Nelson Pass is Right - The First Watt is What Counts

ducati_EL34 said:
I think you missed my point. Sure most driving is done to get from A to B. So it even more important to make that as fun as possible. So yeah, driving a high strung car with a manual gearbox is way better than driving a Camry. Like I said, “life is too short to drive boring cars”. I drive a WRX, and I am in boost every time I accelerate!




And to me, making it less of a grind is what it is all about!
Click to expand...
Good luck with that.
 
restorer-john said:
<snip> Here's the ACA's performance, according to NP's own AP plots and the THD+N is absolutely appalling <snip>
Click to expand...

Mr. Pass himself has said in a forum post that he didn't design the ACA to be the best sounding or performing amp but more to be the easiest to build class A amp he could make as an introduction to DIY.
From the history of the ACA: "The Amp Camp Amp was created by Nelson Pass for a special event called "Amp Camp" in which 20 people built their very first amplifier from start to finish in one afternoon." It had to be able to be assembled and fully functional in an afternoon and run off of an easily accessible power supply. It is a well documented and easy step into the DIY world.
My M2 will run circles around my ACA but it was much more involved tracking down parts and building the M2 and easily 3+ times more expensive. It also has 0.05% distortion which is much better than the ACA: http://www.firstwatt.com/prod.html
 
invaderzim said:
Mr. Pass himself has said in a forum post that he didn't design the ACA to be the best sounding or performing amp but more to be the easiest to build class A amp he could make as an introduction to DIY.
From the history of the ACA: "The Amp Camp Amp was created by Nelson Pass for a special event called "Amp Camp" in which 20 people built their very first amplifier from start to finish in one afternoon." It had to be able to be assembled and fully functional in an afternoon and run off of an easily accessible power supply. It is a well documented and easy step into the DIY world.
My M2 will run circles around my ACA but it was much more involved tracking down parts and building the M2 and easily 3+ times more expensive. It also has 0.05% distortion which is much better than the ACA: http://www.firstwatt.com/prod.html
Click to expand...

Indeed. Nelson Pass has done a lot for the DIY community. There's good reason why he's referred to as "Papa" on the diyaudio forum.

-Dave
 
twiiii said:
All this proves is that we each love our own special distortions.
Click to expand...
THD, key word being "total" distortion is such a useless metric to provide any useful information as to how well an amplifier reproduces music. My current VTLs run 3% at full output but sound significantly more like music than SS models with far less. It is profoundly easy to achieve low metrics - just add boatloads of corrective feedback which decidedly compromises behavior in the dynamic domain. Just like cheap op amps.

The challenge is that those metrics are based upon their performance using uncorrelated sine waves. I guess if that's your choice of listening material, it might be meaningful. :)

I challenge anyone to test their ability to discern THD by visiting this site. I took the test about ten years ago with my Dell laptop and Shure E3 earbuds using the following criteria:

Under "Reports", select "6" driver" for Fast Car. Then select "Music T. Chapman"

Try it repeatedly until you get the best results. At the end of each trial, you'll get your results. After about twenty trials, here is my best:

dist.jpg


-36 db corresponds to a distortion level of about 2%. Post images of your results! :)
 
Last edited:
E-Stat said:
THD, key word being "total" distortion is such a useless metric to provide any useful information as to how well an amplifier reproduces music. My current VTLs run 3% at full output but sound significantly more like music than SS models with far less. It is profoundly easy to achieve low metrics - just add boatloads of corrective feedback which decidedly compromises behavior in the dynamic domain. Just like cheap op amps.

The challenge is that those metrics are based upon their performance using uncorrelated sine waves. I guess if that's your choice of listening material, it might be meaningful. :)

I challenge anyone to test their ability to discern THD by visiting this site. I took the test about ten years ago with my Dell laptop and Shure E3 earbuds using the following criteria:

Under "Reports", select "6" driver" for Fast Car. Then select "Music T. Chapman"

Try it repeatedly until you get the best results. At the end of each trial, you'll get your results. After about twenty trials, here is my best:

dist.jpg


-36 db corresponds to a distortion level of about 2%. Post images of your results! :)
Click to expand...


Listen to T Chapman, twenty times?!? No thanks. :D
 
ducati_EL34 said:
Listen to T Chapman, twenty times?!? No thanks. :D
Click to expand...
Not my choice either. I was *challenged* to taking the "Audiophile Repellent" test by a particularly strident acolyte for the mediocre. Yes, I bettered his score by 6 db. His threshold was about 3%.

Note my reply about the content. :)

I reference this test for those who *believe* that double or triple aught distortion figures are: A- audible and B-meaningful. Not a single one of that crowd had demonstrated their ability to actually discern such over the years. What looks good on a piece of paper has little relevance to the world of listening to music.

It's not for those who possess intellectual honesty.
 
I took E-stat's reference to dynamic domain to mean time domain, at least that is the way my mind translated it. I am not sure that is what was meant but there is definitly merit in looking at that as Fourier doesn't really translate the whole picture.

As to how much FB is too much, it depends on the circuit and how it is implemented. You can get away with a whole lot more if it is just over a single stage than across 3 or more stages. The amount of reactance in the loop also has a lot of impact.

One thing that is often neglected is what is going on inside the loop as the FB tries to correct the errors. Overload behavior can easily be impacted in ways that are hard to quantify with our normal measuring techniques.
 
mashaffer said:
I took E-stat's reference to dynamic domain to mean time domain, at least that is the way my mind translated it.
Click to expand...
That and my chosen content for listening is to musical instruments which generate a rich bath of harmonics. The trick is reproducing a complex and dynamically changing signal like that. I have informally observed an inverse relationship between those who *believe* in the importance of simple distortion metrics and their ability to empirically discern it.

Stereophile typically uses a 50 hz sine wave for its distortion distribution metrics. Here is one such example using a VTL Siegfried II amplifier:

sig.jpg

I confess that my listening to 50 hz sine waves is quite limited. :)
 
Last edited:
There has been concern about the degree of feedback correction and audible dynamic compression. I have noticed that some amps sound more "alive" than others.
 
Pio1980 said:
There has been concern about the degree of feedback correction and audible dynamic compression. I have noticed that some amps sound more "alive" than others.
Click to expand...
The VTL amp to which I referenced in post # 92 allows for variable NFB. Fremer's observations support that assertion:

"Low Damping Factor produced a deeper stage and a pleasing softness in the bass. Repeating the track with Medium DF brought MacLeod's voice forward, but increased the focus and tightened bass transients. High DF flattened the image, dried up the ambience, and seemed to limit dynamics."
 
@E-Stat There is a statistical anomaly at -15 dB. My guess is that 12 dB and 18 dB are the first and second points in the test where they end it if you are not doing well, i.e. have made two mistakes. The results distribution graph is less interesting to me than my individual answers, where I got my first incorrect at -12 dB, but later got -12 dB right.

Anyway, I get your point! :)
 
Nelson certainly isn't afraid of some feedback. The F5 uses "generous" amounts of feedback, about half the open loop gain in fact, and I've never heard anybody describe the F5 as lacking depth or dynamics. I doubt the feedback and damping factor were the real culprits. It's more likely the character of the distortion altered the sense of depth. Healthy helpings of low order distortion, even with "generous" feedback, produce ample depth and width with robust, articulate bass.
 
Same old story different day. If you have distortion going in you will have distortion going out. If your amp or other pieces add more distortion then you will hear it if you don't you are better off. A piece that adds even ordered distortions some might find tolerable or even pleasant, those that add odd odd harmonics just the opposite. I Was listening to some early CD's last night and the frequencies above 2 kHz were not correct at all, bright, edgy, and glaring. Then listened to some recordings made in the late 90's by the same company and the sound was so effortless and faithful. I mean a snare drum and cymbals sounded totally authentic. Then I pulled out some pop LP's, yes the mids and highs were very nice just like to days best digital recordings, the bass below 200 hz or maybe it was closer to 150 was all wrong. In the 70's I blamed bad bass on my speakers. Most of it I'm afraid was on the recordings.. Out of all the pop LP's I own there are very few, mostly D2D, that have accurate bass. The producers of D2D Lp's seem to take the extra effort to get authentic sound. Commercial POP guys are to busy trying to be artistic. Oil portraits on canvas can be very pleasing but if I want accuracy I'll take a photograph, that has hasn't been photo-shopped.
 
You must log in or register to reply here.
Back
Top Bottom