I would like to formally express my thanks to the Olivier Gillert and his Shruthi-Team.
One year ago I started to develop my Stereo Synthesizer “Wave 1”. At the beginning of the 90s i want building a digital synthesizer.
My first project was a 8Bit soundsampler from DOEPFER Musikelektronik GmbH in germany. My next step was a digital
synthesizer on the basis of a digital sound chip SAM9203 from dream company. But I could not finish the project, because
my development tools was not not professional enough.
Now I have found many good ideas and tools for the construction of a synthesizer on the internet. Great help was the hardware
of opichenettes shruthi-synthesizers. My “WAVE 1” Synth is a Stereo-Version of an Hybrid-Synthesizer and have two analog Filtes (SSM2044) and
VCA-Chips (SSM2164). For sound editing is a small built-in touch-scree.
My Projekt-Site has a lot of information for musicelectronics and DIY synthesizers.
Geetings from old germany Rolf
PS: The new forum is very nice But in my old posts are no more pictures. For this reason I have saved my old mutable posts from 4 years on my google cloud. Its free for everyone. Here is the link: https://drive.google.com/open?id=0BxbpDqwYdkvEMVBfd29JS1pTUk0
I’m guessing that you will be moving away from the devboard?
I would personally add 4 potentiometers or encoders for editing of on-screen parameters, since tactile feedback is nice, and more precise… Then again, that might just be me?
The PWM-Frequency is verry high and you can’t listening on the audioout of my synth. In the new Software-Update for my Synth, i have change the pwm-Signal to 8Bit and 125KHz.
As it has been discussed here before the Problem with high frequency PWM carriers may occur when redigitizin the signal. Try running your signal thru a (decent - not a Behringer…) compressor and you will see what i mean, it will compress even if theres no audible signal. On my old MX-8000 console i can just see those signals on the VU-Meter. Maybe you take a measurement and eventually think about filtering a bit more…
The most sensitive point is the VCA CV (you’re already using a DAC for the oscillators, good!).
Here, you have a generous amount of filtering in place: 100k / 33n (cutoff at 48 Hz) ; vs a 125kHz carrier, so about 60dB of attenuation. However, I am unsure about the 100k / 33n combination for the RC filter - to me it looks like this would lead to excessive smoothing of the VCA modulation (and hence a lack of punch in the envelopes).
The shortest time for attack-phase is 4msec and for release-phase is 2msec.
The longest attack-time is 20sec with attack-level 127.
The interrupt-timer in my Xmega-Prozessor for ADSR-Signals is 1msec.
I have looking for a simple and practical storage solution for the wave files and
Sound parameters in my AVR-Synth. I found the “ALFAT OEM Board” from
GHI Electronics (see pic 1).
By ALFAT board you can transfer datas on the XMEGA spi interface to SD & MMC cards and USB flash drives.
The maximum data transfer rate is up to 4 Mbytes per second specified.
ALFAT-concepts
A FAT file system for an SD card or USB stick requires a lot of resources on the microcontroller
in my synth. The SoC processor on the ALFAT module allows a fast and easy access storage via
UART, SPI or I2C interface.
Some features
Built-in 2-port USB Host controller.
FAT16 and FAT32 File system.
No limits on media size, file size or file/folder count.
LFN (Long File Name), licensed by Microsoft.
Friendly user interface through UART,SPI or I2C.
Programmable UART baud-rate.
Up to 8 simultaneous file access.
SD/SDHC card support, no 2GB limit. GHI electronics is an SD association member.
High speed 4-bit SD card interface.
Up to 4000 KBytes/sec file access speed SD cards/USB.
On Ambika I use ChaN’s FatFS code (I rolled my own FAT reader for the bootloader, though, because it has to fit in 4k) - it was very easy to adapt it to my project.