Improving clipping

[Neutrality]

Wanted to make some improvement to clipping performance of my BJT LTP, BJT VAS, BJT Driver, Lateral FET output amplifier.

Simulating with a 2 Ohm load to get a worst case scenario(my speakers are 8 Ohm dipping to 4 Ohm in the bass area).

Clipping, driving an 2 Ohm load occurs at about 3.7V p-p input, which is about 80 V p-p output.

To take it to the extreme I raised the input to 6 V p-p to get some serious clipping.

Attached is result without any measures to improve clipping performance

[Neutrality]

First improvement, Baker clamp from VAS to driver Collector.

Regular 1N4148 diode.

Slight improvement but nothing special

[Neutrality]

Second improvement, changed the Baker clamp to a schottky diode.

Used a 1N5819 Schottky diode.

Better, but still not good enough.

[Neutrality]

Third and last improvement, added a 1N4148 diode in the collector of the driver but before the Baker clamp connection.

Shows a big improvement with no sticking or overhang. Just shows a slight kink in the waveform but nothing to be worried about.

[MiiB]

why worry about clipping with an amp with that voltage-swing....Do you plan to drive it harder than a CD player can..???

[Neutrality]

Originally Posted by MiiB

why worry about clipping with an amp with that voltage-swing....Do you plan to drive it harder than a CD player can..???

Ever heard of designing for worst case scenarios?

I think many preamplifiers can output more than 3.7 p-p....

[Neutrality]

Originally Posted by Neutrality

Ever heard of designing for worst case scenarios?

I think many preamplifiers can output more than 3.7 p-p....

Adding to that, in a real world scenario, when an amp starts to clip continously it will start to sound distorted and unpleasant and people would turn it down.

So 6 V p-p is not realistic at all.

But for short music transients that might make the amp clip on peaks, having some sort of improvement on the clipping performance is a good thing.

"Sticking" on a clipped waveform, even if it just on very short transients peaks, would not sound that good.

The 6 V p-p was just taking it to the extreme.

[jcx]

Originally Posted by Neutrality

Adding to that, in a real world scenario, when an amp starts to clip continously it will start to sound distorted and unpleasant and people would turn it down...

today with Loudness War levels of compression so common many may no longer recognize clipping - its just the way the "music" already sounds

you can find several anecdotes about amateur DJ or others turning up systems with audiophile levels of reserves until speaker meltdown because the sound wasn't "getting fat" from clipping