If it works as an interlectual discussion is as they say for the birds. Also remember this design is broken in a moment if the other logical steps to change it are taken. That's either broken or sonically broken. Either way it isn't a debate, it's broken.
Depends on whether you use similar transistors. Because shown compensations are all related with that. Even (!) the imbalance 1K-22K is part of the compensation. (With different transistors it may become 1K-1K or even 1K-470.)
Whether it will be close to the quality of original Naim, Traderbam will disagree
Each compensation type has its own drawback. You will never want to use your suggested compensation type with Naim topology, trust me. Even with other topologies (which is not as limited as Naim topology) I don't use such compensations because of its drawback and the fact that it is not necessary.
The key is to do it with base stoppers first, then input filter, then you will have little else remaining to fix. My BJT fast amp can run with only 22pF at the input and the PM is 71 degrees (100pF will bring it to 74 degrees), if Nyquist is the only metrics that matter.
If that design is yours, and if you have the LTspice file, I may be able to show my compensation work. You cannot use exactly the same compensation used by Naim while power supply and transistors are different.
This topology is very hard to make right. It's challenging.
If you have access to copies of Jan Didden's Linear audio publications, you'll find DS has covered output inclusive compensation (OIC) already, though his implementation may bring a lot more with it than you may have been looking for. The article alone can be downloaded for a few Euros here: https://linearaudio.net/articles?title=&page=1
The scheme was actually conceived and encouraged by P Baxandall in personal correspondance, if the attribution of ideas is important to us.
The Baxandall Papers: Transitional Miller compensation
This is working amp using OIC, very stable, very nice sounding, a bit complicate compensation not to everyone taste.
http://www.diyaudio.com/forums/solid-state/243481-200w-mosfet-cfa-amp-92.html#post4580555
Member
Joined 2009
Paid Member
I could write a book all about feedback in audio.
Frequency compensation is only one aspect of many and there are no end of engineering textbooks on all possible variants of frequency compensation. It has been analyzed to death. It doesn't take much effort to plagiarize and write about. No free lunch here.
Frequency compensation is only one aspect of many and there are no end of engineering textbooks on all possible variants of frequency compensation. It has been analyzed to death. It doesn't take much effort to plagiarize and write about. No free lunch here.
Last edited:
Member
Joined 2009
Paid Member
Never mind reading about compensation for these amplifiers. I don't believe there are any real gains to be made in the sound performance of Class AB amplifiers. They have peaked already. They can be made to sound glorious and the knowledge to do so is available for FREE on the internet. We're into diminishing returns and intellectual gratification from exploring myriad 'new ideas' that in reality bring nothing to the table in terms of real performance gains - just different ways to skin the same cat. Improvements may come from a better understanding of psychoacoustics to suggest some optimizations but don't expect any serious investments in Class AB power-amps. Some of the best power output transistors are still those developed in Japan 20yrs ago. If you are a young person interested in amplifier technology go spend time with Class D, that is where you will still find money being spent on developing new technology. Leave the Class AB stuff to the oldies, we'll be out of here in a decade or so.
There, I've said it
There, I've said it
Member
Joined 2009
Paid Member
If you look at the younsters interest in vinyl I would bet it has very little in common with the audiophile voodoo of the oldies. It's somewhat self-evident, there simply wouldn't be such a revival if these youngsters thought it necessary to own 301's. And they'll move away from MP3's over time - they are interested in convenience and that's what MP3's provided - it gave a new life to a dying industry.
Can you point to anybody investing money in Class AB that isn't 'Marketing hype' ?
Can you point to anybody investing money in Class AB that isn't 'Marketing hype' ?
Thank you for that! So well said.
People are arguing over 1, 2, 3-transistor VAS properties and ppm distortion and in the end we close LTspice, disconnect the resistive load from our new built, connect a pair of the thousands possible speakers in one of the billion possible room acoustics and suddenly it's all hit or miss.
Having a deaf guy on the console in the recording studio, having bad day or good day, high or low atmospheric pressure - countless ways to hear things differently.
For me personally it's more about the enjoyment of having a hobby: music collection to your taste, well built equipment to take care off (which suggest it shouldn't be of the Plug&Play type) etc...
As for sound it can always be better. I've been to live performances that sounded worse than my home sound and others that sounded fabulous.
I wish it was about the electronics only as the hi-end industry tries to tell us...
Are you including amp designs found in diyaudio? The circuits I've seen in here look to have plenty of opportunity for real performance improvement. I sort of see a touch of the deaf leading the deaf in most of the recommended improvements; even those that folks quote from local "gurus" and their publications. There are some good ideas and a lot of garbage ones that keep getting recycled like distortion in badly implemented feedback.
So I think it's premature to throw the soldering iron in the bin. There is hope for the budding DIYer knowing that there are undisclosed ways to get really good performance. Naim performance and better.
Last edited:
Member
Joined 2009
Paid Member
ha, well, let's hope so !
But Class D powered fully integrated wireless sound bars may rule the roost !gahhhh!
I'm just thinking that in general, here or elsewhere. I don't think anybody has produced a significant step forward in sound quality from Class AB for several years. There have been some amazing amps developed by people on the forum, far more talented than me. I'm just not sure that the results are pushing the perception of sound beyond the superb quality already available from similarly superb Class AB amplfiers. Unfortunately, I haven't listened to many of them so I'm only casting an opinion here.
In other words, I suspect it's a mature technology (I'm not so sure the Naim S1 offers something revolutionary except for it's extreme levels of implementation - it's a work of art.) and the challenges are now to be found in other areas such as Class D, speaker-room optimizations etc. Technology of today is pretty good, the biggest weakness in my listening pleasures were mostly the quality of the recordings - technical issues were a loooooong way below on the totem pole.
Not to imply Class AB isn't fun, interesting and rewarding - just like Ruwe says, it's really enjoyable! Designing and building your own is the ultimate expression of it. This thread, the TGM10 - NAP160 is not going to create one of those superb SOTA Class AB amps, but it is a journey through history and I find it fascinating.
Last edited:
Have anyone ever designed something (amps, speakers) and when very small change was done, the sound quality changed abruptly?
A little change to a design can make or break. But why? Because there are rules to follow. Like I said before, there are secrets in audio design that designers don't share. You can see how very protective some designers only with their amp schematics. How to make the schematic is more important than the schematic itself.
A little change to a design can make or break. But why? Because there are rules to follow. Like I said before, there are secrets in audio design that designers don't share. You can see how very protective some designers only with their amp schematics. How to make the schematic is more important than the schematic itself.
Member
Joined 2009
Paid Member
Last edited:
I don't think there is any secret in electronic design and audio design in particular. Compared to RF, microwaves or multiphase power electronics audio is actually simple.
It's more about what people know.
I work in R&D of new products (non-audio) and I'm surrounded by about 200 engineers, many of them with PhDs - way, way smarter than me in their respective fields. If there is something that is unexpected we spend time to measure it, understand it, explain it and make it repeatable. Then they re-adjust their design. There are so many stages before a product is on the market: modules designs and tests, system tests and adjustments, engineering prototypes, certification prototypes, manufacturing prototypes and so on.
For the end user or the occasional DIYer it may look like something secret, but anything is there, because it has to be. Otherwise, any engineer would save 1$. Of course, the sales people like the special "secrets" of the product, because that's what sells. Just few hundreds years ago people thought the Earth was flat and all was explained with god's intervention...
I wish to believe that serious audio companies (Naim?) design in similar fashion. Otherwise, it's one man "secrets", because "he said so".
And a competent team always outperform one guy. We see that on this forum. It's rare that someone shares a design that others won't somewhat improve. What is the final quality is another story...
It's more about what people know.
I work in R&D of new products (non-audio) and I'm surrounded by about 200 engineers, many of them with PhDs - way, way smarter than me in their respective fields. If there is something that is unexpected we spend time to measure it, understand it, explain it and make it repeatable. Then they re-adjust their design. There are so many stages before a product is on the market: modules designs and tests, system tests and adjustments, engineering prototypes, certification prototypes, manufacturing prototypes and so on.
For the end user or the occasional DIYer it may look like something secret, but anything is there, because it has to be. Otherwise, any engineer would save 1$. Of course, the sales people like the special "secrets" of the product, because that's what sells. Just few hundreds years ago people thought the Earth was flat and all was explained with god's intervention...
I wish to believe that serious audio companies (Naim?) design in similar fashion. Otherwise, it's one man "secrets", because "he said so".
And a competent team always outperform one guy. We see that on this forum. It's rare that someone shares a design that others won't somewhat improve. What is the final quality is another story...
Last edited:
I made a set of bookshelf speakers several years ago to use in my garage - where my DIY system lives. I have made changes to the crossovers and played with the vent tuning and acoustic wadding and was never satisfied.
I never got as far as finishing them properly however something changed when I put some small rubber feet on them to make them more tuneful and live - it was an un-looked for discovery as in the garage environment bits of grit had gathered underneath and spoiled the furniture finish of what they were sitting on.
Many audio designers are protective of their schematics, because they want people to buy the pretty box and enjoy the sound, without seeing how lame/uninspired/ubiquitous is the actual thing that "sounds soooo great".
Read Stereophile! It's free online. I love it! Read the reviewer report and then check the measurements section with the comments of John Atkinson.
I always have fun comparing the two reports, especially when phenomenal sound is subjectively reported, and then the 65k$ unit completely fails even basic 1/3 power level soak and shuts off due to thermal issues. Even more funny are the letters that the manufacturer sends later to explain themself...
Audio is the only branch of electronics where saying something about secrets will be accepted without loud laughs in the room.
Last edited:
One thing that is hard to understand is the psycho-acoustic: Psychology, Biology (Physiology), Electronics (Physics).
Relating the measurement with perception is very difficult. Especially if you can't perceive sound well. No professors can help you there.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.