About 7 years ago, right after I first met a good friend of mine, he asked if I could fix a MIDI controller for him. It was a Waveidea Bitstream Pro.

I opened it up, and ordered the PIC16F887 I thought was the right chip. After that didn’t work I realized I should have purchased the PIC16F877 instead. I had ordered the wrong part.

Well, I just made the same mistake again. I ordered a dsPICFJ64GP804 instead of a dsPIC33FJ32GP304 while fixing my Hertz Donut MKII by The Harvestman.

I had to do a little debugging and got the attached image as the output from the part of the circuit that handles the knob input values.

I learned a lot about how the board was designed though so it wasn’t a complete run around. I ordered the correct part and hopefully will have it working by this time next week.

In other news, I was approached by someone asking me whether I could write USB interface firmware for USB controlled sound cards! Well, I’m already working towards that goal… but I didn’t have to write a single line of it! The PIC32MX line of chips comes with USB example code for doing just that… but it’s messy code and doesn’t actually do the conversion of the audio stream off of the USB endpoint into PCM that can get shoved out over I2S.

However, I ordered a few extra Wolfson CODEC chips for the Hertz Donut MKII repair anyways and since my SPI 12bit ADC/DAC conversion is failing at high frequencies I’ll probably end up using some of the same architecture as the Hertz Donut. It’s interesting that I’ve converged on pretty much the same choices as The Harvestman in my search for a digital signal path that will let me build my sequencer/cv recorder…

The Harvestman: I swear, I’m not ripping off your ideas or designs!

I don’t know jack shit about DSP design anyways either…

This is the signal going into the ADC pin on the dsPIC. All of the knobs are turned down on the front of the module except the main osc’s pitch. That peak is that knobs value.: