Cascode Output Stages

[AMV8]

Pietjers

You ask about sonic benefits. My experience is that the benefits of cascode topology are in the front and mid stages of a power amp. I believe cascode topology can provide sonic benefits in all stages of a pre amp. Of course this all depends on the actual topology and devices used. This is based solely upon the amplifiers that I have designed and built for myself over the years. I have found that cascode designs are easy to listen to.

With the advent of power mosfets, I find myself agreeing with Ha, that the benefits of cascoding in the output stage would not apply any more.

Don

[Lumba Ogir]

Hi Hannes,
some of Self`s statements regarding distortion are confusing to me, anyway, his lack of genuine sense for audio is obvious (in terms of designing good sounding amplifiers).
The distortion in output stages strongly depend on the output power and load characteristics, which are typically nasty. You know what I hold about static measurements.
Also, I am proposing fully balanced topology decreasing the demand for rail voltage levels.

Next thing is that especially mosfets like high Drain-Source voltages for linearity

No. Please read about the decisive parameters for linearity. These are: L, VGSth, Gm and Crss.
Ode to Lateral MOSFETs

[Bob Cordell]

Originally posted by h_a
Cascoding for the outputstage is not very practical in real life.

If you want best linearity, you use the transistor in a very small voltage range, thus small voltage swing. Consequently you have greatly reduced output power.

It's the same for the cascoded power jfet; how much power can you get from power jfets, cascoded to about 5 volts to keep dissipation down? C'mon!

If you enlarge the outputswing, you get reduced linearity. Next thing is that especially mosfets like high Drain-Source voltages for linearity, so usually it's better to drop the cascode and get the larger voltage swing.

But of course feel free to build and see for yourself!

By the way, read D. Selfs book, the outputstage does not produce so much distortion compared with the VAS or the input. While most here will not agree with his high-feedback attitude, he shows undebatable measurements.

Have fun, Hannes

Properly cascoding the output stage can provide some nice benefits, as Nelson demonstrated. I do believe that the cascode is best driven from the input signal, rather than fed back from the output, and this choice can affect stability.

I think the key point about whether or not to use a cascode output stage is the price to be paid in power supply voltage headroom and in increased power dissipation, both at idle and under signal conditions.

One possibly fair way to ask the question is as follows: let's design the amplifier with the cascoded output stage to produce a given output power, say 100 W at 8 ohms, and set the rail voltage accordingly.

Now let's build that same amplifier with the same rail voltages, but without the cascode. Let's assume we have enough SOA in both cases.

Which one will sound better?

Not sure, because the output stage in the second design has extra headroom at the disposal of the output transistor doing the work. Up to the 100W power level, that transistor may not deteriorate in performance as much as the one with the cascode above it. I'm not saying I know the answer, I'm just saying that it may not be as obvious when looked at this way.

Let's now put in a clean Baker clamping circuit to clamp the second amplifier at the same 100 w maximum power level as the amp with the cascoded output stage. The output stage proper will never clip or saturate.

Which amplifier will sound and perform better?

Not sure.

The point is, if we are going to take a rail voltage and power dissipation penalty, there may be more ways than cascoding to spend that headroom in making the amplifier perform better.

Good food for thought.

Cheers,
Bob

[pietjers]

Thanks again for the interesting replies.

Bob,
Because the cascode has very different electrical properties it will have a different sonic character, regardless headroom, maximum output power or whatever. A different circuit topology will have a different character, you will hear that right away.

What I am actually looking for is a unity gain cascode output stage to use in a hybrid power amplifier with a tube front-end. I am currently building/designing an amplifier with a triple darlington emitter follower output stage. The prototype is really wonderfully sounding with its tube front-end!! Without any doubt a lot better than the normal darlington emitter follower I used before. One of the benefits of the triple is the enormous amount of current gain. My measurement s show a current gain of 6.4M. I never tried the CFP because the emitter follower does its job so well. What I hear is the magic that tube amplifiers are so famous for.

But the fact is that I am fascinated by the unique properties of a cascode output stage.
Does anyone know an example of what I am looking for (unity gain)?

Peter

[ppa]

Hi all,

there is one error in the schematic; the led diode must be inverted.

[h_a]

Hi Lumba,

given your expertise in modeling theory, why don't you develop an amp according to your principes? I guess quite some people here would be interested.

But please don't call sine waves static

All the best, Hannes

[Sigurd Ruschkow]

Here's a design that uses a cascoded output stage. It is a 120W/)ohm power amp that was designed by the guys at the Danish hifi magazine "high fidelity" in 1984.



Sigurd

[Sigurd Ruschkow]

It is a good amp that can sound very good when hooked up with a good PSU and good components.

Photo shows non modified version.



Sigurd

[Lumba Ogir]

Hi Hannes,
it is exactly what I want to do, but I would need some help, maybe you could provide that...
I do not believe that anyone would be interested.

[pietjers]

Here is the circuit that is in my head for a while. Perhaps it is more a “totem pole” than a cascode as JCX stated in one of the previous posts.

The Vbe multiplier and variable CCS are only drawn as a function, but that won’t be the toughest part.

Would this circuit work well? What do you think?

Peter