2. amps-preamps
  3. 1409251739

Class A vs. AB vs. D... can you hear a difference?

All things remaining constant is there an audible difference?

I do not mean tube vs. Solid state.

All solid state.

Some Class A amps go to class AB after so many watts... is there an audible change?

I ask because I have a class AB amp and was thinking about going to a class D set up front for home theater

Thanks

Bill
baranowski

Showing 3 responses by kijanki

kijanki's avatar
kijanki
4,446 posts
 
It depends greatly on amplifiers used. To me class D
sounded smoother and less splashy (darker) than class AB.
Midrange was better but lower midrange was leaner (had to
change speaker cables). Lows were tighter while highs
initially sounded less extended, but they just sound
different - cymbals are less splashy. Amplifier (Rowland
102) needed few hundred hours of break-in to sound smooth.
Initially it was a little harsh/forward. This amp also
sounds much better at lower volume (resolution, extension).
At high volume peaks it keeps composure better but it might
be due to regulated (SMPS) power supply and not the class of
amplifier.
kijanki's avatar
kijanki
4,446 posts
 
Ngjockey, you have to realize that linear power supply is in reality a primitive switcher that pollutes AC even more. Jeff Rowland designed his latest class AB power amps with SMPS to reduce noise. For the same reason he designed his Capri preamp with SMPS (where efficiency is not important). Benchmark was able to improve signal to noise ratio from 116 to 126dB by changing power supply from linear in DAC1 to SMPS in DAC2 . Recent power amp from Benchmark is also powered by SMPS. As for the highs in class D, I agree - they sound different, more natural. Brass sounds less splashy and more like brass.
kijanki's avatar
kijanki
4,446 posts
 
Linear power supply draws current from the mains in narrow spikes of high amplitude. These spikes contain a lot of harmonics that might induce noise in any LC circuit inside as well as coupling to other cables. I called it a primitive switcher because diodes switch on/off when voltage is the highest. It also produces diode switching noise when diode is suddenly reverse polarized and conducts for a moment in opposite direction to finally snap back. In addition to high frequency noise it also produces 120Hz noise that is very hard to filter out. It requires large transformer that produces mechanical noise especially in presence of any DC. It requires a lot of capacitors to clean 120Hz and to keep voltage steady since it is unregulated. These large capacitors contain inductance compromising amplifier response at higher frequencies. Adding non-inductive capacitor in parallel helps but creates parallel resonant circuit (with inductance of main capacitors) that rings.

Modern zero-voltage/zero-current switching SMPS operate at high frequencies (Rowland latest SMPS operates at 1MHz) that are easy to filter out. In addition they often contain Power Factor Correction presenting resistive load. In addition they are line and load regulated with instant response and are not sensitive to presence of DC (can even operate from DC). Jeff Rowland uses SMPS in preamps where efficiency is not important. Benchmark's new power amp contains ultra quiet SMPS resulting in overall S/N=132dB.

http://benchmarkmedia.com/products/benchmark-ahb2-power-amplifier

Bel Canto wrote white paper highlighting advantage of SMPS

http://www.belcantodesign.com/pdfs/EfficiencyandPerformance.pdf

Jeff Rowland also wrote many papers explaining use of SMPS:

http://jeffrowlandgroup.com/kb/questions.php?questionid=145
More to discover