Krill - The Next Generation

[Steve Dunlap]

Originally Posted by AKSA

Interesting, Steve, something of an epiphany.

Your contention all along was correct, then? The outputs never turn off?

Could it be that the simulator has problems with driving the collector of a transistor?

Hugh

Hi Hugh,

I'm not sure where the differences lie, other than in the models. LTspice gives essentially the same results as my copy of Intusoft. That, plus what I measured back when I could, does lead me to believe I was correct all along.

I'm sure there will be more debate on this. I am not willing to go into the evolution of my design until this is mostly settled. I don't expect what I presented here today to change all minds, but hopefully it will sway the ones on the fence.

[jkeny]

Very well done, Steve, you know what they say, if you want the job done right ...........

[tinitus]

Excuse me, I have no say in this
But I really dont understand this debate
I understand this amp has been built and sounds fine
How can you debate if it works or not
Either it works or it doesnt
No point in explaining it to me, because I wouldnt understand
Or have you just moved to questioning why the sims and theory doesnt work
Its a bit confusing
But a pitty if this is a fine amp that was lost in a stupid game
But Im afraid its not the first time
Do I need to say who loose the most

[AKSA]

Steve,

Based on my listening tests with very similar topology amps in Class AB and Class A, it may be that the switching or otherwise has little impact on the sound quality anyway. I actually find the Class A amps to sound slower, less dynamic. Most of the differences can be sheeted home to gm doubling at the crossover rather than the switching per se. I'm not too sure how you eliminate gm doubling without lapsing into Class A.

Sliding bias schemas seem to have their problems, too.

Tinitus,

The amp works fine. That's not the issue. It's whether the outputs turn off at all which is under discussion.

Hugh

[stinius]

Originally Posted by Steve Dunlap

but seem to be in my favor.

Interesting, but if it seems to be good to be true, maybe it's something else?

Not meant to criticize, but could you post the biasing volts and currents.
Is it biased in class A/B or maybe A?

stinius

[Steve Dunlap]

tinitus;

Of course the amp works. I think the debate has been over whether the output transistors turn off or not. Many people have had trouble running a simulation on this circuit. I had problems with that myself at first. Some sim programs don't seem to like this circuit and don't give results.

[Steve Dunlap]

Originally Posted by stinius

Interesting, but if it seems to be good to be true, maybe it's something else?

Not meant to criticize, but could you post the biasing volts and currents.
Is it biased in class A/B or maybe A?

stinius

In my sim program the outputs are biased at 202.77ma for a total output stage of 55W with 4 pair of outputs.

I am off fro the night now, not avoiding answering questions.

[Allen Wright]

Steve,
Interesting concept.

We played with an amp 30 years ago that also didn't turn off - it was an adaption of the "Blomley" circuit pubished in a Wireless World magazine circa early 70's, and classes as a low standing current pure Class A.

I don't have the article any more, but still remember the concept. It used a fairly conventional front end that drove a pair of complimentary commonbase trasistors that acted as current directional switches - the output triples were biased on in very low AB1 (circa 10mA) and when a given common base tr was conducting, that output triple was powering the load, on the other half cycle, the reverse happened. But no visable crossover distortion, even in the residue of a distortion analyser.

Was the best sounding amplifier we ever had heard in the 70's, and still sounds great today.

Regards, Allen

[tinitus]

Originally Posted by AKSA

Tinitus,

The amp works fine. That's not the issue.

It's whether the outputs turn off at all which is under discussion.

Hugh

Ah, sorry, stupid me, I understand
Its about output "biasing", whether theres switching or not

John Broskie is writing a lot about class-AC right now
Might that be related ?

Hugh, you mentioned your concerns with classA
Funny, the few supposedly classA amps I have built have always sounded better running with lower bias
But I always thought it was more related to "poor" design, or more correct a design not fit fore classA, rather than it was related to classA in general
That said, I always let my ears decide the best bias point
But seems to allign well with stable DC as well
Not sure how thats connected

[syn08]

Originally Posted by Steve Dunlap

I think the debate has been over whether the output transistors turn off or not.

Not as far as I recall.

First, it was challanged that the outputs not switching off has any relevance for the crossover distortions. It does not.

Secondly, it was challanged that the Krill output stage will do anywhere significantly better (that is, to justify the extra complexity) compared to a EF triple (like the Locanthi), biased near the "Bernie Oliver" point.

Third, the posted performance (of 50ppm @ Ic=25mA, if memory serves) was challenged.