New class A biasing (with non-switching class AB overflow)

[powerbecker]

Originally posted by anli


Here it is. Also modified LTspice's lib/cmp files are included.

Thank you very much, I´ll try them later!

[powerbecker]

Anli,
please can you explain how to get the nice pictures from Ltspice
you post here?

All your circuits are transconductance amps with FB so I take the best (the simplest, with THD20=0.23% ) for some improvements.

I add 2 drivers ,now all have a power bandwidth of 1MHz,
the simplest one has a THD20=0.01% which is fairly low for such a simple schematic !

Also I try to add a AD8099 for further FB, and at least the inverting example goes down to 0.000 12% without loosing the bandwidth.

Of course that´s only simulations but I am fairly confident that the behaviour in reality will be similar.

Thanks for your nice idee!

Regards
Heinz!

[anli]

Originally posted by powerbecker
Anli,
please can you explain how to get the nice pictures from Ltspice
you post here?

All your circuits are transconductance amps with FB so I take the best (the simplest, with THD20=0.23% ) for some improvements.

I add 2 drivers ,now all have a power bandwidth of 1MHz,
the simplest one has a THD20=0.01% which is fairly low for such a simple schematic !

Also I try to add a AD8099 for further FB, and at least the inverting example goes down to 0.000 12% without loosing the bandwidth.

Of course that´s only simulations but I am fairly confident that the behaviour in reality will be similar.

Thanks for your nice idee!

Regards
Heinz!

Heinz,

1. You have not set mosfet current to something about 150-200mA (with simply added follower you have ~2.7A).

2. I'm not sure THD itself does mean too much.

[powerbecker]

Originally posted by anli

Heinz,

1. You have not set mosfet current to something about 150-200mA (with simply added follower you have ~2.7A).

2. I'm not sure THD itself does mean too much.

OHHH, sorry, you are right how awkward!!
Please don´t look at my simulations!!

[powerbecker]

Now I changed 4 resistors and test the circuits again with 200mA quiescence current.

And of course bandwidth and distortion gets a bit worse.

When I added the driver I don´t realize, that a adjustment for different Gate-Source voltages would be necessary.

I hope I´ve done no further mistakes.

Heinz!