Output Configuration

[aged one]

Hello all

Question from a newbie......The output of the some amplifiers use complimentary devices (Leach, Digi125,etc) while others use a pair of similar transistors (JLH,Leak,etc).

What are the advantages of each configuration and can they be interchanged?

Thanks

[cunningham]

A quasi-complementary pair consists of two alike devices stacked from emitter to collector of other one. The emitter/collector connection is the output. The AC load line for the input is different for each device, making each side, + & -, having different feedback componants, if global feedback is used. one device is driven like a voltage amp and the other is an emitter follower current amp, with no voltage gain. these circuits are trickier to design.

Complementary pair uses NPN and PNP devices and have a more similar AC load line for input and equal, but opposite feedback componants. Either configuration can be used, but the complimentary pair is more popular because the +part of the driver circuit is same as -part.

[aged one]

Thanks Cunningham

I have a bunch of NTE327 power devices and had hoped that converting from complementary to quasicomplementary was just a matter of some minor circuit rearrangement.

Guess I need to go do some reading

[darkfenriz]

The answer is: they can never be interchanged

The quasi-complementary amp has two similar power transistors and they are opposite driven (the phase is changed by 180 degrees). You can get it with a npn tansistor with the same collector and emmiter resistors to the rails - one transistor is driven directly from the emmiter and the second from the emmiter so that they are opposite.

The complementary pair of transistors are driven with the same signal - the opposition(is it formal english . by the way) occurs becaouse the transistors themselfs are kind of 'opposite'.

differences:
-as the quasi-complementary amps use pair of the same transistors it is going to be less distortive (still under dispute-personal opinion )
- espetially with the MOSFETs: P-channel are worse, slower and much more expensive >> using two N-channel seems tempting
- quasi-complementary are somehow more comlex commonly consider old-time and tube-styled (as we have no comlementary tubes- one using electrons and the second positrons )

cheers

[cunningham]

If you want to make an amp using just one type of device, I would make a class A power amp. Biasing the output with a current source. Capacitor or transformer coupled.

BTW if you have any audio output transformers, you can make a push-pull amp and bias class AB with these devices.

[Upupa Epops]

go man, who in this time can develope quasi amp . It's relict of age, when was on market only NPN ones

[smoking-amp]

Hey! Its not so stone age yet. Look up the "Circlotron" configuration. (Search in Tubes section mainly, but SS too) No output transformer needed, and has the benefit of identical output device characteristics too. Easy to add Hawksford error correction to output also since it only needs one diff. amp in the feedback. (Search "Hawksford" in tube section, then translate back to SS design.) Also requires a phase splitter circuit.

Don

[Upupa Epops]

Connections of type common colector and common emiter have not the same and identical characteristics .

[smoking-amp]

Oh! You are correct, I missed the "quasi" in your earlier post.

I was just pointing out that it is still possible to make a good amplifier design just using NPN outputs by using the Circlotron approach where both output polarities use the same transistor configuration (half CC and half CE).

Don

[sam9]

If you decide to make a qusi-comp output consider using a "Baxandal Diode" parralel to the driver RE resistor on the negative half of the circuit (assuming the output devices are NPN). When I first saw a quasi with this diode my first thought was "I bet a cap across the diode would be better stll". Later I discovered that Linseley-Hood had thought of that decades ago which was kind of reassuring.

In simulation, at least, an inverse quasi works just as well. I.e., PNP output devices at least in the case where you are comparing it to it's complement. The circuit it just laid out other way around. Of course, there is no particularly good reason to do this - but then there is no good reason not to, either.