Unusual Bias Method

lots of software ignores (or mistranslates) the JPEG rotation and mirroring tags.

On a sunny day (Sun, 24 Nov 2013 16:44:32 -0700) it happened Jim Thompson wrote in :

Gov had no work for you lately ;-) ?

I prefer things that work, and keep working when users abuse it. Audiophiles use golden crowbars.

It is a crap amp, increased distortion at low frequencies too due to the bootstrap.

The art is trying to understand perfection, trying to understand crap is like trying to understand dog language. It will never work, will not help you, and waste the dogs time.

On a sunny day (Sun, 24 Nov 2013 19:01:02 -0700) it happened Jim Thompson wrote in :

In that time amplifier technology was 10 years ahead of you, if not more:

Current Dumping Power Amplifiers

QUAD 405 2013 1975 to 1982 64,000 units

(From Wikipedia

You could learn from THOSE designs!

BYE!

On a sunny day (Sun, 24 Nov 2013 22:14:44 -0700) it happened Jim Thompson wrote in :

In physics somebody once wrote, "before coming up with a new theory, or criticize Einstein etc, you d*mn well better understand the current theories in depth".

You 'design' and your fear it seems to _learn_ by studying what others did before you, something you call 'cut and paste', shows you have not mastered the 'audio amp design'.

You brought it up and thought everybody would be amazed, many have already pointed out it is not really any good, or better than what was done 10 years earlier, except for the not needed use of opamps (with its own input dangers).

get a life, if that is still possible with Alzheimer.

One thing I figured out is that that circuit, h-bridged and jammed into a car, is one of the things that destroyed his hearing.

At Tulane, I was a subject in a noise/hearing loss study (mostly interviews and hearing tests) so I got warned about loud noises. I used to wear my motorcycle helmet in steamship engine rooms (the huge aux diesels were the worst) when everybody else was wearing hard hats.

That's what I figured. ...Jim Thompson

Why should I bother.

For the true engineers here... run it on LTspice (RG=100 as requested), I doubt you will see a spike at all... LTspice simulation defaults are too loose. Then run at 100ns max timestep, then 10ns, etc (as I did in PSpice). The spike grows and grows in amplitude. Tells me there's a very narrow but quite high voltage spike. Per Hobbs, the energy IS low, so no immediate blowing the gate oxide... just gradual deterioration... after 30,000 cycles ?>:-} ...Jim Thompson

Yes. It's unusual. How does it work? ...Jim Thompson

On a sunny day (Mon, 25 Nov 2013 07:17:44 -0800) it happened John Larkin wrote in :

Yes, and his circuit could have well oscillated above audible frequencies. Them darlingtons can be nasty, I have worked in a ship engine room once, we had voltage regulators there that needed adjustment, It was way too loud, and it was not even running full power. My hearing is exceptionally good still, but some loss of high frequency, IIRC above 12 kHz last time I tested,

Because you made the claim. You won't bother because you'd look stupid if you tried. You cut your losses by refusing.

I suspect that the bias servo loop oscillates at audible frequencies. The entangled current sinks are probably there to minimize that. In an h-bridge, the two sides oscillate independently! For stereo, make that four.

Nope, that's your talent. My talent is making you fume. And you still haven't been able to understand the amplifier bias scheme. I believe your "can of worms" comment proves you can't cope with anything more than a couple of transistors before you become lost.

And your statement, "The bias setting stuff is component-rich and silly. The usual Vbe multiplier worked fine and took a lot less parts." proves you have no clue.

The Vbe multiplier has dead-band OR thermal run-away... take your choice. (There's an analysis clue in there ;-)

You're a cut-and-paste "engineer", certainly no "designer".

Am I getting your goat yet ?>:-} ...Jim Thompson

You and Jan are clueless. I knew I would fully establish your stupidity level. Didn't realize I'd snag some other bloviators as well ;-) ...Jim Thompson

Even a verbal analysis of HOW? Jan is clueless.

Stone throwers without any hint of analysis. Such "engineering" >:-} ...Jim Thompson

It's right there in plain sight. What's not to understand?

I don't know if/how the bias loop oscillates, and how much of any oscillation gets into the output. Since there'a a comparator in the loop, it must oscillate. And since you went to some trouble to cancel the current sinks, it must.

I also don't know how well it works in real iife. A sustained audio power output will make it think that the bias current is high, so it will ramp down the base-base voltage over time; the duty cycle of seeing the minimum current will be low or possibly zero. That will cause a lot of crossover distortion, for quiet passages, for a while after the big output stops. It would take work to quantify that.

It wouldn't take long at all for you to demonstrate that 80 volt mosfet spike.

Here's a 17 kilowatt MRI power amp that has no crossover distortion or biasing problems, and PPM noise levels:

And a couple of boards with more than a few transistors:

When I did complementary followers, I liked to use relatively large emitter resistors paralleled by power diodes. No possibility of runaway and the forward transfer function was continuous, nice and linear once the feedback loop is closed. Any complementary follower is some sort of nonlinear if both transistors are on at idle.

I prefer using the collectors/drains as the outputs, with closed-loop control of each output device. That can be made to have rigid idle current control and no crossover distortion. None of the transistors turn off... one side swings hard while the other remains at idle current. You get basically rail-to-rail drive, too.

Trying to make friends here?

On a sunny day (Mon, 25 Nov 2013 09:32:36 -0700) it happened Jim Thompson wrote in :

y>>>Yes, and his circuit could have well oscillated above audible frequencies.

mm bloviaters, you are forgiven because of your alzheimer, realy sad, show us ONE working thing you ever made, just ONE.

You live in your world of (non real) simulations, go round insulting great engineers, and have nothing to show for all the claims you made.

Darlingtons in the output of audio power amps have their own problems. And likely somebody make your circuit die with just some weird input,

But again, as you are deaf as a ? you are forgiven as you could not hear the oscillations and distortion. Should have scoped it, do you even have a scope? and a load resistor? And decent speakers?

I mean .....

Designing under influence? :-)

Note that he posted a nearly 40 year old circuit, but will say nothing about it. He probably doesn't remember.

On a sunny day (Mon, 25 Nov 2013 09:33:28 -0800) it happened John Larkin wrote in :

Yes, and now he wants us to help him figure out why it blew up all the time.