"But in an emitter follower there is no time delay"
It is to laugh.
I've built and personally measured subnanosecond ECL circuits, which use emitter followers extensively, and I can assure you: the emitter followers delay the signal.
Or if you don't care about IC circuitry: Audio power amplifier people talk about "excess phase" of emitter follower output stages (see Cordell p.84). This is time delay. Even a Charles Hansen designed power amp has time delay in its emitter follower output stage.
It is to laugh.
I've built and personally measured subnanosecond ECL circuits, which use emitter followers extensively, and I can assure you: the emitter followers delay the signal.
Or if you don't care about IC circuitry: Audio power amplifier people talk about "excess phase" of emitter follower output stages (see Cordell p.84). This is time delay. Even a Charles Hansen designed power amp has time delay in its emitter follower output stage.
No BTW it isn't, how you can pack so many misconceptions in 13 words is amazing.
I've been waiting for the feedback argument to start! 😀
We might need to start a separate thread called "What's wrong with Op Amp Feedback?" because in my experience the audio world often overlooks the feedback loop inside of delta-sigma ADCs and DACs.
Now I'll go back to enjoying my properly functioning endocrine system, which relies on multiple nested feedback loops.
We might need to start a separate thread called "What's wrong with Op Amp Feedback?" because in my experience the audio world often overlooks the feedback loop inside of delta-sigma ADCs and DACs.
Now I'll go back to enjoying my properly functioning endocrine system, which relies on multiple nested feedback loops.
I admire the speed at which you read the whole thread and decided which model was referenced. 🙄
No, but you can hit the center of the oak tree with greater accuracy, albeit perhaps correcting your path some nano or pico seconds too late?
Isn't that the idea...
of course given that "chips be cheap" why no feedforward correction?
The straw man model. No one I know starts out with a linear, infinite BW assumption to do anything useful (i.e. linear = no distortion in the first place) unless they are setting up a straw man to "prove" feedback can't work. This is the same technique used to prove Fourier analysis is "flawed".
BTW equating phase and delay leads down a dangerous path. A true delay is non-minimum phase, phase alone is part of a steady state solution which any normal amplifier at audio BW reaches.
I happened to find one of those Cambridge Audio phono pre-amps in my stuff and took off the cover. The PS is simply a 12V wall wart with a double Cockroft-Walton circuit and 3pin regulators to get +-18 volts explaining some of the noise on the output. So I suppose this doesn't bother the op-amp rollers?
Last edited:
I fear that you're the one doing the strawman, more precisely in combination with missing the point. The thread never got past the models where beta and A are constants, or, at best, a function of frequency. It was quickly interrupted by the same allegations of "Fourier denial", glossing over the fact that the discussion was getting into non-linear systems and/or with memory. Plus "the hive" postulating that amplifiers can't have memory / time delays, and it's all about phase. Pretty much like the same direction you're trying here.
As someone previously said in this thread, when the hive is buzzing it means there must be honey around.
As someone previously said in this thread, when the hive is buzzing it means there must be honey around.
Hives may also buzz when they smell shite.
No allegations made, guilty? Simple analogy to, "the Fourier integral is from plus to minus infinity so it can't apply to real audio signals". Do I really need to search out how often all these tired arguments are made.
Scott, why don't you go and do something useful like design the next super-duper opamp? 😀
Jan
Jan
Are you sure you can quote me saying that?
If not then what's your point? Discussing third parties and inferring my guilt by association?
I suppose its all sort of a moot point for me, I'm using a Sziklai Hybrid now for pretty much every stage. It cleans up the codec/op-amp issues and is super simple. You could say I've found the perfect "op-amp". Its my "happy place" at least for diy audio. The TPA311x makes it all Nirvana.
So goodbye data books , app notes, Google Advanced Search, "last buy" notices and those nasty lead times.
-Bruce
So goodbye data books , app notes, Google Advanced Search, "last buy" notices and those nasty lead times.
-Bruce
No I never said that you made this statement, so let's start over, posit an argument that feedback does not work and I will respond.
The question is, is there such a thing? THAT contacted me a couple of years ago but I take conflict of interest seriously maybe now an idea to toy with.
You're missing the point again. The allegation wasn't that "feedback does not work".
Let me state the question in a different way:
- do you believe that, for the purpose of designing/analyzing/characterizing audio amplifiers, it is sufficient to use just the small signal model of the components?
I modded one of those once. The Cockroft-Walton is not so much of a bother to me as is the 3 terminal regs. it wasn't such a bad unit after the mods. I put my goto 600v/us LT1358 in the front end. I left it socketed so the owner could take it out and use the AD823 prescribed by modder Hypnotoad. The owner left the LT1358 in.
- Status
- Not open for further replies.