My amp design schematic

[fab]

I've always received the same characteristic group for both pnp and npn but if you don't then you would use a trim pot to balance the offset.

I use symetric complementary input stage with NPN and PNP pairs and mosfet as outputs, and match all 4 inputs bjt for hfe to within about 15%, and without adjustment I can get repeatedly (5 same driver boards design) less than 10 mv dc (usually between 1 to 6 mv).

Fab

 

[markp]

I use symetric complementary input stage with NPN and PNP pairs and mosfet as outputs, and match all 4 inputs bjt for hfe to within about 15%, and without adjustment I can get repeatedly (5 same driver boards design) less than 10 mv dc (usually between 1 to 6 mv).

Fab

Sounds like my results too. With a decent match the offset is very low and even just using random pairs I've never seen more than 30mv offset.
How do you like the mosfet outputs? Have you tried a bjt set to compare? I worry about the infamous 'mosfet haze' if there is such a thing anymore.

 

[fab]

...How do you like the mosfet outputs? Have you tried a bjt set to compare? I worry about the infamous 'mosfet haze' if there is such a thing anymore.

To be honest, already at school, I got convinced from the beginning that the V or lateral mosfet would be hard to "beat" when used as output transistors of power amp because of their high current capabilities and low required drive - which also simplifies the output stage of a power amp. Thus I have built only a few protopype amps using BJT as output. After my first built own design using mosfets, I was so amazed by its sound compared to the previous attempts with BJT that I have not really got back to BJT as output devices for power amp since then. My first built "mosfet" design was: "dual series differential" with quasi-complementary mosfet output (N channel). It had a lot of degeneration thus low feedback. Its THD was 0.04% at max power (40 W). I have used it in my home for about 10 years. Today, I use complementary mosfets.
But to get back to your question, I have not objectively compared BJT against mosfet in a similar driver board topology.

If the power mosfet are considered infamous, I would be curious to know why.

Fab

 

[markp]

Mosfets have been said to put a very slight haze on low level detail in the past. Some swear by them others refuse to use them. When I was in college, there were only N channel mosfets available and I'm a big fan of complimentary so I never got into them. Maybe I'll take a look into the new generation to see how they work vs bjt.

 

[AKSA]

Mark,

What sort of VAS stage current and quiescent are you running on this design?

I ask because I'm trying to figure out if the outputs have sufficient beta to drive to clip into an 8R load; with say 10mA VAS current, and a beta of say 100 in the outputs, you could cope with around 500mA peak into the load. This would only be 8Vpp, which is 1W, so either VAS current is high, or output device beta is very high.

Cheers,

Hugh