New type Zobel = <THD

[Zero D]

Well new to me anyway 😀

Been tinkering again 😉

I've only shown the relevant parts, but the full simmed Amp shows lower THD with this configuration. Not much lower i admit, but lower = Lower 🙂

@ any level below clip it went from 0.004% - 0.003% @ 1kHz. @ 10kHz it went from 0.034% - 0.020%

I havn't extensively tested this, or changed the caps/res much, as i've just done it. The values of the Zobel components as shown are just examples of what was achieved so far. Other values/components may be substituted.

Further options "could" be as in screenie 2 & 3 for eg, and/or variations on the same theme.

If it's not new, why don't vendors etc do it ?

 

[djk]

Mark Levinson ML3.

 

[AndrewT]

Why did you choose emitter resistor at 4r7 for a power amp?

 

[davidsrsb]

Real loudspeakers are not 8R2 resistors. Up to a few nF of cable capacitance is the problem

 

[Zero D]

@ djk

Thanks, i'll try & find the circuit diagram ! Interesting that it does actually work 🙂 Even if i wasn't the firstt o discover it.

@ AndrewT

It was a headphone amp sim i had on hand.

@ davidsrsb

Indeed 😉

 

[JMFahey]

It was a headphone amp sim i had on hand.

Fine, but then it might be useful to re-simulate with real Power Amp values.
Using, say, 0.47 to 0.1 ohms (which would be the actual value range) makes me think that the already minuscule distortion reduction will be even smaller.

 

[Zero D]

@ JMFahey

I tried it with 0.25 Ohms & you're right it @ 1kHz anyway it was still 0.003% !

My motto is, if you don't experiment, you might never know what's Actually possible. Sometimes good things have been discovered just by tinkering 😉

 

[JMFahey]

Fully agree 🙂

But I had a gut feling that a smaller ballast resistor would put both feedback points closer to each other.
Thanks for checking.