedback.
I’ll dive into my Zynthian one of the next days and I’ll report back!
Cheers

#14

Good news! Probing around I found the TFT power supply line to cause the noise. Adding a 1uF cap between 3.3V and GND close to the PiTFT fixed the problem.
Thanks all for being so active on the forum!

8 Likes
#15

I told you… :wink:
(I’m psychic on certain problems)

2 Likes
#16

There will, of course, be the traditional charge of one sound sample to thank the good people of this fine establishment for their efforts! (and don’t think sound samples in early posts count… Especially if I don’t notice them!) :face_with_monocle:

#17

@wyleu You are as undaunted as the sinking Titanic orchestra :rofl:

2 Likes
#18

Only that there was hope on the Titanic, that they would stop.

3 Likes
#19

There is no hope with our dear wyleu …
:rofl::rofl::rofl::rofl:

1 Like