Class D at low volume levels

[mfong]

Recently hooked up a second Arjen Helder t-2020 amp board to my living room system. It sounds great at usual listening volumes. In fact, I don't hear the power limitations until I get close to too loud. I'm also noticing for the first time that it just doesn't sound good at very low as in barely audible volumes.

I know with traditional class a and a/b, distortion usually goes up as the power increases. I remember seeing a claim that class D goes the other way and there's more distortion at low volume. I've never attempted to measure, but do remember a similar phenomenon with my sonic impact amps.

It could also be that there's some source of noise in my system that isn't masked by lower volume music.

We rarely talk about listening at low volume. Usually, when I evaluate electronics I turn things up. Am I the only one who hears this?

[Workhorse]

Again T-amp .............

Change your class-D amp instead

[mfong]

So,are you saying that this is just a problem with the low power tripath chips as opposed to class D?

[BWRX]

Have a look at the THD+N vs output power plots (usually only given at a specific frequency) for any amp. In general, the distortion gradually drops as output power rises then sharply increases as the amp reaches its maximum output power. That doesn't necessarily mean that most amps will sound bad at low volume levels. Your speakers are just as much a part of the equation.

[Workhorse]

Originally Posted by mfong

So,are you saying that this is just a problem with the low power tripath chips as opposed to class D?

Listen to UCD at low level volume and compare it with any **** chip T-amp

[81bas]

The problem is, that the T-amp uses fully digital processing. And if it does even has 24 bits DSP, then at low level there will be used only 5-6 bits maybe, which is not enough.
Non-digital class-D amps (UcD, for example, as mentioned by Workhorse) are having more deep resolution, and their theoretical lower limit is only the noise level of the devices used...

[Workhorse]

Originally Posted by 81bas

The problem is, that the T-amp uses fully digital processing. And if it does even has 24 bits DSP, then at low level there will be used only 5-6 bits maybe, which is not enough.

ERRRRRRRR........................

From which angle T 2020-amp becomes digital and from where these 24 bits DSP came in jumping for resolution

http://www.kafka.elektroda.eu/pdf/tripath/TA2020.pdf

T-amp uses spread spectrum analog modulation, no DSP with 24 bits

[wwenze]

Doesn't the loudness of the input signal (which is determined by the volume pot and not the amp) affect only the duty cycle in a class-D amp? Where does bits come in?

[Saturnus]

mfong, the problem you experience is common to all low-voltage amplifiers no matter what configuration. Doesn't matter if it's class A, A/B, D, T or what letter is in vogue at the moment.

[sendler]

I have noticed this with my modified Sure 2X100 amps as well. The sound seems to be a little less than satisfying at early morning (other people are asleep) listening levels. The good news is that at normal (loud) listening levels, this is the best sounding amp I have ever heard. Much better than my Ucd180.