namke
wonkystuff
electronics and sound, what's not to like?!
Posts: 686
|
Post by namke on Oct 26, 2021 20:55:01 GMT
and ... it's another fail. Still have lots of artifact in output sound, can't figure out where is the issue. It's working perfectly on the Uno board, but not on mine. It's not on the output filters as the artifacts are already there on the Atmega pin before the filters. It's not the timing, I'm using a crystal oscillator like on the Uno board. I guess it's not the power supply, I'm using the same for the Uno board. Already tried to exchange the working Atmega on the Uno board with one I bought, in case I had fake ones or bad quality despite I got them from a good shop. any ideas guys ?  Is the processor running at the speed that you are expecting? for the ATTiny85 there is a fuse setting which means that the clock runs at an eighth of the speed (so 8MHz runs at 1MHz). If this was the case then what you might be hearing is jitter because your sound-generating code is not running at the correct rate. If you have an oscilloscope it would be worth simply toggling a pin on the IO port every time you enter the interrupt routine (e.g. PORT ^= 1; to toggle the lowest bit) and then seeing how fast the pin toggles. |
|
namke
wonkystuff
electronics and sound, what's not to like?!
Posts: 686
|
Post by namke on Oct 26, 2021 20:58:00 GMT
|
|
|
|
Post by m4vrick on Oct 27, 2021 12:23:32 GMT
Yes I checked the fuses, which are some kind of confusing.
On the Uno board the Lfuse is 0xFF, on my board 0xF7.
As I understand with an external crystal we should use 0xF7, dont know why the Uno is using 0xFF and working better than my board.
Anyway I tried both without more success.
I could not check the real working frequency for some days, I'm on holiday with my family. That's a good idea, will try it for sure.
|
|
|
|
Post by m4vrick on Nov 5, 2021 10:42:47 GMT
Finally found the root cause ! And it's even more disturbing... The Atmega is working perfectly, the problem comes from the mozziAnalogRead(). On my board I rearranged the analog inputs to simplify the PCB routing, and it seems that's the problem. Even if all the analog input are the same from A0 to A5 when using them as on my board the sound is bad. I tried to rearrange the input according to my board on the Uno board and the issue is the same  When using the same input order as on the original project the sound is good. I can't understand why, it simply using mozziAnalogRead from 0 to 5. A simple ADC read... |
|
|
|
Post by keurslagerkurt on Nov 6, 2021 16:13:34 GMT
Very happy you found the cause, but that is indeed very weird!! I have no clue on how the Mozzi functions work exactly, so I can't be of help here unfortunately..
|
|
|
|
Post by m4vrick on Nov 6, 2021 17:15:44 GMT
It's working right now... but I dont know why  There is something around the mozziAnalogRead() function, I can't figure out what exactly. Anyway the most important is the result, it's sounds great now. I'll post an example soon. |
|
|
|
Post by m4vrick on Nov 7, 2021 18:17:48 GMT
So, here is a short video with the sound of the two modules.
First one is additive VCO, second one is Chord.
I have to adjust the output level, I'll wait to test them on a friend's system to confirm.
|
|
|
|
Post by moruial on Nov 7, 2021 19:39:26 GMT
The Chord sound really great as I've told you. It as some SOLINA vibe 
PS: sound only on left ear, or is it just me?
|
|
|
|
Post by m4vrick on Nov 7, 2021 19:47:31 GMT
Yeah, it's only on the left channel on the master module stereo jack...
|
|
tgergo
Junior Member
Posts: 59
|
Post by tgergo on Dec 12, 2021 10:51:46 GMT
How is the additive VCO project coming along? Any news on the chord module? I have had an idea for a chord module in my mind for a while too. The idea was a module where you could enter a set of chord parameters (type and inversion), then select combinations from these with a CV input for each and the module would generate the 3 or 4 1V/Oct values based on the combination of the 2 parameters. I'm just putting the idea out there because who knows if I ever get to this, so maybe I can inspire someone else to do it.  |
|
|
|
Post by m4vrick on Dec 12, 2021 13:42:59 GMT
Thanks for asking news  I'm actually doing some testing and will have soon others guys testing the two modules. Your idea seems interesting but I'm not really sure to fully understand it, as I understand it this could not work on this particulare module as there is no way to generate multiples CV signal or having several chords. |
|
