-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
NAD 2200 amp Soft Clipping setting
- Thread starter TSHRED56
- Start date
- Thread Starter
- #3
I use a remote controlled SMSL DO100 DAC to control the volume. I set the preamp volume levels med/high but as I think about this the soft clipping feature I realize it could come in handy. Worse case scenario is the DAC messes up and allows full feed suddenly. Might be good to have that protection turned on.
staticV3
Master Contributor
- Joined
- Aug 29, 2019
- Messages
- 8,025
- Likes
- 12,873
Soft clipping does not protect in any way.
It artificially increases harmonic distortion to make the transition to clipping a bit more "graceful".
If you rarely if ever max out the Amp, then I'd leave Soft Clipping turned Off.
It does not protect in case your DAC messes up.
- Thread Starter
- #5
Okay. Thanks. I try to keep the system volume within a decent peak to prevent a spike caused by a DAC failure. I didn't realize Soft Clipping wouldn't help in that scenario.
DVDdoug
Major Contributor
- Joined
- May 27, 2021
- Messages
- 3,038
- Likes
- 4,005
It could potentially make distortion worse if you are close to clipping
It doesn't know how high the wave-peak is going to be so it has to start pushing-down the peaks before it actually clips.... Once the wave is squared-off it's too late. I don't know where it kicks-in but lets say -2dB below clipping... If the normal wave-peak goes to -1dB (over that made-up 2dB threshold) the wave will be compressed-limited and maybe the peak will be "pushed down" to -1.5dB. It still remains "smooth" and "curved"... It's not hard-clipped but it's slightly distorted. If the normal wave-peak is "trying" to go to +2dB it will be limited to 0dB (maximum power) but "pushed down" and smooth so not as-badly distorted as hard-clipping. If you push it too far it will still be squared-off at the top and bottom and only partially "softened".
I don't know if any amplifiers work this way, but with DSP it's possible to attenuate the whole signal (with no waveform distortion) for a short period of time after an excess peak is detected, so the next peak isn't distorted. But without a delay you can't "look ahead" so at least one "wave" will be distorted. (i.e. Compressors have a release time so a short time after the signal returns to below the threshold, the level is no-longer altered.)
It doesn't know how high the wave-peak is going to be so it has to start pushing-down the peaks before it actually clips.... Once the wave is squared-off it's too late. I don't know where it kicks-in but lets say -2dB below clipping... If the normal wave-peak goes to -1dB (over that made-up 2dB threshold) the wave will be compressed-limited and maybe the peak will be "pushed down" to -1.5dB. It still remains "smooth" and "curved"... It's not hard-clipped but it's slightly distorted. If the normal wave-peak is "trying" to go to +2dB it will be limited to 0dB (maximum power) but "pushed down" and smooth so not as-badly distorted as hard-clipping. If you push it too far it will still be squared-off at the top and bottom and only partially "softened".
I don't know if any amplifiers work this way, but with DSP it's possible to attenuate the whole signal (with no waveform distortion) for a short period of time after an excess peak is detected, so the next peak isn't distorted. But without a delay you can't "look ahead" so at least one "wave" will be distorted. (i.e. Compressors have a release time so a short time after the signal returns to below the threshold, the level is no-longer altered.)
I used to have a NAD 3140 back in the 80's and I engaged soft clipping when I was providing music for parties, running the amp at full power (according to the LED meters) for many hours. This was really just to protect the tweeters in the speakers I built for party use. It seemed to work well for this purpose.
I here because I'm planning to modify my NAD 902 Power Amp. It sounds pretty good, but I identified a few areas where improvement may be possible.
The availability of 'soft clipping' necessitate the use of an extra electrolytic capacitor in in the imput line whether it's switched on or off. So effectively there are two 22uF caps in series. That's not good for sound. There is a bridging mode switch which when off is completely disconnected. This is not the case for the 'soft clipping' which connects the input through two reversed diodes to a filter set up regardless of whether the switch is on or off. It's a very strange circuit which involves a zener connected to the positive supply. I'm guessing that when the switch is on the two diodes are taken to the verge of switch on and a large input will take them to a biased on situation. If that's true then NAD may think it's ok in sound terms, but a look at the forward characteristic of a silicone diode shows that a tiny current flows non-linearly before the switch on voltage of around 0.7.
I'm minded to remove the diodes and replace the input cap with a single high quality 10uF one then see if I can detect an improvement in the sound.
The availability of 'soft clipping' necessitate the use of an extra electrolytic capacitor in in the imput line whether it's switched on or off. So effectively there are two 22uF caps in series. That's not good for sound. There is a bridging mode switch which when off is completely disconnected. This is not the case for the 'soft clipping' which connects the input through two reversed diodes to a filter set up regardless of whether the switch is on or off. It's a very strange circuit which involves a zener connected to the positive supply. I'm guessing that when the switch is on the two diodes are taken to the verge of switch on and a large input will take them to a biased on situation. If that's true then NAD may think it's ok in sound terms, but a look at the forward characteristic of a silicone diode shows that a tiny current flows non-linearly before the switch on voltage of around 0.7.
I'm minded to remove the diodes and replace the input cap with a single high quality 10uF one then see if I can detect an improvement in the sound.
