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 A, B and C amps are all traditional amplifiers with a linear gain relative to input signal. Class D is a different beast and uses pulse-width modulation to trigger (at very high rates, typically 10x the input frequencies) a binary switch. Because of the binary nature of the switch, to me and many other engineers, this qualifies as a digital system.
To be semantically accurate - it is not a "digital amp," technically it is a "digital switch circuit" that is being used to mimic what we have traditionally called an "amp."
There are several types of analog to digital convertors (ADC's) which use a ramp to first convert an input amplitude to a time quantity, a variable width pulse. Up to this point in the process, there is some similarity to a class D amplifier. Both convert voltage to time.
So far though, the ADC is incomplete. Next it needs a counter and a high speed clock signal to drive it. The counter begins counting at the beginning of the pulse and continues counting for it's duration; at the end of the pulse, the counter stops and it's value is stored. We now have a digital representation, a numerical value, that corresponds to the input amplitude.
The actual analog to digital conversion is not in the voltage to pulse width conversion; that is just a conversion from one analog quantity (voltage) to another analog quantity (time). The actual conversion is in the measurement of the pulse width by the counter. And it is this critical part of conversion to digital that is missing in the class D ampifier.
Also, the high and low states of the class D modulated pulse are not properly zeros and ones; they are logic levels. The numerical values of '0' and '1' are an interpretation of a logic level. Furthermore, you can't store the stream of highs and lows in a digital medium i.e., computer memory, as you could a true digital encoding. The bit stream (if you want to call it that) looks the same regardless of the class D amplifier input:
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 ...
Now if you had a time value for when each transition from 1 to 0 (or high to low) occured, you could store that in memory. But to do that you would first need an analog to digital convertor, and as explained above, it is simply not present in a class D amp.
And finally, a true digital encoding can be copied and reproduced endlessly with no degradation. You can't do that with the stream of width modulated pulses. You could record them, but it would still be an analog recording. The time position of the pulse edges would become increasingly blurred with each successive copy.
-
So far though, the ADC is incomplete. Next it needs a counter and a high speed clock signal to drive it. The counter begins counting at the beginning of the pulse and continues counting for it's duration; at the end of the pulse, the counter stops and it's value is stored. We now have a digital representation, a numerical value, that corresponds to the input amplitude.
The actual analog to digital conversion is not in the voltage to pulse width conversion; that is just a conversion from one analog quantity (voltage) to another analog quantity (time). The actual conversion is in the measurement of the pulse width by the counter. And it is this critical part of conversion to digital that is missing in the class D ampifier.
Also, the high and low states of the class D modulated pulse are not properly zeros and ones; they are logic levels. The numerical values of '0' and '1' are an interpretation of a logic level. Furthermore, you can't store the stream of highs and lows in a digital medium i.e., computer memory, as you could a true digital encoding. The bit stream (if you want to call it that) looks the same regardless of the class D amplifier input:
1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 ...
Now if you had a time value for when each transition from 1 to 0 (or high to low) occured, you could store that in memory. But to do that you would first need an analog to digital convertor, and as explained above, it is simply not present in a class D amp.
And finally, a true digital encoding can be copied and reproduced endlessly with no degradation. You can't do that with the stream of width modulated pulses. You could record them, but it would still be an analog recording. The time position of the pulse edges would become increasingly blurred with each successive copy.
-
on page 3 and still reading this ridiculousness.... WHY DID I CLICK ON THIS THREAD
Poor guy has no friggin idea what he walked into here.
Link Removed
Right, but "RMS" is never correctly applied to watts, only to voltages and currents. "RMS watts" is an incorrect term that we've become stuck with. What the FTC standard really calls for are "continuous average watts". They're typically computed from an RMS voltage, but that doesn't make them "RMS watts".
The best explanation I've seen so far is here:
http://www.eznec.com/Amateur/RMS_Power.pdf
Reid
Last edited:
Thirty-six to be exact. Every Bassman I had on the bench clocked 36 watts, and there were a lot of them.
No. Valve amps smell "warmer" than SS amps. SS amps smell "cleaner"; some would say "sterile". Vintage value amps smell even more than newer valve amps. That's "mojo". (If you've ever wondered what mojo smells like, now you know.)
Class D amps don't smell at all because they lack the "heft" of tube and SS amps. Unless you pee on them, in which case they become "Class P" amps, and reek of mojo.
Not when it has beer spilt into it, the tube amp smells like pancakes cooking.
