Open Theremin v3 - bad sound

Posted: 8/20/2020 12:55:27 AM
Bayaneer

Joined: 8/20/2020

Hello!
I tried to build Open Theremin and made a few of shields but all they soundd terrible.
Middle register is normal but when I try to play in lower register I hear a lot of different noises but not normal sine. In upper register there are high-frequency noises, especially when the register pot is in higher position.


Device calibrates normally and works well except bad sound.

What can I try to fix this problem? Or it`s normal?

Posted: 8/20/2020 2:10:40 AM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

The high end is clearly aliased, the low end seems to be suffering from limited resolution.  How many shields have you made?

Posted: 8/20/2020 9:32:38 AM
Thierry

From: Colmar, France

Joined: 12/31/2007

That's definitively not normal. There can be few mellow aliasing ghost tones in the highest register setting, but only in the very highest octave of the 5 octave playing range. The gargling in the lowest register setting sounds to me like a dirty power supply, interference, or a missing ground issue. That's definitively not expected behavior. 

You say that the calibration procedure runs normally. Would you mind sharing the detailed debug output, though?

Which software release did you install? The initial v3 had some bugs. That's why I heavily reworked it which led to the release of the HD Audio version which you can download here. The release notes are as follows:

1. Fix a wavetable addressing issue (found by @miguelfreitas)
2. Use the Arduino's hardware SPI to control the DACS and use the Latch signal to reduce audio jitter
3. Improve the register switch to transpose by clean octaves and keep the tone spacing and pitch tuning consistent
4. Improve the volume response to give a smoother start and wider dynamics (*)
(*) This relies on a recent gcc compiler version. Make sure to compile it with the Arduino IDE >= 1.8.10


And a - more general - question, rather for my scientific curiosity: Why do people take the effort on them to build their own shields if you can get a readily built and tested shield including antennas for only 89€ ? Take the price of the materials and the hourly wage of an electronics technician into account for the hours you spend with it and you will see that you can never build it at such a low price.

Posted: 8/20/2020 1:12:24 PM
dewster

From: Northern NJ, USA

Joined: 2/17/2012

"Why do people take the effort on them to build their own shields if you can get a readily built and tested shield including antennas for only 89€ ? Take the price of the materials and the hourly wage of an electronics technician into account for the hours you spend with it and you will see that you can never build it at such a low price."  - Thierry

Yes, I wonder about this too - it seems many take the DIY route and end up having trouble (and no wonder, there's a lot going on, and one little thing wrong can kill it).  Though the soldering job in those photos is a bit rough!  Are the antenna connectors banana plugs?

Posted: 8/20/2020 7:36:22 PM
Bayaneer

Joined: 8/20/2020

Thank you for your answers.


How many shields have you made?


I made three shields with different chips but they all sound the same. But I used same DAC`s in all devices because I couldn`t find another in my city.
Also I can`t measure capacitors. Perhaps, I bought wrong or bad-quality capacitors.


The gargling in the lowest register setting sounds to me like a dirty power supply, interference, or a missing ground issue.

I measured the 5v with different power supplies. USB power seems worst than power from battery but I don`t hear any difference in sound.
USB power:


Battery:


missing ground issue.


I tried to ground it different ways but sound was the same.


You say that the calibration procedure runs normally. Would you mind sharing the detailed debug output, though?

OK

PITCH CALIBRATION

Arudino Freq: 15996794.00

Pitch Set Frequency: 499400
Frequency tuning range: 480381 to 517726

DAC value L: 0 Freq L: -19005
DAC value H: 4095 Freq H: 18330

DAC value L: 4095 Freq L: 18322
DAC value H: 2085 Freq H: 289

DAC value L: 2085 Freq L: 287
DAC value H: 2053 Freq H: -78

DAC value L: 2053 Freq L: -75
DAC value H: 2059 Freq H: -4

DAC value L: 2059 Freq L: -7
DAC value H: 2059 Freq H: -7

VOLUME CALIBRATION

Volume Set Frequency: 460157
Frequency tuning range: 443864 to 475501

DAC value L: 0 Freq L: -16276
DAC value H: 4095 Freq H: 15351

DAC value L: 4095 Freq L: 15340
DAC value H: 2108 Freq H: -4223

DAC value L: 2108 Freq L: -4249
DAC value H: 2536 Freq H: 959

DAC value L: 2536 Freq L: 951
DAC value H: 2458 Freq H: 63

DAC value L: 2458 Freq L: 49
DAC value H: 2453 Freq H: 11

DAC value L: 2453 Freq L: -15
DAC value H: 2452 Freq H: -6

CALIBRATION COMPLETED

I also measured the frequency on testpoints, there is square signal: on w3 is 499.938Khz, on w4 is 460.754Khz. Is it normal?
w3:


w4:


Which software release did you install?

Now installed the latest release. I tried to install previous release but i had the same problem.

And a - more general - question, rather for my scientific curiosity: Why do people take the effort on them to build their own shields if you can get a readily built and tested shield including antennas for only 89€ ? Take the price of the materials and the hourly wage of an electronics technician into account for the hours you spend with it and you will see that you can never build it at such a low price.

I know that I will spend more than 89Є and of course I can just buy it. But it`s not interesting It`s like a some difficult Lego for me.

Posted: 8/20/2020 8:11:02 PM
Bayaneer

Joined: 8/20/2020

I also took some photos of my Arduino and shield, perhaps you will see something wrong.


Antennas: http://accordion.in.ua/ot/3.jpg
Arduino: http://accordion.in.ua/ot/4.jpg
http://accordion.in.ua/ot/5.jpg

Posted: 8/20/2020 8:17:24 PM
Mr_Dham

From: Occitanie

Joined: 3/4/2012

And about the arduino itself, is it an original one or a clone?

There is something about arduino clone in the FAQ :
"Q:  Why do I have strange tone vibration on low frequencies (like buzzing flies, around note E).

You may be using a Arduino clone with a cheap ceramic resonator instead of a quartz crystal. Replace your Arduino with an original Arduino. Or then try to replace the oscillator yourself as shown here. (Thanks to Sergey Anokhin for figuring this out.)"


Posted: 8/20/2020 8:46:36 PM
Mr_Dham

From: Occitanie

Joined: 3/4/2012

I was writting at the same time as when you posted the pictures.
Looking at them, it seem that the arduino is a real one(Check list here).
Maybe forget this track...

Posted: 8/21/2020 2:21:20 PM
Bayaneer

Joined: 8/20/2020

And about the arduino itself, is it an original one or a clone?

I bought it as an original one in a local shop but it`s Ukraine and all is possible I will try to replace the crystal.
I also used the smd Arduino clone with normal quartz crystal, but sound was the same again.

Posted: 8/21/2020 4:24:43 PM
Mr_Dham

From: Occitanie

Joined: 3/4/2012

I saw some counterfeit arduino, your one really look like an original one. I would not go in this way as a first attempt.

Calibration seems normal.

Which Timbre do you use (1st, 2nd, ... ,8th) ?
It doesnt sound (and doesn't look) like sine wave.
The richer it is, the more aliasing you will have in high register.

Original 1st timbre looks like sine wave and have normally very soft aliasing in the highest register as Thierry says.

Maybe it worth posting a sound sample / video with 1st Timbre if it is not the case of the 2 videos.

Out of curiosity: how do you get this "Atari Punk Console" sound at the begining of the second sequence ?


 

You must be logged in to post a reply. Please log in or register for a new account.