YM2612 & distorted sound output
Posted: Wed Jun 10, 2020 7:30 pm
Hi,
Does anyone have any experience to share regarding the output waveforms of the YM2612/YM3438 ?
I've built a hardware synth based around this chip and I'm seeing unexpected distortion on the audio output, even with nothing (other than a scope) attached.
For example, I've turned everything 'off' except operator 1 on channel 1. Algorithm is 7 (all operators in parallel). TL=0 (max volume), AR=15, DR1=14, DR2=14, RR=4. All the other registers for all operators and channels are set to 0 (apart from the TLs which are 127 for minimum volume). So no AM, FM, detune, MULT etc.
With a single operator I would expect the output to be a pure(ish) sine wave, with the amplitude controlled by the envelope, and I've verified this using a VST version of the YM2612 by setting the voice parameters to be the same. And I see a pretty good sine wave.
But the output of my chip looks quite different. The envelope is similar but it looks almost like half-cycle amplitude modulation of a higher frequency sine wave is going on. And if I buffer the output and feed it into an amp it sounds much harsher than a sine tone.
My first thought was that my 'write' function to the chip wasn't correct but I've checked it against online projects. And I can see the envelope change on the scope as I tweak the values. I'm reading the BUSY bit and waiting for that. So I'm reasonably sure I'm writing to the chip correctly.
I've tried two chips now and they both behave more or less the same. Anyone have any ideas please ? Could I just have damaged chips ?
Thanks in advance.
D.
Does anyone have any experience to share regarding the output waveforms of the YM2612/YM3438 ?
I've built a hardware synth based around this chip and I'm seeing unexpected distortion on the audio output, even with nothing (other than a scope) attached.
For example, I've turned everything 'off' except operator 1 on channel 1. Algorithm is 7 (all operators in parallel). TL=0 (max volume), AR=15, DR1=14, DR2=14, RR=4. All the other registers for all operators and channels are set to 0 (apart from the TLs which are 127 for minimum volume). So no AM, FM, detune, MULT etc.
With a single operator I would expect the output to be a pure(ish) sine wave, with the amplitude controlled by the envelope, and I've verified this using a VST version of the YM2612 by setting the voice parameters to be the same. And I see a pretty good sine wave.
But the output of my chip looks quite different. The envelope is similar but it looks almost like half-cycle amplitude modulation of a higher frequency sine wave is going on. And if I buffer the output and feed it into an amp it sounds much harsher than a sine tone.
My first thought was that my 'write' function to the chip wasn't correct but I've checked it against online projects. And I can see the envelope change on the scope as I tweak the values. I'm reading the BUSY bit and waiting for that. So I'm reasonably sure I'm writing to the chip correctly.
I've tried two chips now and they both behave more or less the same. Anyone have any ideas please ? Could I just have damaged chips ?
Thanks in advance.
D.