What causes oscillation at clipping?

I often stick stuff together at random just to see what it will do. That's how I've come up with all those wild (but working) current mirrors I've devised over the years. ...Jim Thompson

I kind of had my heart set on getting that weird output stage to work properly, and the other circuitry was shoehorned around it which lead to a mess. The boostrapped cascode seems to work well enough, maybe I'll keep that part and try to go with something more "conventional." I just want to avoid the standard differential pair, voltage stage, push pull output buffer with biasing diodes setup that's so common. Though perhaps it's so common for a reason.

Because you are creating a phase shift with the caps in the loop.. You start off with (-) feed back but end up generating (+) feed back and only needs some trigger to start it. Like clipped signal for example..

Jamie

Actually, I've found that if I connect up the boostrapped cascode directly to the "weird" output buffer via an emitter follower I can get everything to work pretty nicely, so long as I don't need more than a few volts output. :/

Work in smaller pieces. When you tested the output stage alone, you realized it was a dud. ...Jim Thompson

What changed when you did that? Hint: Current drive versus voltage drive ?:-)

You need to learn and use node and loop analysis, until you can "see" problems just by observation.

Also assume, and analyze, states... driving up, driving down... saturated?, etc.

I've been building circuits dating back to when my only skill was the ability to read a schematic and solder (I grew up in a radio and TV repair shop).

But I was already starting to be able to analyze when I went off to MIT. MIT honed my skills. I also worked as a technician while a student at MIT. ...Jim Thompson

Stick with a design that uses all DC coupling through out..

Seeing this is a Head phone amp, just a bridged output. The fewer the caps you pass your signal through, the better off you are.

Jamie

Thanks again for your advice and assistance. I'm going to give it a rest for now, and give it another try when I'm fresh.

Darlington...ugh.

Sometimes a resistor in series with the base of the output device gets rid of oscillations. However I'd suggest going mos on the output stage.

=A0 =A0Q9:b

=A0 =A0 =A0 =A0 =A0 =A0...Jim Thompson

ulation

al

=A0 =A0 =A0 =A0 ...Jim Thompson

to

=A0 =A0 =A0 =A0...Jim Thompson

=A0 =A0 ...Jim Thompson

=A0 =A0| =A0 =A0mens =A0 =A0 |

=A0 | =A0 =A0 et =A0 =A0 =A0|

=A0|

=A0 =A0 =A0 |

Spice is a verification tool, not a design tool.

off

igh

the

for

Most headphones have a common ground, which makes bridged designs impossible.

Part of the reason I use it is because of my total inability to draw a schematic consisting of more than a few parts and have it come out looking decent. I'd like to have some software that would let me come up with a neat printout of the basic topology and then work out the rest with pencil and paper.

I could just use a printout from LTSpice I suppose, but there's always the temptation to start simulating. Better some other program that doesn't have that capability.

^^^^^^^^^^^^^^^^^^^^^^

Isn't that what I said? (Underlined above.) ...Jim Thompson

Until the late '80's, though I had a simulator (2G6) on a VAX, I didn't have a schematic-entry tool.

So I hand drew schematics, numbered the nodes, edited a netlist, then ran it on the VAX.

Later under DOS.

I still sometimes draw partial schematics on paper, just to doodle thru the math as I design... as a sanity check. ...Jim Thompson

Echo...echo...echo...

Well, I have a better idea for a design in mind, and this time I intend to just rough out the basic topology in LTSpice (using basically as an illustrator) and fill in the rest with pencil and paper. I've found a design for an output stage that I like and that I think will work well for a headphone amp.

Doing a design like this from scratch is certainly a humbling process. Being innovative is _hard_!

Naaaah! That's the fun part ;-) ...Jim Thompson

I've been fortunate/cursed to have grown up in a time where some form of computing was available to the masses; even as far back as second grade I can remember the classrooms having a Colecovison or an Apple II.

By the time I went to college circa 1997 in my rounds as a network technician I still encountered _some_ teenagers/20 somethings that were complete virgins to computers - they had never had one in the home and the secondhand Mac Performa that they arrived with was their first machine. Desktops were the rule, I remember dealing with only perhaps a half dozen laptops in my entire time working in that capacity. We were, around that time, probably the last generation of students who generally didn't have a laptop or cellphone or PDA in the classroom and mainly did all our note-taking and assignments with paper and pencil.

Even so, there's still the temptation for me having grown up on that cusp to want to do everything "inside the box" and not bother sitting down to do serious calculation.

It's hard to garner new skills once you leave school. Though my mind is quite good at thinking up nice math algorithms to solve engineering issues I have to resort to asking my oldest son to write the software program. Wish I could do it myself. ...Jim Thompson