OK the original question was is class D digital? The answer is no. "Class D may or may not be digital" is wrong it is never digital. The signal may be converted to digital for whatever reason but it is converted to analog to amplify. Switching is not "digital" Digital is a code of ones and zeros that can be processed but not amplified...yet. As I have said there have been experiments but nothing commercial has come of it. That is my last word on this. My other comments were of a simple and general nature and as such could be picked apart because the are always exceptions and low level artifacts that are contradictory. The last posts were of much better accuracy than some of the earlier nonsense.
Enjoy the benefits of a premium membership!
Upgrade Account
-
TalkBass has been independent since 1998. Add your voice.
Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
Join free Log in
Want zero display ads or expanded classifieds tools? Compare plans.
Class D popping up everywhere. Innovation?
- Thread starter Volume4
- Start date
bongomania
Commercial User
Did you notice how almost all of those last "better accuracy" posts made a point of saying everything you said was wrong?
- Dec 18, 2000
- 13,201
- 340
- Disclosures
- greenboy designs: fEARful, bassic, dually, crazy88 etc
If you mean the pre-amp input on the Crate Power Block - it has diodes for clipping. Seems little old diodes are used for the key ingredient in many soft clipping circuits.
For my tests, I always bypass the preamp via whatever inputs are available. IIRC the Power Block had some sort of power amp input jacks. I should have compared with and without my dummy load.
FYI, A digital system uses discrete (discontinuous) values. ON or OFF. Any switching system is considered digital. A class d amp can be considered digital. In fact it is very similar to a 1-bit digital to analog converter. It can not be analysed, simulated, or treated like a continous analog system.
True that the power section of a switching amp does not use binary numbers to represent values, but it still uses two discrete states to produce its output.
True that the power section of a switching amp does not use binary numbers to represent values, but it still uses two discrete states to produce its output.
I think what you're talking about -- a system with two discrete states -- is more commonly referred to as "bistable" rather than "digital." An old fashioned home thermostat is bistable but not digital.
But don't give up on simulation just yet. What program have you tried? Here's a schematic and analog simulation result for a 1-bit delta-sigma modulator.
Invalid Link Removed
Now, here's where it gets interesting. When the author of this circuit posted it at the www.diyaudio.com forum, someone else pointed out that it still works if you remove the 1-bit quantizer, but it becomes a self-oscillating Class-D amplifier -- a purely analog circuit! The output pulse train will still have discrete values, but the pulse timing is continuous. I performed that exercise using the analog simulation program LTSpice, and it works just fine. Several people have published analyses of the self-osc design concept. At least one commercial bass amp works this way.
I'm not a guru, but AFAIK the thrust of contemporary Class-D design is to take that simple concept and make it work in the real world. Only a few people have done this successfully. Designs with more elaborate digital innards have largely fallen by the wayside.
In my view, the discussion of whether Class-D amps are "digital" should revolve around real world circuits, otherwise throwing that buzzword around illuminates absolutely nothing. Showing real circuits eliminates any need to quibble over terminology. In the words of Thelonious Monk, "Let's not talk about it, let's play it."
If you want to be very fussy about the matter, class-D is still analog.......
What happens is an ANALOG translation from "analog amplitude" to an "analog time duration", and back again.
Analog signal amplitude is directly transformed to analog pulse width (PWM), and then transformed back to a continuous analog signal at the output. The signal representation is at all times analog.
There IS a discrete time component to this, but it is most similar to the old analog type 'flanger" circuits, which functioned as an analog delay line even though they did 'sample" the signal. The signal representation is always analog, and never represented as any sort of digital "word".
Even the two-state nature of the signal is not a "1" and "0" stream, since the analog duration of the "1's and zeros" is the actual signal.
You CAN do PWM with a digital input, directly transforming the digital signal in any form it comes, into analog pulse widths and then into a continuous analog signal. In fact John Ulrick and the folks at Spectron which I mentioned earlier, have patents on methods of doing just that.
BTW, some of the Ampeg/SLM class-D units were of the "self-oscillating" or "natural sampling" type. There are considerable advantages to that as opposed to the clocked sampling type.
What happens is an ANALOG translation from "analog amplitude" to an "analog time duration", and back again.
Analog signal amplitude is directly transformed to analog pulse width (PWM), and then transformed back to a continuous analog signal at the output. The signal representation is at all times analog.
There IS a discrete time component to this, but it is most similar to the old analog type 'flanger" circuits, which functioned as an analog delay line even though they did 'sample" the signal. The signal representation is always analog, and never represented as any sort of digital "word".
Even the two-state nature of the signal is not a "1" and "0" stream, since the analog duration of the "1's and zeros" is the actual signal.
You CAN do PWM with a digital input, directly transforming the digital signal in any form it comes, into analog pulse widths and then into a continuous analog signal. In fact John Ulrick and the folks at Spectron which I mentioned earlier, have patents on methods of doing just that.
BTW, some of the Ampeg/SLM class-D units were of the "self-oscillating" or "natural sampling" type. There are considerable advantages to that as opposed to the clocked sampling type.
Circuits of this type are often labelled "mixed signal". Its hard to argue that a bucket brigade delay is either entirely analog or entirely digital. Its a discreet time system, with a continuous amplitude.
Class D amplifiers could be argued to be like this. Though some class D designs have many fewer digital-ish bits than others. Some have none at all.
It would even be possible to have an almost entirely digital signal path after an A/D conversion, with a digital PWM controller, though I can't imagine why you'd want to.
Well take any 3rd year linear systems class in when you get your BSEE and you would understand (if you passed 
). Actually a smoke signal (smoke digitally modulated with a blanket) or even Morse code (got these from wiki, sorry) are considered digital. Look up the strick definition. A digital system has discrete time points when things happen and the mathematics of continous linear systems do not apply. Period. You can not apply linear equations or use continous time math ( for example Laplace transforms, etc) or integrals. There are digital analogies you must use instead. These are use for the simulations. The simple fact that the digital switching speed is an issue and/or the artifacts are filtered-out should be a huge clue. A digital system may have any number of discrete states. Binary is the term related to two digtal states. Also Bi-stable means digital. There is no stable condition between states. And yes, a class D is a "mixed signal" design and it implies both digital and analog circuits exist in the same system, however usually it is used to refer to an integrated circuit that uses both analog and digital circuits. PWM is a digital modulation technique. A pulse train of a variable rate modulates a binary switch (transistor).
Oh and BTW anything we actually implement today has both digital and analog properties since the switching times are greater than zero. However even though a digital system has analog properties, it can not be considered analog for analysis.
). Actually a smoke signal (smoke digitally modulated with a blanket) or even Morse code (got these from wiki, sorry) are considered digital. Look up the strick definition. A digital system has discrete time points when things happen and the mathematics of continous linear systems do not apply. Period. You can not apply linear equations or use continous time math ( for example Laplace transforms, etc) or integrals. There are digital analogies you must use instead. These are use for the simulations. The simple fact that the digital switching speed is an issue and/or the artifacts are filtered-out should be a huge clue. A digital system may have any number of discrete states. Binary is the term related to two digtal states. Also Bi-stable means digital. There is no stable condition between states. And yes, a class D is a "mixed signal" design and it implies both digital and analog circuits exist in the same system, however usually it is used to refer to an integrated circuit that uses both analog and digital circuits. PWM is a digital modulation technique. A pulse train of a variable rate modulates a binary switch (transistor).Oh and BTW anything we actually implement today has both digital and analog properties since the switching times are greater than zero. However even though a digital system has analog properties, it can not be considered analog for analysis.
OK, so what do you see as an oscilloscope trace when you connect the leads to the output of a Class-D amplifier? Let us, for simplicity, assume you use an old-fashioned ANALOG oscilloscope. The input signal could be for instance a 1 kHz sine wave, with 100 mV amplitude. Inquiring minds want to know...
+1
For example, if we could see output signal of an Orange AD200 (all tube) and also the Bass Terror (same preamp but Class D).
That would be interesting....
With an o-scope you would see what looks like a 1KHz sine wave. And for your eyes and and most likely your ears, thats all fine and good. But..with a spectrum analyzer you would see artifacts of the PWM process attenuated by the natural sync function and the implemented filter plus speaker response. These artifacts should go unheard by all except maybe a lapdog. These distortions are considered out-of-band. Of course an analog amp has different types of distortions (like crossover, non-linear, etc). Both digital and analog types of amps can be forced to clip by being driven too hard, either at the pre-amp stage (too much input) or at the power-amp stage (too much signal OR load for the output devices and/or the rails).
Have we decided yet how many analog angels can dance on the head of a digital pin?
This thread lost its relevance in arguing about details some time ago....
This thread lost its relevance in arguing about details some time ago....
I guess you are right but relevance is relative
. I'll bow out now since the pay is terrible ...IMHO there are many non-engineers here who have an impressive understanding and "feel" for electronic systems so I just wanted to help out.
Good point about checking the output with a spectrum analyzer. Or, you could take the rig, including your cab, into an anechoic chamber and run a frequency sweep at different power levels and see how the spectral and distortion characteristics change with volume. That could give us some comparable data from one rig over to the other. But, what actually would matter the most, is the audience response playing live. That's a complex equation that cannot be measured objectively.
That's the key - same signal to the speaker. That's not going to change - speakers are going to need an analog signal for years to come.
Sound wise - Class-d amps are the same as other class amps except for artifacts that can't be heard but would be measured with unsuitable test gear.
Can't stress enough that the amp class should be black boxed. And just the audible measurements considered for a sound wise comparison. And nobody should care about the class.
Consider the weight, efficiency, reliability, price. Class-D, PWM, PDM, amps are here to stay.
$0.16 a watt retail and falling.
