Agree about Spice, LT or other, though there seem to be people over at diyAudio that are pretty good at it. Truthfully, I'm such a dinosaur I never even considered it when I made the Bode suggestion. I'm Mister Physical Measurement, sworn enemy of Mr. Fancypants Math much of the time. Though I've never done it, I know there are techniques used for opamps to measure it open loop, or somehow derive it. Bob Pease probably had something to say about it, and I'll check my notes here.
On the speaker distortion issue, and this is purely my own unsubstantiated belief, speakers are a physical spring and mass system of fairly low frequencies, albeit somewhat complicated. They have a hard time making "unnatural" and unpleasant sounds. Amplifiers, OTOH, have no trouble at all generating harmonics you wouldn't find in a speaker. Amplifiers I've heard that keep THD below 0.01%, and don't have other flaws, sound about the same to me. Amplifiers above that start to please me less and less. I've heard the "you can't hear it below a few tenths of a percent at best" argument, but I remain a skeptic. The best speaker systems I've had were tri-amplified with active crossovers and were very low distortion.
I was banned by Diyaudio, I was too cocky before, I since learn my lesson.
Sure people are more technical oriented there, but I found a lot of them are quite close minded. I started a lot of discussion, I don't think I got fruitful result as this thread. I like to throw out ideas and think and talk, but over there, people seems to be fast in throwing out other people's quote and "trusted" circuit.
I actually spent a lot of time asking help in the LTSpice yahoo group. Those are the true serious LTSpice people. I did talked in Diyaudio about LTSpice before and found they are not as knowledgeable as I expected. I stop doing a lot of simulation lately because the result is totally inconsistent from the real circuit. One thing is I feel very confident on my layout because I layout all my boards in my whole career, I did circuit much much more critical than this, from multi GHz to nV, pA low level signals. I think I did everything I can on my layout with point to point 0.1uF and 10uF bypass cap for every single transistor, power and ground planes and very short connections on critical paths. I don't think my layout cause any problem, at least I trust my layout much more than the LTSpice.
I studied the Spice models, their models are way too simple, there is no lead inductance and parasitic in the model. In the output stage, you are dealing with a few ohms impedance, every little lead inductance can become important in a few MHz. All the fixes that are so effective in the real circuit do not do anything in the simulation. In fact, when I put 3300pF in my amp simulation, it didn't do squat!!! But it's the major killed in real circuit.
I think LTSpice just treat the output impedance of the output stage using r'e= 1/gm and just parallel all the stages together to give almost 0ohm output impedance, nothing you can put on the output can affect the stability. BUT in real life, lead inductance comes into play, output impedance gets funky when you are in a few MHz arena with transistors that are 30MHz or so.
I just change the Zobal from 50ohm to 22ohm and 0.022uF at the output terminal of the amp, it really improve the stability, I can relax the compensation and it's stable with 0.022uF right on the amp's output connector with NO 8ohm load. This is really good.
I even removed the 10ohm 0.1uF Zobal on the OPS board. I am sure if I put the 10ohm and 0.1uF Zobal on the amp's output connector, that should do it. And I am going to reverse the other Zobal and stuffs to see how far I can reverse back to the original circuit. My point is to find the sore spot of the circuit and I think I am getting close.
I like to really get into this experiment and design as this is my hobby. I found when I raise a question in DIYaudio, people are too quick to give an establish solution and say just use it. I don't find this work out all the time. This is the exact example where Zobal on the board didn't help at all, but a lighter Zobal on the output connector do magic. this is through talking and thinking which I don't get from DIYaudio before, people just offer what they consider an acceptable solution and tell you to follow it.
Another thing about compensation, all the schematic of people designed their amp in Diyaudio mainly stress on the miller cap of 33pF and some variation. I found it is critical to have a RC lead-lag network across the collector of the LTP to set the low limit of the capacitance at the LTP. Say if you use 33pF miller feedback capacitor, you get 33000pF when the open loop gain is 1000, BUT you only have about 60pF loading the LTP when open loop gain is down to unity as miller effect is gone. You need a lower limit of about 220pF to keep the stability.
This is the same for the VAS where the second part of the miller capacitance, it's only equal to 33pF, no more. You need a little more on that. You cannot increase the miller capacitor as it will increase too much on the LTP. So you need a lower limit on the VAS. I put a 330pF on it. These two really help. But I can see if I talk like this in DIYaudio, I'll get laugh at!!!! I am producing result on the bench, yes, I might have to re-invent the wheel, wasting a lot more time, but I think I learn. The last one week of playing with the stability taught me a lot.
I think you get very deep into the this too, you like to stop and think and research yourself to find the answer. That's why I like to talk to you as both of us tend to be "trust but verify" type!!! Believe in "seeing is believing" and spend the time to look at the theory behind it. That's what make it interesting and to learn.
EDIT:
Last week when I was struggling with the stability, it felt so hopeless and running out of ideas. I was tempted to use another name to join DIYaudio again, but I just kept telling myself to be patient, think it through, experiment with it and try to think out of the box. That's when I came up with the idea of loading the EF driver and pre-driver with Zobal. I am glad I stayed the course and find my own way. I know this is nothing new, but I like to discover this on my own.
Did I tell you that even though a have been an EE for 30 years, my degree is Chemistry, I almost never have formal education in EE. I studied a lot, hours a day for years on my own. I always find my own path, I might waste more time, but I think I can do it with power amp. This is why electronics is my single passion of life and learn and discovering is part of what make it interesting.
