Are BBD chips considered Analog or Digital, or even neither or both?

[bushidov]

I used to believe they were analog, but I have also read they are digital, and not analog, and in one article, that they are neither. What say those in this forum?

Claiming it is digital:
https://www.thegearpage.net/board/index.php?threads/why-they-call-bbd-analogue-when-it-is-not.2244153/

Claiming it is neither:
https://neunaber.net/blogs/neunaber-audio-blog/13830765-the-analog-myth-chorus

I am still not convinced that it isn't analog, just due to my experience with microcontroller and microprocessor electronics, but I also don't have a solid answer against it.

[Fancy Lime]

Is it that time of year again already?

BBDs are analog but operate in discreet time intervals. Digital also operates in discreet time, hence the confusion. But what makes digital digital is not the discreet time (although that is one of the prerequisites for it to work) but the encoding of signal levels in discreet values that can be represented in binary code.

Andy

[bushidov]

That's what I thought too, however folks that argue they are digital state that is irrelevant, as the act of sampling time means it definitely isn't analog, and in some instances necessitate that it is digital.

I don't agree with those arguments, but I don't feel I have a full enough understanding to say the opposing view are incorrect. So I am here to see if I can get a better understanding from those who know much more than I.

[Fancy Lime]

A digital signal can be expressed as a string of numbers, or digits. Hence the name. You can read a digital signal and print it out as a long string of 1s and 0s. That is not the case with bucket brigades. The values are not digitized. A BBD is basically a lot of sample-and-hold circuits stringed together, each consisting of a transistor and a holding cap and connected to it's neighboring elements in such a way that the held voltage is passed along the chain. Sampling of analog voltages does not make it digital if the voltages are not subsequently digitized. In a BBD, each step that happens is analog. Current goes from one cap to the next, millions of times a second. I suspect that to some people BBDs "feel" digital because they sample a signal at a rate higher than the audio band. But at no point are there any 0s or 1s, just analog signal levels.

HTH,
Andy

[antonis]

I think we mix up "digital" with "quantized"..

[bushidov]

I suspect that to some people BBDs "feel" digital because they sample a signal at a rate higher than the audio band. But at no point are there any 0s or 1s, just analog signal levels.

I agree. That's been my thoughts on it as well.

I think we mix up "digital" with "quantized"

I would agree, but could you elaborate?

Is there any weight to the statement:

"Analog" means analogous to another continuous signal. In audio, this means analogous to sound pressure. Once you quantize the signal—in either time or voltage—it is no longer analogous to sound pressure

Which in turn uses the definition of Analog Signal here, as its reference:
https://en.wikipedia.org/wiki/Analog_signal

[Mark Hammer]

Consider your average sample-and-hold circuit, as found in the Maestro FSH-1 or similar pedals.  A clock turns the gate of an FET on and off, storing the voltage from a noise source at that moment in a capacitor, that is isolated enough such that the stored voltage remains constant for a while.  The noise source is not digital.  The op-amp or timer-based clock turning the FET gate on or off is not digital, and neither is the cap that holds the sampled voltage.

The sampled noise voltage is not quantized in any manner; quantization being a "rounding" up or down to some specific step within a scale.  The noise voltage at time X is the voltage, the range of which could be anything, with no particular "steps" defined within the range.  The noise was continuous.

BBDs are essentially sample-and-hold devices, that use an audio signal as their "noise source", and tiny storage caps to "hold" the sampled voltage.  A network of FETs is used to allow the sample to be moved along to a second, 3rd, and nth storage cap.  The absolute voltage in the storage cap cannot be held for very long before it drains, and must be moved along, from stage to stage to the final output, before that drain happens and the stored voltage stops resembling the original sample.  Were it a digital sample, what would be stored and passed along is the code for the sample -something that makes its way to some eventual output without any error, no matter how long you hold onto it.

[Josh?]

Here's a funny thing to complicate matters even more: what's the smallest possible unit of electric charge? It's an electron with a charge of "-e" (or if you wanna get nit-picky, a quark with a charge of 2e/3 or -e/3), so if current is measured in terms of how much charge goes through something per some amount of time, it's already quantized in terms of number of electrons, but not in terms of time. So if you were able to quantize that time exactly, would you be (technically) able to represent current in binary?

[antonis]

You should be proud for turning "Building your own stompbox" section into "Philosophical frying pans" respective one, Erik..

[Keppy]

"Analog" comes from the same root as "analogy," meaning using one thing to represent another. Like using voltage to represent sound pressure. Or like using bits to represent numbers which represent voltage.

My point is that by any reasonable definition, digital IS analog, "analog" doesn't mean what we use it to mean, and words are stupid.

[Fancy Lime]

The definition in the Wikipedia article is a bit misleading. It states at the time of me writing this:

"An analog signal is any continuous signal for which the time-varying feature of the signal is a representation of some other time-varying quantity, i.e., analogous to another time-varying signal. For example, in an analog audio signal, the instantaneous voltage of the signal varies continuously with the pressure of the sound waves."

The crux for BBDs is "instantaneous". Some people seem to mistake this mentioning of "instantaneous voltage" to mean that the output follows the input instantaneously. Analog signals are not generally instantaneous in that sense. As soon as there are caps in the signal path, there is a phase delay of one kind or another. The output still moves analogous to the input but not instantaneous. The time-dependence of one part of the analogy need not be identical to that of the other, as long as they both are time-dependence in some way. Which is the case in BBDs. A digital signal level is not time dependent. 0011010110110111 stays the same until the end of time, the digital sample rate only defines when it's time to switch the output to the next value.

Andy

P.s., let's postpone the "is time quantized?" discussion for a moment, shall we

[Fancy Lime]

...doesn't mean what we use it to mean, and words are stupid.

That describes most debates on the internet pretty accurately, I'm afraid...

[antonis]

"Analog" comes from the same root as "analogy," meaning using one thing to represent another.

If I'm allowed to define Greek root words, I'd say that "analog" means:

1. corresponding/respective/equivalent
2. appropriate/suitable/proper
3. similar/alike/of a kind
4. which increases or decreases in a similar or inverse manner to another size

Many people confuse "analog" with "continous", hence segragate digital from analog..

[Mark Hammer]

The signal source is continuous.  It just gets interrupted a lot.

[iainpunk]

Here's a funny thing to complicate matters even more: what's the smallest possible unit of electric charge? It's an electron with a charge of "-e" (or if you wanna get nit-picky, a quark with a charge of 2e/3 or -e/3), so if current is measured in terms of how much charge goes through something per some amount of time, it's already quantized in terms of number of electrons, but not in terms of time. So if you were able to quantize that time exactly, would you be (technically) able to represent current in binary?

electric tension is divisible in smaller increments than quark charge, since super position can cause the charge of quarks to be divided between multiple places, we had a whole lecture on the topic once, it was really interesting and confusing, quantum physics is a bit above my level of comprehension.
aslo, electrons aren't anywhere most of the time, they are actually clouds of chance most of the time...

cheers

if you actually are Josh from JHS, check this out:
https://www.diystompboxes.com/smfforum/index.php?topic=126052.msg1204019#msg1204019

[amz-fx]

The signal source is continuous.  It just gets interrupted a lot.

Each sample is stored and actually used twice. Once going to Out1 and then delayed by an additional stage before going to the second output (Out2). The two output signals are combined equally into the final audio signal through a reconstruction filter.

If you are using the same sample twice, at different moments of time, is it continuous?   



Best regards, Jack

[iainpunk]

The signal source is continuous.  It just gets interrupted a lot.

Each sample is stored and actually used twice. Once going to Out1 and then delayed by an additional stage before going to the second output (Out2). The two output signals are combined equally into the final audio signal through a reconstruction filter.

If you are using the same sample twice, at different moments of time, is it continuous?   



Best regards, Jack

i guess its just as analog then as tape delay, since you use the signal more than twice if it has feedback or multiple read heads.

cheers

[amz-fx]

...never mind...

[Rob Strand]

That's what I thought too, however folks that argue they are digital state that is irrelevant, as the act of sampling time means it definitely isn't analog, and in some instances necessitate that it is digital.

I don't agree with those arguments, but I don't feel I have a full enough understanding to say the opposing view are incorrect. So I am here to see if I can get a better understanding from those who know much more than I.

Every signal that you can measure on an oscilloscope is analog.   It exists in the real world.

Digital requires that something is digitized.  That means converting it to numbers (which can be stored  and used later).

One blurred boundary is the signals we would measure on a digital circuit.    The signals themselves are analog.   The interpretation of the meaning of the signals is digital.   The signal becomes digital when we assign a meaning say value 1 when the voltage level is greater than 1V and value 0 for a voltage level less than 1V.    After that the specific levels have no other meaning. For example we represent the state of something with a digital value.  That could a binary value with one wire or a multi-state value with many wires.      The analog interpretation of the signal is it has many values  all have the meaning of the level of the signal at a given time.

There is a special pigeon-hole for *IC's* which process analog and digital signals, it's called mixed signal.  It's a grey box 

Sampling is a whole different issue.   The signals are analog.   If we take an analog waveform and sample and hold it's value the resulting signal at the output of the sampler is analog.    We can put an oscilloscope on it.    There is no digital interpretation until it is converted to a number.   Analog signals tend to be continuous.  Even when output of a sample jumps from one sample to the next, in the real world the signal on an oscilloscope will pass through all values between the start-point and the end-point.  It might do it quickly but it is still continuous.

[niektb]

I'd say that a chip is fully digital when it processes a digital signal: a signal that has been sampled at discrete intervals and quantized.

In this sense, Logic gates are not fully digital (they quantize signal to be above their threshold) (unless they have a clock input) and neither are BBDs (because the signal is not quantized)