I've read where running a true balanced (differential) amplifier as such sounds much better than running it single ended (I'm assuming the same amp has both balanced and single ended inputs here).
Why would that be the case? Is it merely the improved SN ratio from being balanced, or is it something circuit related?
Why would that be the case? Is it merely the improved SN ratio from being balanced, or is it something circuit related?
Balanced signalling is used to reduce noise and hum pickup in the cable run and particularly injected noise due to ground loops between units. Its not to do with the amp as such, although fully differential amp circuits are sometimes used as the same advantages exist (although the problem is much less severe inside a unit)
Balanced amp has two output stages for one output. DC out must be balance for each output for equal clipping. More circuit cost and complexity, and in PCB layout. But as a bridged circuit, it takes power from the rails at full wave instead of alternating half wave intervals for class B type OPS. This is a big advantage for PS regulation. After all, an amplifier is just the modulator of the power supply.😀 Also SR is doubled, but so is the load impedance. 🙄 Bridged mode is popular in class D as the "inverted" digital signal is easily derived, usually internally. As with everything there is always compromise.
Balanced amplifiers with two outputs are very rare, except in specialised telephony applications, and I imagine all those have succumbed to digital. The 'bridged' amplifier configuration has two outputs, but these come from two separate power amplifiers.
Inside an amplifier, a balanced (diff pair) input stage is almost always used because it has good DC accuracy, higher transconductance, and if properly set up completely cancels second-harmonic distortion. That is often claimed for circuits but this is perhaps the one place where it really works reliably.
Inside an amplifier, a balanced (diff pair) input stage is almost always used because it has good DC accuracy, higher transconductance, and if properly set up completely cancels second-harmonic distortion. That is often claimed for circuits but this is perhaps the one place where it really works reliably.
Very well said!
Also #4, but unquotable yet...
Actually I consider it rather strange not to drive the loudspeaker from a bridge. As if the 'other side' roots in a everything swallowing swamp (the 'ground') never causing any trouble. Really?
After my post #4 is quotable, but well...
Last edited:
A balanced input forces good grounding because the signal return is not ground. Obviously this cancels noise but it also avoids distortion from ground currents. Grounding problems are very common.
Canceling even harmonics is a mixed blessing because some times even harmonic distortion is deliberately added to "enhance" music, but of course there is more to it than any random even harmonics.
Canceling even harmonics is a mixed blessing because some times even harmonic distortion is deliberately added to "enhance" music, but of course there is more to it than any random even harmonics.
Most of the issues are well described in the attached article from Bruno Putzeys.View attachment WP_The_G_word.pdf
From my limited experience balanced and differential are two separate things:
* balanced connections - ie cable runs, using differential signalling
* differential amps - amplifying both differentials, the resulting difference is then amplified and that isn't cancelled out and becomes sound.
The most common use for differential amps (other than cable runs) seems to be before a push pull.
Common noise on both signal paths that are equal magnitude will cancel, the residual magnitude difference and non-common noise plus the signal then amplified.
Also it seems prudent to treat power supplies as another signal source, and design for common noise from the supplies to cancel out.
I never thought of it from the grounding perspective - you learn something every day 🙂
* balanced connections - ie cable runs, using differential signalling
* differential amps - amplifying both differentials, the resulting difference is then amplified and that isn't cancelled out and becomes sound.
The most common use for differential amps (other than cable runs) seems to be before a push pull.
Common noise on both signal paths that are equal magnitude will cancel, the residual magnitude difference and non-common noise plus the signal then amplified.
Also it seems prudent to treat power supplies as another signal source, and design for common noise from the supplies to cancel out.
I never thought of it from the grounding perspective - you learn something every day 🙂