Roland Alpha Juno-2 pwm sawtooth

[TechnoManiac]

There is really nothing to understand about the aJuno chip which was used in many other synthesizers that could lead to any clearly audible difference.

There are some nuances that I haven't heard in other analog synths with DCO. I have already described these moments above. However, I did not see a reasoned answer. I'll repeat my question.

Please listen to these examples that I recorded from Alpha Juno 2:
https://drive.google.com/drive/folders/ ... N7i_TV5_aa

So...

File: 2 - LOW Notes - strange squeak on hight freq.wav

In this example, you can hear that the saw in the lower notes sounds with some additional high-frequency squeak, as befits some old lo-fi samplers. This accompanying high-frequency sound is not observed on the SAW of Juno-106

If DCO of Alpha Juno has an analogue waveshaper (waveform generator) - where does this high-frequency sound come from?

High-frequency sound not very loud, in relation to the useful signal, so to hear it, you need to make it louder. It can also be seen as a straight line on the spectrogram. The line changes on different notes. If you zoom in on the waveform, you can see that it is not smooth, but in the form of small steps, like a sample with a low number of bits

- - - -

File: 3 - PORTAMENTO on off - change saw spectrum.wav

A very digital nature of the portamento, in addition to the attack, the very nature of the timbre changes.

In this example, I alternately turn portamento on and off, and you can hear that the very character of the timbre changes, although it would seem that portamento should not affect the spectrum of the timbre, but only the pitch of the attack. But when the pitch is equalized, the timbre no longer sounds the same as without portamento. I can't understand the reason. It seems to me that such an artifact is difficult to explain in an generator with analog waveshaper. However, this is easily explained in digital oscillators: when a sample from the lower zone is transposed upward through portamento, it has many high harmonics that are usually cut off by the filter in each sample's own zone. A purely digital artifact that can often be heard in romplers when using portamento. That is, it’s as if samples from different zones sound.

Is this possible in a generator that generates a signal using the analog method by analog waveshaper? There must be a logical explanation for this

[aciddose]

You're linking audio from a youtube video, you do not own the synthesizer. I'm telling you I do own it, and I can not reproduce those in my testing. There is likely a collection of serious flaws in the author of the youtube video's hardware.

The synthesizer does not use a sample rate other than 12 MHz. Aliasing would not show up at this frequency at levels much above the noise floor except in exceptional cases. Most of the effect is likely due to non-linearity in the signal path such as the input to the OTAs used in the filter and amplifier.

Please don't use phrases like "as you can see, you can hear". You can either speak technically or just don't bother asking me for a response. I am deaf. I hear nothing. "It sounds like a ..." silence.

[TechnoManiac]

If it is, where is the DAC and the PCM data? There isn't any.
...because I have an alpha juno PCB in front of my face right now.

In the DX7 oscillator the quarter sine samples is located directly on the oscillator chip

[aciddose]

Nice alt account.

[TechnoManiac]

You're linking audio from a youtube video, you do not own the synthesizer.

Did you carefully read what I wrote?

Please listen to these examples that I recorded from Alpha Juno 2

Let's do it again:

I recorded from MY Alpha Juno 2

Is that more clear?

What does this have to do with the YouTube link? This was another example... but you can hear exactly the same distortions there too

[aciddose]

Did you carefully read what I wrote?

Only the first time, then I laughed.

[TechnoManiac]

This article covers the JU-, but not aJU-:
http://www.electricdruid.net/?page=info.junodcos

Some of the descriptions are a little off but close enough.

Yes... there are a lot of inaccuracies here.

Under the concept of DCO, Roland in the Juno-106 manual means ONLY a counter based on the Intel 8253 chip. This is a DCO oscillator. The waveshaper (SAW, PWM etc) based on the MC5534A chip is not a part of DCO!!! This is clearly shown in several diagrams of the Juno-60 and 106 manual. There is a separate counter (Intel 8253) called DCO, and then there is a separate waveshaper MC5534A. These are two different parts.

But in the chip DCO of Alpha Juno we see a combined waveshaper. In this case, the waveshaper is part of the DCO in one chip. Therefore, when I say “analog oscillator” here, I primarily mean analog waveshaper (aka Waveform Generator). The fact that the generator has a discrete resolution along the X axis related to the masterclock frequency, and the fact that it is not a VCO, is not even discussed, it goes without saying. This is what the DCO concept already assumes. But nowhere in the DCO concept does it say that the waveshaper must be analog, even in the first Junos he was separate from DCO

[aciddose]

I've operated under the assumption that the 1 and 2 are essentially the same with only a few additions to the 2. It is possible there are differences in the audio generation portions of the circuit for the 2 that would produce some effects that never occur in the 1 as I own.

I can assure you that your descriptions of these are extremely non-technical, demonstrate clear subjective bias and you seem to jump to conclusions without any supporting evidence.

Here is the best I can do without having evidence myself, since I am a technical sort of person and I'm not interested in subjectively judging anything without measuring it directly.

#1) "high frequency at low frequencies of oscillator output", are you certain you're measuring only the oscillator output?
#2) "portamento changes the timbre", are you certain the frequency and all other parameters with portamento enabled are exactly identical?

[aciddose]

A quick review of the service manual for the juno2 shows they are essentially identical.

[TechnoManiac]

Nice alt account.

are you confusing me with someone? )

#1) "high frequency at low frequencies of oscillator output", are you certain you're measuring only the oscillator output?

No....it's just a low note from the line output. The signal passed through the entire path of the synthesizer. I did not measure it directly from the chip output. Do you think that this could be interference penetration before the divider?

#2) "portamento changes the timbre", are you certain the frequency and all other parameters with portamento enabled are exactly identical?

I just take the same note, and I just turn the portamento on and off. In addition to the intended attack, the timbre itself changes. When the pitch is equalized, the note frequency is the same as without portamento. Well, at least that's how it should be. But the timbre of the wave changes. It seems as if two waves with different phases are superimposed. However, there is only one oscillator. I can't figure out where the parallel sound could come from.

.... you seem to jump to conclusions without any supporting evidence.

I haven’t made final conclusions... I’m just giving approximate probabilistic assumptions, and looking for a reasoned explanation why this might happen one way or another

[TechnoManiac]

...

[aciddose]

are you confusing me with someone? )

To be honest, at first I couldn't tell if you were serious or not given some things you've said, but, the thing that did make sense was this low-note high-frequency, which would be exactly as expected given either an integrator-based or DAC-based output. The real issue is without being able to both isolate and exactly control (bit-for-bit) the chip ourselves, it's very difficult to just guess.

#1) "high frequency at low frequencies of oscillator output", are you certain you're measuring only the oscillator output?

No....it's just a low note from the line output. The signal passed through the entire path of the synthesizer. I did not measure it directly from the chip output. Do you think that this could be interference penetration before the divider?

From my own testing, I do recall these tones but I've never managed to be absolutely certain what causes them. What you might notice is that given the same chip is outputting six voices at once ... you should test like this: Play multiple notes spread across the whole range of the keyboard. Transpose the preset and the key transpose as low as possible. Set the pitch bend range to maximum, set the pitch bend input to maximum negative. You should measure approximately 5.83~ Hz and the first harmonic may be at approximately 1024 Hz.

Now, the important thing is the timbre and the location of these harmonics depends upon which other notes are assigned to the other voices! So we can't say for certain one way or another what the source is, other than confirm that the other five voices interfere directly with the output of the one we're measuring. I'm in this case also not isolating the output because that's a pain and I don't see it as particularly valuable or necessary to do anyway.

Given the frequency ratios we can measure, we can attempt to guess at the "bits" that would exist to produce the harmonics that are phase-aligned and frequency-linked to the voice we're adjusting. This is easy to test, just wiggle the pitch-bender up and down and you'll see some harmonics track the oscillator frequency and some do not.

Since we know the ratio of those that track, we can determine the Nth harmonic (~175.64~?) and sort of guess at the number of bits, assuming this is quantization we're hearing ... only we'd expect something like 256 or 128, not 175.64!

#2) "portamento changes the timbre", are you certain the frequency and all other parameters with portamento enabled are exactly identical?

I just take the same note, and I just turn the portamento on and off. In addition to the intended attack, the timbre itself changes. When the pitch is equalized, the note frequency is the same as without portamento. Well, at least that's how it should be. But the timbre of the wave changes. It seems as if two waves with different phases are superimposed. However, there is only one oscillator. I can't figure out where the parallel sound could come from.

Again, it's hard to say I'll need to test this further myself, but, yes mine produces similar effects if I recall correctly along with clicks & pops from the filter or other modules. I'd describe the effect of timbre change as "interference", and it's likely the result of the different divisors the oscillator chip gets with portamento on or off. In addition, if the CPU is producing the integration (linear) portamento effect, this may be impacting the clock rate at which it can output address/data to the oscillator chip and strobe the data latch pin. This could dramatically impact the performance of the oscillator or even every module in the synthesizer due to changes in timing on the CPU alone.

.... you seem to jump to conclusions without any supporting evidence.

I haven’t made final conclusions... I’m just giving approximate probabilistic assumptions, and looking for a reasoned explanation why this might happen one way or another

Well, as I said to be honest I wasn't sure if you were even being genuine because you said a few very silly things. It's fine though, you might just be a bit confused and not as experienced. I've gotten to the point in messing with these sort of things that the first thing I know is, whatever I think at first, that is most definitely wrong! Even any conclusion is pretty much guaranteed to be completely wrong, or often we end up looking at things in completely the wrong way. So the best course of action is all about the "Philosophy of Science", which says evidence is our best bet (though never a sure one!)

https://en.wikipedia.org/wiki/Evidence# ... of_science

There is a lot more to discuss on why it's "never certain". Suffice it to say the hope is that by stumbling and crawling bit-of-evidence by bit-of-evidence (as if they were rock-climbing foot and hand holds) we can eventually climb up and over and get a clearer picture of things. It's quite common to see an amazing display of stupidity by those who misunderstand, talk about "proof" or "right" or "wrong". The reality is there is always another next higher perch, and we're always merely stuck in the next highest valley with our view obscured by yet another infinite collection of peaks. We can't ever see the sky, we're merely crawling through tunnels toward sparkles and glints of light through an infinite darkness.

[aciddose]

I suspect, although this is just speculation ... that it's very possible there is some form of quantization due to the fact the harmonic tracks with the oscillator frequency. Another explanation though is that this is merely the same frequency in another division step previous to the final division. There may be initially a division to set the semi-tone and fine-tune. This is probably followed by an additional division to set the oscillator "footing/range" parameter. So we're hearing not just interference from other voices, but also interference from another division of the master clock before the final division and wave-shaping produces the output waveform.

Great experience for understanding the effects of various pulse-shaping, gating and division circuits is with an old divider organ from the 1970s.

[aciddose]

In addition, I can note that I have never added additional digital/analog bypassing to the circuit. I've never replaced the PSU capacitors or electrolytics used for supply bypassing on any boards in my alpha juno 1. Therefore I would feel very safe to bet that some or all of these effects are related to those capacitors being badly worn out and decayed.

[aciddose]

Incidentally, one of my own analog synthesizers (designed & built myself) has a linear/log glide switch. By flipping this, it can impact the "sharpness" of the gate & CV signals in a very minor way and slightly offset the output pitch CV. So that has additional impacts on the trigger detection circuit and other circuits most likely due to the larger capacitor used. In my MIDI to CV software implementation of an identical switch, it produces different effects but this time due to differences in the quantization and coefficients. Both the analog version and digital version of the same circuit are flawed in unique ways!