Greetings fellow audio to synth enthusiasts!
Some of you are already aware that the ACO160 chips were first powered on only about two weeks before the first day of NAMM in January, and the firmware for converting the pitch information to MIDI was working only the day before the show opened. You can imagine how relieved I was that our demo worked out as well as it did, and I am thankful to everyone who helped to make it a success.
Before going into more detail on the ACO160 production timeline I want to announce that I am very proud to be presenting a paper on the fundamental frequency tracking circuit used in the chip at the 146th AES Pro Audio Convention in Dublin this March 20-23. My paper will be part of the P18 – MIR (Music Information Retrieval) session on the afternoon of Friday the 22nd. When the paper is published I will be sure to share it here for the benefit of those unable to attend the conference in person.
And now, the production news! The ACO160 introduced several improvements over the earlier ACO100, most importantly the new fundamental frequency detector and envelope follower based on dual peak detectors. As is often the case in chip development, the first version of this circuit was not completely without “features,” a.k.a. bugs. I was hard at work for the first three weeks of February tracking down and fixing these bugs and I’m happy to report that I finished by the next prototype chip submission deadline this week and will have ACO160EVKs available this July as promised, and ACO160 chips in production most likely a few months later.
Here is a summary of improvements I made to the ACO160 chip to make it production ready:
- Eliminated one source of high-frequency noise due to a mistake in layout that was fixed by simply swapping two signal lines. This will reduce the on-chip noise floor, increasing the dynamic range of the frequency detection circuitry and will also reduce jitter in the frequency output independent of the audio amplitude, eliminating the “note clusters” heard when the MIDI pitch is quantized to semitones.
- Corrected the timing of the SPI outputs to fix a setup time violation when communicating with the STM32F100C6 micro-controller we chose for the generation of MIDI output. This setup time violation resulted in the system entering an undesired state where the frequency words would all be improperly read by the uC and required a reset for the system to recover. We were relieved that most of you at NAMM (probably?) didn’t notice us doing this when it happened. 🙂
- Improved the balancing of switched capacitor clock coupling in the decay circuit of the peak detector. Unbalanced clock coupling was causing the envelope output to bottom out around 110mV. Amazingly this was traced to a mere 1fF (that’s 10-18 Farad!) of imbalance in capacitance from clock signals to the summing nodes of an op amp. This is why the bottom meter LED was on almost all the time (it turns on for envelope voltage higher than about 100mV). Fixing this should bring the envelope floor down to about 20mV.
- Disabled output of frequency data via SPI when the gate is low. This was not a “must have” but rather a “nice to have” that I decided to implement since I had the opportunity to make changes anyway.
- Increased the hysteresis in the gate comparator to eliminate some glitches I observed in the gate output for certain threshold levels.
- Changed the clock of the polarity detection comparator which compares the positive vs. negative envelopes to determine which one has higher amplitude (the ACO160 adaptively selects the polarity for frequency detection based on which polarity has higher amplitude); these envelopes used to be compared on the falling edge of the reference clock but are now compared, for obscure but proper reasons (ask me if you REALLY want to know) a fraction of a cycle after the rising reference clock edge. This was another “nice to have” rather than “must have” change.
If you read this far there is probably a good chance I can convince you to follow Second Sound on YouTube, Instagram, and/or Facebook to catch our latest demo videos as well as some exciting demo videos we have planned in the near future. And if you’re not already subscribed to our newsletter, please do so HERE.
I look forward very much to designing some revolutionary products with the manufacturers among you and hearing the musicians among you make sounds that you only dreamed of making until now.