What defines a digital or analogue circuit?

Discussions about anything analog, digital, MIDI, synth technology, techniques, theories and more.
User avatar
tallowwaters
Moderator
Moderator
Posts: 4998
Joined: Thu Aug 25, 2005 4:11 am
Gear: LC-MS/MS
Location: snake's belly in a wagon rut

Re: What defines a digital or analogue circuit?

Post by tallowwaters » Fri Jun 22, 2012 6:26 pm

Bitexion wrote:You guys need to stop going after the Person, and stay on the topic instead. Making a dozen posts about why some guy is wrong derails the whole thread.
Not when it's a single person derailing a thread. Feel free to keep your 2 cents.
Brains can be used like a "stress ball," but only once.

User avatar
samuraipizzacat29
Supporting Member!
Supporting Member!
Posts: 819
Joined: Mon Jan 31, 2011 6:40 pm
Real name: Nate

Re: What defines a digital or analogue circuit?

Post by samuraipizzacat29 » Fri Jun 22, 2012 6:34 pm

Ok, so hopefully this is my final long-winded response. Please bear with me.

First, let's explain my logic and what led me astray, and then we'll see the positive information we can gather accordingly. The first problem is this from the juno-106 service manual:
Image
Image

when you look up the datasheet for the 82c53 (previously linked) it states it's a cmos timer. Therefore, deductive reasoning says "if roland clearly intended to call that CMOS chip a DCO, then what makes the Korg Poly 800 CMOS chip any different?" The answer is, they're not, but there will be more on that in a moment. also, i made an assumption considering the waveshaper, but we'll get back to that.

The second thing that led me astray was my thinking on what makes a digital oscillator. My thinking was pretty strictly this:
a waveform represented in irregular bits output to a DAC
http://www.electricdruid.net/index.php?page=info.dds
supplemented by my working definition of what i'd consider digital spelled out rather nicely below:
rschnier wrote:A digital circuit is a type of electrical circuit where the voltages inside the circuit are interpreted, by whether they fall into two or more predefined ranges, as a means of representing discrete states. This discrete state representation is needed in order for the state of the circuit to represent a discrete numeric/count-type value (hence the term "digital" - as in the digits on your hands).

All digital circuits are analog(ue), but not all analog circuits are digital. The difference lies in whether or not the circuit is capable of representing discrete states via encoding of voltages or voltage ranges. (Or currents or current ranges -- which is just as valid, but designers usually think in terms of voltage ranges.)

Thus, a pulse/square wave is an analog wave, but if you define the top of the wave as "one" state and the bottom of the wave as "the other" state, then you're using the wave "digitally" -- that the amplitude of the wave is capable of representing one of two different numbers. Same wave in either case; it just depends on how it's being used and interpreted.
Deductive reasoning again in my head, "if the claim is that something like a poly 800 is a digital oscillator, then the definition of a digital oscillator says that there must be something to interpret it's state on the output end"
this isn't entirely true either....

so I went back to the beginning and started checking stuff over again. EUREKA! I realized what I did wrong. I made an assumption about what Roland calls the waveshaper in the service manual. I saw this:
Image
DARN YOU ROLAND! DARN YOU ROLAND! DARN MYSELF! DARN IT! DANG IT! POOP!
I fully knew the definition of a DCO: a digital clock feeds a frequency divider that resets a vco. But I was more concerned with what Roland called a DCO and comparing it to what Korg calls a DCO and all I got was :? I assumed that if roland is calling it a dco, that that's where the DCO is. and if they're calling it a waveshaper, then all it's doing is waveshaping. ERRRRRRR, NOPE! The "waveshaper" is the actual oscillator. That's where the VCO lives. The "dco" input is fed to a capacitor tied to the base of the transistor triggering an increase/decrease in voltage so that it oscillates. I feel like I knew this, but IDK why I didn't look at it more closely. Sorry if this is the essence of what caused the head bashing.
Just to verify my suspicions and what we're all saying (in the opposite direction), I checked out a poly 800 service manual:
Image
yep - the output of the cmos is definitely mixed together and fed to op amps. no vco there.....

So, what is this "thing" that Roland and Korg are both calling a DCO that is not actually a DCO? and why did they do that.....? solely for the purpose of confusing me and pissing you guys off? thanks a lot roland j**k and korg j**k.... "My life is on the line and you've got butterfingers!"
I went back to tom's page on the juno, and it didn't elaborate.
http://www.electricdruid.net/index.php? ... o.junodcos
the direct digital synthesis was a bit of a dead end (as we're not outputting to a DAC).
I went back to DCO, but the wave we're talking about is created digitally and doesn't "drive" anything else, so the definitions there don't suffice.
Then I went back to the "Numerically Controlled Oscillator" page and found something in the talk (previously linked). Now, this is obviously just people talking about stuff, but I think it's very relevant:
wiki wrote:A NCO and a DDS in my opinion are not equivalent. A DDS digitally synthesises a waveform (The Direct digital synthesis article even goes as far to suggest the waveforms final destination is a DAC). A NCO creates a periodic waveform of a given frequency and sample rate.
I think it is accepted that NCO's are an integral part of a DDS. However its not implied that the output of an NCO will ever see a DAC, or be part of a DDS system. e.g. The output may be used for further digital modulation, or as a phase reference for an another signal. The latter example I wouldn't consider to have any relationship to a DDS.
(this opinion is then agreed upon by 4 people, one of which is Tom from Electric Druid. I'm not citing him as a scientist, but as clearly an invested professional in the field)

sooooooooooooooo, a "Numerically Controlled Oscillator" outputs oscillations that could be said to have a digital use and an analog use, eh? That would work with what Roland says "the resultant at the output of each counter will be a rectangle of audio frequency". That also explains why the Korg implementation can just mix it and send it to the vca. which reminds me, if anyone can find a datasheet on that njm1405b or whatever is right after the opamps in the p800 schematic, I'd like to know.

so, the way I see it is this:
Direct Digital Synthesis - describes complex waveforms by sending bits of information to a DAC that decodes the information and outputs it according to its output resolution. an NCO may be used in the middle of the process as a frequency counter for digital operations.
Numerically Controlled Oscillator - this oscillator receives instructions via bits and can either output to a DAC or output them directly. If the output is direct, it could be referred to as analog output. If it's intended to be interpreted by a receiving device, it is a digital output.
Digitally Controlled Oscillator - an NCO sends clock information which is used to trigger a VCO
Voltage Controlled Oscillator - an output oscillation in potential (voltage) caused by an input potential of possibly irregular waveform, phase, and frequency.

WHEW! do you have any idea how hard it is to define a VCO without using language that could misconstrue it as digital? There's not really a definition like that out there that I could quickly find, so I hope I did well.....

on to the conclusions:
firstly, more obvious, definite conclusions. A Poly 800 oscillator is an NCO, a Numerically Controlled Oscillator. A signal created from something taken to be digital (A clock from a microprocessor for instance) goes through a logic function that changes its state, the output of which could be processed digitally or be used as an analog square waveform at audio rates.
second, both Roland and Korg knew they were misappropriating the ics used. Roland did it not to further confuse the issue at hand (as a DCO was a tough sell already, now I have to explain what an NCO is too?!?!). Korg did it because they figured "if Roland calls the ic itself the DCO, then we're going to do the same thing." but that disregards the fact that the clock is outputting directly, and not being used as the impulse to trigger a VCO, which is what we all define a real life DCO as. Additionally, Roland knew that the output wave could be used for audio output since they based the sub-oscillator on that and output it directly to the VCF/VCA, but they chose to send it to a VCO to "flavor" the signal, thereby making it imperfect.
What's more, Direct Digital Synthesis is always digital representations of COMPLEX waveforms represented in a digital wave (irregular) output to a DAC. If a Direct Digital Synthesis implementation outputs an alternation of high and low states, that implementation could also be called a Numerically Controlled Oscillator if the output of that alternation is intended to be a "perfect" square wave.

Here's some indirect conclusions, or rather questions for you to answer for yourself, or rather rather the musings of my head that may or may not piss you off. edit: i'm playing tennis in my head for the sake of arguing with myself :)
First: a Numerically Controlled Oscillator, in perfect operation, is the only oscillator the likes of which could be accurately described as "entirely digital" or "entirely analog" depending on the perspective you take. Or only partly digital or partly analog at the same time. Both "entirely" opinions are extreme. "entirely digital" is extreme to the right because it's widely accepted that "our ears can't hear digital!". yet, a square wave that is intended to be a perfect representation of high and low states in alternation is accurately described as both analog and digital. Alternatively, the position "it's entirely analog!" is extreme to the left as it disregards the fact that a microprocessor created the wave, that the wave is composed of high and low states, and that the wave could be used to express something digitally. Also, "it's entirely analog" disregards that there are no changes in frequency, phase, or waveform. So, I can say it's entirely digital but make a case for it being analog and vice versa :)

a few other questions. Let's say we're in a vacuum and we clock a VCO with another VCO. The clock VCO does not drift, the waves are perfect, and it doesn't go out of phase (we're in a vacuum here where whatever I say is true is true :) ). Is that implementation a VCO clocked by a VCO or is it a DCO? I know the world isn't perfect, I'm just asking "what if"?
Or, let's say we use an again perfectly stable vco sent to the input of an NCO (a cmos chip). or not perfectly stable, it doesn't matter...... Since an NCO is defined being clocked by something digital, or numeric, could we accurately call that either an NCO or a VCO? either one wouldn't be entirely true.
There's a host of reasons why something can or can't be digital and a (perfectly) square wave makes and breaks all the rules :)

I'm sorry if your head wants to explode, this is really fun for me.

anyway, I do have an analog oscilloscope, a plan b model 15 vco, a poly 800 and some time this afternoon....... hopefully those responses soon.

thanks for your patience, just trying to clear some things up.
Nate

User avatar
theglyph
Junior Member
Junior Member
Posts: 144
Joined: Fri Mar 31, 2006 6:25 pm
Location: Passing thru a BBD

Re: What defines a digital or analogue circuit?

Post by theglyph » Fri Jun 22, 2012 7:14 pm

samuraipizzacat29,

It shouldn't require that much thought/response when it comes to a relatively simple question.
Eurorack modular, Voyager w/351,MF101.102.104.105B, LP Stage, Juno-106, Siel DK600, Paia 9700, Nord Electro 2, Ensoniq MR76

User avatar
rhino
Supporting Member!
Supporting Member!
Posts: 2607
Joined: Sat Aug 02, 2008 5:00 pm
Real name: bill
Gear: keepers:
Kurzweil K2500x
Ensoniq TS-12
Yamaha SY-99
Alesis QS-8
Roland JD-800
Roland JX-10
Akai AX-80
Ensoniq SQ-80
Korg DSS-1
Moog Mini
Fizmo
Location: kentucky hills

Re: What defines a digital or analogue circuit?

Post by rhino » Fri Jun 22, 2012 7:17 pm

Image

Old guy wading into the deep part of the s**t. IMHO, the VCO part of the DCO is the green circle: The CV charges the cap in a linear (+/-) slope and the pulse to the xistor (from the digital counter) shorts and discharges it for the next cycle.

A true digital circuit has only 0's and 1's (their logic levels) involved in the workings. Intermediate levels are deliberatly surpressed and ignored (as noise).
An analog circuit makes an "analog" of a voltage or current - that is, the level of the signal represents a "picture" of the intended results.

With the usual "exceptions to every rule", oscillators used in synths fall into classes:
(1) True analog - MiniMoog
(2) Divide-down additive - Poly-800, most organs
(3) DCO - most vintage Rolands, Akai AX80
(4) Digital - Korg DW8000, Ensoniq SQ80
(5) Fully digital - most VAs - entire sound generation is done by a CPU.
NOTE: I am referring to the oscillators ONLY. All will have filter and level control which may be A, D, or combo. Also, digital signals must pass thru an D/A converter of some sort before being sent out as audio.


DCO's seem to fall into two classes also:
(1) VCF with constant digital correction of pitch - again,most vintage synths
(2) VCF with intermittant pitch correction (autotune) - SC SixTrak

Reached the top of my boots. Going back to shore.
When the wise man points to the stars, the fool looks at the finger.
- Confucius

User avatar
b3groover
Senior Member
Senior Member
Posts: 965
Joined: Thu Apr 10, 2008 7:41 am
Real name: Jim
Band: THEO
Location: Lansing, MI
Contact:

Re: What defines a digital or analogue circuit?

Post by b3groover » Fri Jun 22, 2012 7:43 pm

theglyph wrote: I define digital as software based sound generation/manipulation requiring a DAC to communicate with the outside world and analog as all other methods.
It doesn't matter how you define it. That doesn't change what it actually is. My estimation of this thread? Some people need to learn about electronics before engaging in arguments.

User avatar
nathanscribe
VSE Review Contributor
VSE Review Contributor
Posts: 2889
Joined: Fri Aug 11, 2006 1:03 pm
Location: The right side of the Pennines
Contact:

Re: What defines a digital or analogue circuit?

Post by nathanscribe » Fri Jun 22, 2012 7:50 pm

rhino wrote:the VCO part of the DCO is the green circle: The CV charges the cap in a linear (+/-) slope and the pulse to the xistor (from the digital counter) shorts and discharges it for the next cycle.
Yep, the voltage controlled part is that integrator. The cap charges, and is reset by the transistor, which receives its opening nudge from the clock pulse, which appears at a frequency that is determined by the pitch of the note being assigned to that voice.

Note also, the "range" IC before the integrator. The octave setting determines which of those resistors is used to feed the CV to the integrator - the higher the value, the slower the cap will charge.

The part to the bottom right of that is the waveshaper, which derives the pulse wave from the ramp.

User avatar
samuraipizzacat29
Supporting Member!
Supporting Member!
Posts: 819
Joined: Mon Jan 31, 2011 6:40 pm
Real name: Nate

Re: What defines a digital or analogue circuit?

Post by samuraipizzacat29 » Fri Jun 22, 2012 7:56 pm

rhino wrote:IMHO, the VCO part of the DCO is the green circle
Thanks for your input :) I have a hard time understanding the inner workings of the "actual" oscillation, I was merely referring to the physical action of the capacitor discharging a trigger mechanism to the base of the transistor. I knew what this looked like because I'm having a *(#)@#) of a time figuring out what's wrong with the triggering on one side of a CGS114 i'm building. The fact that they called the initial IC a "DCO" and the combination of other IC's the "waveshaper" is highly confusing considering the actual mechanism that the ic is doing is only triggering, just like we say in our definition of a DCO. I took for granted it was just being shaped, because they DO call it a waveshaper, but it's not.

thanks nathanscribe for further information on those parts too.... I thought i posted this and then got a notification of another post and realized I didn't....
rhino wrote:Also, digital signals must pass thru an D/A converter of some sort before being sent out as audio.
This is what I pointed out as not being entirely true but not entirely false either in the comparison above between Direct Digital Synthesis and a Numerically Controlled Oscillator (what you referred to as the divide-down circuits). There's little information on these other than this paper from intersil I found and what you can divine from the opinions on the wiki. if someone else has found a paper, website, chat, etc please share!

User avatar
nathanscribe
VSE Review Contributor
VSE Review Contributor
Posts: 2889
Joined: Fri Aug 11, 2006 1:03 pm
Location: The right side of the Pennines
Contact:

Re: What defines a digital or analogue circuit?

Post by nathanscribe » Fri Jun 22, 2012 8:14 pm

samuraipizzacat29 wrote:I have a hard time understanding the inner workings of the "actual" oscillation, I was merely referring to the physical action of the capacitor discharging a trigger mechanism to the base of the transistor.
That's not quite right - an "integrator" is basically where you have a current fed to an opamp (in this case) which has a capacitor across it - the cap charges as the current flows. The output is "read" from the output side of the opamp/capacitor, more or less is how I'd put it - your waveform appears at the pointy end of the triangle. This changing charge means you get a slope. Then along comes the timer to say "that's enough of that at this particular pitch" and trasmits a pulse to the transistor. The transistor turns on, short circuits the capacitor, and the capacitor empties, taking the charge to nothing and there you have, at the output, a straight edge. The transistor closes and the charge repeats. Cycle this and you have a sawtooth, or ramp, or whatever you want to call it.

The rate of charge of the capacitor is determined by a few things, but *basically* a higher voltage feeding the circuit will charge the cap faster.

In a traditional VCO, instead of that clocked transistor doing the resetting, there's another little circuit that measures the wave as it comes out of the integrator. When it reaches a particular level, *it* provides the pulse to open the transistor.

BTW, the capacitor labelled on that particular Juno diagram above is not the charging cap - it's a triggering cap, which in combination with the resistor connected to it just gives a short pulse to the transistor. That's not the complete circuit - it's worth studying the whole thing if you're interested. Take it piece by piece though.

It's also quite educational building such things on breadboard so you can try elements for yourself and make changes to see what they do.

Glad you've got your head around it now.

User avatar
samuraipizzacat29
Supporting Member!
Supporting Member!
Posts: 819
Joined: Mon Jan 31, 2011 6:40 pm
Real name: Nate

Re: What defines a digital or analogue circuit?

Post by samuraipizzacat29 » Fri Jun 22, 2012 8:30 pm

nathanscribe wrote:Glad you've got your head around it now.
yeah, not really :) I need to spend more time working with a scope and exactly what's going on in that circuit to understand how the terminology relates to what's physically happening.

But the action of a VCO was kind of besides my original point. My original point is best explained by these scope pictures:

first, the output coming out of a plan b model 15 pulse output
Image
as you can see, slightly rounded edges, but for the most part very square. The scope traces get faint on the verticals because there's essentially no interpolation between on and off.

The horrendously bad Poly 800 saw wave at 2 resolutions:
Image
Image

ewww, just ewww.

The equally bad pulse wave from the cmos chip:
Image
Image
Edit: I feel like I could have scoped this better, but it is the output out of the headphone jack. cutoff at 99... I don't feel it's important enough to rip the whole synth apart so I can get to the pcb, demonstrates the argument aptly enough.

in this case, what we're referring to as a digital synth is more accurately described as analog than the "ultra" analog vco that could be dern near described as digital. That is what i've been trying to point out since moment 1.

(along with all the other confusion of course) :)

User avatar
nathanscribe
VSE Review Contributor
VSE Review Contributor
Posts: 2889
Joined: Fri Aug 11, 2006 1:03 pm
Location: The right side of the Pennines
Contact:

Re: What defines a digital or analogue circuit?

Post by nathanscribe » Fri Jun 22, 2012 8:40 pm

... except, the Juno's oscillator is analogue - there's a continuously changing charge that gives a continuously variable output, as a voltage, that is viewable directly as a continuously variable waveform - that's analogue. The CONTROL is digital - the clock - hence, DIGITALLY CONTROLLED oscillator.

The Poly 800 does layer its octave-separated squares using opamps, and at that point we're looking at an analogue signal - but the generation of those square waves is carried out by logic circuits driven by a CPU and associated timing components, which makes them digital up to the point at which they become analogue. :-)

User avatar
samuraipizzacat29
Supporting Member!
Supporting Member!
Posts: 819
Joined: Mon Jan 31, 2011 6:40 pm
Real name: Nate

Re: What defines a digital or analogue circuit?

Post by samuraipizzacat29 » Fri Jun 22, 2012 9:01 pm

Ha, i'll play along. :) I knew I should have ripped everything apart and scoped the direct output of the chip. Since I'm not going to do that, you win this round. Unless you wanna come over and hold the camera :)

That still doesnt address the fact that the best, most accurate vco i and many others have is better described as a digital output than either one of them.

User avatar
nathanscribe
VSE Review Contributor
VSE Review Contributor
Posts: 2889
Joined: Fri Aug 11, 2006 1:03 pm
Location: The right side of the Pennines
Contact:

Re: What defines a digital or analogue circuit?

Post by nathanscribe » Fri Jun 22, 2012 9:07 pm


User avatar
samuraipizzacat29
Supporting Member!
Supporting Member!
Posts: 819
Joined: Mon Jan 31, 2011 6:40 pm
Real name: Nate

Re: What defines a digital or analogue circuit?

Post by samuraipizzacat29 » Fri Jun 22, 2012 9:48 pm

I hope you understand this is frustrating for me as well. Every time i ask a question someone either cant or doesnt want to answer, i receive the underhanded equivalent of "stop asking, youre clearly of inferior knowledge".

It would seem clear that your position is that a definiton of digital requires for a state to be high or low with nothing in between. My understanding of physics says that the bond that keeps current flowing is never entirely broke,n so you can say that all signals are analog. Conversely, every signal is defined by either having and electron or not, so aNlls signals are digital. I have repeated this until my ears bleed asking for someone to prove me wrong or adjust my thinking and I don't get anywhere. I keep posting because I think t's important to the task at hand.

prove to me that the flow of current physically stops and then starts at a higher value and I'll shut the heck up. Prove that the flow of current physically stops and then starts again in a digital application.

I'm not dragging this out because I want to be right, I'm dragging this out because I want to know why I'm wrong so that everyone can learn from my mistake.

User avatar
space6oy
Moderator
Moderator
Posts: 5401
Joined: Tue Apr 10, 2007 5:29 am
Gear: vimpat, citalopram & vitamin D.
Location: stuck in ohio.
Contact:

Re: What defines a digital or analogue circuit?

Post by space6oy » Fri Jun 22, 2012 9:55 pm

all i've got for this thread: :facepalm:

mute
Active Member
Active Member
Posts: 640
Joined: Thu May 08, 2008 6:51 pm

Re: What defines a digital or analogue circuit?

Post by mute » Fri Jun 22, 2012 10:26 pm

samuraipizzacat29 wrote:I hope you understand this is frustrating for me as well. Every time i ask a question someone either cant or doesnt want to answer, i receive the underhanded equivalent of "stop asking, youre clearly of inferior knowledge".
To everyone else it looks like you've been given several detailed answers (on a dead topic) and you refuse to accept them.

Locked