You got me there, I was thinking of FM as radio talkshow broadcast, and didn't include its vast other applications.
Otherwise, seems like any further typing generates more confusion, I shall refrain.
P.S. I have to admit I made another mistake above, regarding the definition of sampling - the reduction of a continuous signal to a discrete signal, so analogue sampling as I typed above - guess this is another mine. Anyone going to pick on this?
Otherwise, seems like any further typing generates more confusion, I shall refrain.
P.S. I have to admit I made another mistake above, regarding the definition of sampling - the reduction of a continuous signal to a discrete signal, so analogue sampling as I typed above - guess this is another mine. Anyone going to pick on this?
Last edited:
Bernie!
You're welcome!
First of all: PWM is not equal to ClassD. Not every ClassD circuit uses PWM.
ClassD is not a signal, nor a circuit, it's a category, it is neither analog, nor digital. A given circuit of this category can be discussed wether it is analog or digital. For example these are digital: http://focus.ti.com/lit/sg/slyt312/slyt312.pdf but many others are analog, like: LM4651, UcD, etc...
wwenze!
In connection with your post I must agree:
You're welcome!
First of all: PWM is not equal to ClassD. Not every ClassD circuit uses PWM.
ClassD is not a signal, nor a circuit, it's a category, it is neither analog, nor digital. A given circuit of this category can be discussed wether it is analog or digital. For example these are digital: http://focus.ti.com/lit/sg/slyt312/slyt312.pdf but many others are analog, like: LM4651, UcD, etc...
wwenze!
In connection with your post I must agree:
We may have been on different pages in our discussion then
Yes I know "Class D" is a category with the "D" an arbitrary letter assignment.I had in mind UcD amps and such when I asserted that it was analog. I myself have a pair of UcD monoblocks. May I assume that the mainstream Class D amps are analog, eg those based on ICE, Spectron etc?
Which Class D consumer amps do you know of are digital? Why would they be classified under Class D, or, what specific distinguishing feature warrants this?
no.
Any output that is a reconstituted version that comes from a chopped or switched waveform cannot be analogue
Why? Because you don't let it to be?
If it is not analog, then please tell me what is its resolution! What is the smallest change that is representable by PWM? The answer is: zero! Hence it is analog. (If it is a simple PWM.)
Bernie7!
I know, but you used the terms "PWM" and "ClassD" in the same meaning if I understand correctly.
I don't know them. I don't buy ClassD amps, I build them.
For example:
http://www.tactlab.com/Products/T2/T2.html
Because they are switch mode (=Class D) amplifiers. The output transistors has two states: fully open and fully closed. That's all.
Officially digital means digital
. If the manufacturer says it's digital, then it is (probably). Your earlier question:
Error correction is pointless in this aspect. Digital error correction can correct only the errors occured in digital signal, but in a chip the probability of any digital error is negligible. In some amplifiers (and in many sound processors) analog signal is digitized in order to be able to feed into the digital signal chain, but this has nothing to do with digital error correction. (Exept if somebody want to transmitt the signal.)
Last edited:
"zero" is clearly either an exageration or not true.
Which?
What is an inherent characteristic of an analogue transmission path?
The signal MUST continuously follow the original waveform throughout the transmission from input to output.
Can that ever be applied to any form of chopped signal representaion of an analogue signal?
No!
At least not as far as my imagination or current knowledge will allow me to follow.
A class D modulator translates amplitude into time.
PWM translates back time into amplitude. Resolution in the time domain is infinite (or limited by quantum physics). Resolution in the amplitude domain is limited by noise floor, and is actually worse than time resolution. Only digital modulators using quantized time may have a limited resolution, but analog modulators doesn't.
What is limited in translations between time and amplitude domains is frequency response. Above half the sampling frequency, components are mirrored back to below Fs/2.
Amplification using the time domain has its advantages, amplitudes get distorted easily (harmonics), time information doesn't (jitter).
PWM translates back time into amplitude. Resolution in the time domain is infinite (or limited by quantum physics). Resolution in the amplitude domain is limited by noise floor, and is actually worse than time resolution. Only digital modulators using quantized time may have a limited resolution, but analog modulators doesn't.
What is limited in translations between time and amplitude domains is frequency response. Above half the sampling frequency, components are mirrored back to below Fs/2.
Amplification using the time domain has its advantages, amplitudes get distorted easily (harmonics), time information doesn't (jitter).
AndrewT!
Simply telling that the other one is wrong shows no intelligence at all. Try to make any argument! And answer the question! You can say as small (positive) value as you wish, the PWM signal will be able to represent.
Continuity has two aspects: continuity in time, and continuity in value set! A signal can be fully discrete (sampled and digitized, Pulse Code Modulation), or discrete in time, but continuous in value set (sampled analog signal, Pulse Amplitude Modulation), or continuous in both domain (simple analog), or discrete in value sets, but continuous in time (quantized=stair signal).
If input signal is bandwidth limited, then you don't have to transmitt every time-points, you can recover the original signal from discrete samples (shannon's theorem). The sampled signal has the full information, and it is analog (despite the fact that it doesn't follow the original signal)! If these samples are not represented by amplitude of the pulses, but by the width of the pulses, the information remains exactly the same, so it is obviously still analog. (However it cannot be recovered perfectly by simple filtering.)
Simply telling that the other one is wrong shows no intelligence at all. Try to make any argument! And answer the question! You can say as small (positive) value as you wish, the PWM signal will be able to represent.
Continuity has two aspects: continuity in time, and continuity in value set! A signal can be fully discrete (sampled and digitized, Pulse Code Modulation), or discrete in time, but continuous in value set (sampled analog signal, Pulse Amplitude Modulation), or continuous in both domain (simple analog), or discrete in value sets, but continuous in time (quantized=stair signal).
If input signal is bandwidth limited, then you don't have to transmitt every time-points, you can recover the original signal from discrete samples (shannon's theorem). The sampled signal has the full information, and it is analog (despite the fact that it doesn't follow the original signal)! If these samples are not represented by amplitude of the pulses, but by the width of the pulses, the information remains exactly the same, so it is obviously still analog. (However it cannot be recovered perfectly by simple filtering.)
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.