Why do specs matter if it sounds good? I don't want to perpetuate the false marketeering of THD. Even the existing audio companies that openly state how BS THD is still post THD measurements only because their customers demand to know, further perpetuating the cycle.
I've occasionally had this daydream in my head. To give you a perspective on my thought process, the way it usually plays out is that I'm just like "here, listen". Initial presentation over. Why waste time with words and specs?

My 2c.

Even somewhat horrible electronic specs sound good in a room with great acoustics and good speakers. HI-FI audio starts with the rooms physical construction, acoustic treatments, traps, speakers (the actual transducer where most distortion happens) and speaker placement, not electronics in a critical listening environment. Room correction software with the best electronics can only polish a turd of a room. I see people spend thousands on equipment with nothing on treatments that are surprised it sounds like crap in the living room.

http://arqen.com/acoustics-101/room-setup-acoustic-treatment/

 

Even somewhat horrible electronic specs sound good in a room with great acoustics and good speakers. HI-FI audio starts with the rooms physical construction, acoustic treatments, traps, speakers (the actual transducer where most distortion happens) and speaker placement, not electronics in a critical listening environment. Room correction software with the best electronics can only polish a turd of a room. I see people spend thousands on equipment with nothing on treatments that are surprised it sounds like crap in the living room.

Unfortunately I have 0 experience with speakers. I use HD800s as a reference. I have 2 unique driver technologies that I intend to experiment on and eventually build a speaker out of one of them, but that is expensive and far away. They both require special amplifiers and the price of entry into the R&D of these technologies is steep. I've only done short proof of concept tests so far.
My first move is to design headphones, I can't see any reason based on the physics that I won't be replacing my HD800s immediately after building one of them. Fingers crossed.

 

Then post your bode plot

Here's the simulated bode plot for the solid state amp 5v p-p into 300 ohms

The tube amp gain of 8

Below is a simulated THD plot of my best tube design, although keeping it stable in real life has proved...challenging. I'm currently using a lesser performing variation.

Dac stage

Looks like there is some low level distortion in the dac stage.

Power supply @200v output loaded with 300ma.
Output impedance starts at 500 picoohms at 1hz and ends at 40 microohms at 20khz.

I'm a noob at reading instability plots. Let me know if anything sticks out at you in any of these.
I have a bunch more designs but they are still in beta.

 

Your first plot is a 10 milliwatt output into 300 ohms?

Why are you not using speaker models or a resistor close to a speakers impedance?

Elliott Sound Project website may be useful to you.

 

our first plot is a 10 milliwatt output into 300 ohms?

21 miliwatts, which is probably enough to destroy the driver. Certainly enough to make you deaf.

Why are you not using speaker models or a resistor close to a speakers impedance?

I am, I specialize in hi-end headphones, not speakers. Speakers are too expensive for me. Sorry I should have mentioned that. I eventually plan to get into speakers but it's just a terrible first step, vastly more expensive to R&D.
I altered the circuit a bit and ran a sim of it running 30v p-p into 8 ohms and the THD came out to 0.000055% which is impressive by normal means but it should be lower, need to reconfigure the design for speaker loads.

 

http://ltwiki.org/LTspiceHelp/LTspiceHelp/_OPTIONS_Set_simulator_options.htm may help you change the number of digits in your analysis.

If your business idea is the power amplifier, you need to have an idea of the specifications you want in a power amplifier. You know, I want the contemporary design specifications: Input ... Power Out ... Load ... THD

Sound good is not an objective specification. That is understood to be the desirable trait. If your building a preamp into this setup, I'll assume you are testing with an inverse RIAA circuit between the generator and preamp input.

on edit: Remember, some audiophiles don't like to play in the sand.

 

If your business idea is the power amplifier, you need to have an idea of the specifications you want in a power amplifier

I have a plethera of designs that are all equally impressive in their own right. All of them have significant technical advantages over equivalent stuff on the market. The things above are just a few examples of some stuff I have. I haven't even touched the surface but I can't really tell you about my other designs without straight up giving away what I did to make them better.

Input ... Power Out ... Load ... THD

I've already factored all of these in of course. Except for THD, I'll say it again, THD is complete marketing BS. Most amps have somewhere around a 0.01-0.001% THD, A 0.001% THD can either be riddled with horrendous amounts of high order distortion well within the DB range of hearing or it can have 0 high order distortion and lots of low order distortion, or it can just have a high noise floor. THD means absolutely nothing, it's just an average of harmonics.

Sound good is not an objective specification

Agree to disagree. It can have 1000% distortion and be home grown in the back garden and as long as it sounds like real life I don't care what the specs are. If my customers care so much they can buy elsewhere, I won't perpetuate falsehoods. Spec obsession is 90% of the reason the entire audio electronics industry is complete garbage right now.

If your building a preamp into this setup, I'll assume you are testing with an inverse RIAA circuit between the generator and preamp input.

Depends on what you mean by preamp. I only have one design that is not capable of gain and that is a version of the tube WCF I designed where I eliminated all of the inherent issues of the circuit other than the obvious tube output impedance which can be dealt with in other ways.
Even my dac stage is capable of gain.
As far as I'm concerned if I need a dedicated preamp it will only be because I want to use it to introduce distortion. It's a bit conflicting because my amplification stage design took me a year to design and is insanely clever in a lot of ways but I may just end up not needing it.
Although I did design an ESL amp that swings 600v p-p with 0.000007% simulated distortion into 1nF or something ridiculous like that, but I don't think I'll be making ESL amps any time soon.

on edit: Remember, some audiophiles don't like to play in the sand.

I was like this too, and people are right to fear the sand. Most often it is not implemented the way it should be. My disrespect for the typical audio engineer is intense. I've never heard a solid state amp I liked ever at any price except the one I built. I still consider myself an electronics novice on the whole. My expertise in electronics is highly specialized. There's no excuse that I should be able to get these results and none of the other "engineers" can. It quite frankly disgusts me.
After going to the last canjam in February I could not believe what was being passed as "good sound" by all of the "high end" vendors. I've never been in such a state of disbelief in my life, it took me weeks to get over it.
I've been to previous events before a few years ago but I really became awake this time. I even asked the engineers about the designs and topologies and most of them were using some of the most textbook designs I can imagine. Some were even using things I could probably pull out of a 1930s "how to" book.
I suspected this for a long time after seeing many schematics of amps people talk about but I did not believe it until I heard it from the horses mouth.
Basically on average what you pay for is a lazy, uncreative design in a fancy expensive chassis.

 

You can't even articulate what your output power is .... Sound good is your only criteria.

Good luck. I'm done.

 

You never asked. You do realize I have nothing close to a final product yet right? I've stated before many times that I have much R&D to do before I'm finished optimizing and create final values.
My output power is whatever you want it to be. Output power is like the easiest spec to design. The output power capability of my current solid state amp prototype is about 1W, it can easily be 10,000W if I really wanted and I had the money.

As I've said I deal in headphones not speakers. No one cares about output power for headphones, even the most difficult to drive headphones require trivial amounts of power.
If you think good sound is my only criteria you haven't been paying attention. Even to a spec junky my stuff is impressive. I can't recall seeing numbers like those anywhere, even in sims.

 

Do you realize you don't have anything written about your design goals. All you do is say it sounds good and has low distortion.

People have asked for your design specifications and you were silent. Now, I will return to my silent mode.

 

Headphones, heavy sigh. Listening to even the best headphones is a different experience than listening to speakers in a quality room. Headphones usually place the sound 'inside' your head while speakers surround you with sound. With headphones things start to sound really weird with some mixes. HI-FI realism where you can feel the music is open speakers in a room.

 

Hey coinmaster........put the real(I) signal in one ear and put the imaginary(Q) in the other.

I/Q audio.......it's a little different than stereo.

 

Do you realize you don't have anything written about your design goals

Um, I've been pretty thorough about my design goals. You haven't been reading or comprehending my posts.
From a sound perspective my goal is to emulate reality while also providing euphony.
From a technical perspective I create ultra low distortion designs and then insert post distortion to avoid having uncontrolled distortion in the system.
If your definition of "design goal" is output power, gain, and load, I can set this to any arbitrary point I want. It's pointless to state these definitively when I only have prototypes built.
They are the easiest aspects of the amp to design, relatively speaking. You seem to imply they are the biggest and most important factors. In my opinion there are two categories of amplifier. Headphone and speaker. In my opinion if either of these classes of amp can't handle 99% of stuff out there then they shouldn't be made, designing for any sane load is not difficult.
There are exceptions, sometimes (or perhaps even most often) more power just makes it sound worse, depends on the design. Low power designs with the specific intent on being low power for sound purposes are okay too.

Headphones, heavy sigh. Listening to even the best headphones is a different experience than listening to speakers in a quality room. Headphones usually place the sound 'inside' your head while speakers surround you with sound. With headphones things start to sound really weird with some mixes. HI-FI realism where you can feel the music is open speakers in a room.

Agreed. The HD800 is the only headphone I've achieved "out of your head" and truly convincing realism on. Most headphones are meh and I've heard all of the good ones except the HE-1, but I heard it wasn't that great from people at the show that auditioned it.
Currently with the HD800 I can be convinced I am at a concert or the singer is in the room with me if I close my eyes. That is difficult to do with headphones.
Depending on the recording it can be scary realistic, I have to look around to make sure there's not someone in the room with me.

 

If your definition of "design goal" is output power, gain, and load, I can set this to any arbitrary point I want.

Yes you can. You just never stated it.

20 mW is NOT a POWER AMPLIER.

I'm not a dentist and I don't feel like pulling every little bit of info from a specification sheet out of you.

 

20 mW is NOT a POWER AMPLIER.

Nope, just the load I was using for the measurement as I pretty clearly stated. 20mw will destroy most headphones.
You're statements make no sense. It should be auto-assumed that whatever amp I design is capable of driving whatever you have. You even admit you understand this. Not sure what your obsession is with output power.
I've done my due diligence and I know my designs function well into 8 ohm loads so there no reason I should care whether or not my headphone amp is not what you consider a "power amplifier" (even though by definition it is).
Scaling up to higher powers is trivial.

 

I'm not comparing apples to oranges. Good luck.

 

I'm not comparing apples to oranges.

No idea what you're talking about. You're not making yourself very clear.
Anyway I gave a speaker amp more thought and this is the result
30vp-p into 8 ohms or about 30W.

It's a thing of beauty....I challenge you to find anything that approaches that performance anywhere at that power level, or at any power level.
Spice just shows 0% distortion and I can't seem to figure out how to increase the number of digits via the link you gave me.
I don't think any real world scope can measure that far down anyway and the practical noise floor will be higher.

 

Now I'm a little confused. Are you saying part of the novelty of your design is this ability to reach very low distortion? I think that, demonstrated in a real device, this would be easy to communicate and get attention. I thought your novelty was ... something else?

 

I thought your novelty was ... something else?

Well, if we are going on a novelty basis I have several of them. Ultra low distortion is just the only one I can and/or am willing to show with a visual depiction over the internet.

 

30vp-p into 8 ohms or about 30W.

How is 30 V pk-pk 30 W?

Quit mixing subscripts.