This is becoming a "can of worms".
Indeed, topology IS the important thing, as it is for speakers.
I... think that perhaps I need to clarify my opinion on this matter. I agree that typology is important, but Typology isn't the only important element. I believe this is valid for amplifiers and speakers.
I remain convinced that an optimized typology for a loudspeaker crossover is essential to the good results. I would much rather have mediocre component quality with a well optimized crossover typology than great quality components with a poorly implemented typology. My very first DIY speaker kit used a SEAS T25-001 tweeter and a Scanspeak 8545 midwoofer and it sounded terrible - until a gracious gentlemen redesigned the crossover. Then... oh, my!! It was wonderful!
When I started in this hobby 7 years ago there weren't many folks capable of creating an extremely good loudspeaker crossover in the DIY realm. Dennis Murphy, John K, Phil Bamberg, Geoffrey Dillon and George Short were very capable. Certainly Mr. D'Appolito and Mr. Linkwitz were capable on the commercial front too. I heard a very solid rumor that many commercial folks were using a modeled crossover (not rigorously tested) in their commercial speakers (more on this later). There were certainly other capable folks but not too many of them.
Today there are many more solid designers. Heck, I met a young guy named Chris D'Allesio (spelling?) that did very good work. Also, Scott Sehlin is solid, and so is Curt Campbell. There are also the Packer fans; Scott, Nate and Matt. And... Mike Bengfort. All of these guys are very capable grass-roots DIY guys, and this is just in the Midwest region. And... Dave Ellis managed a "quickie" crossover using the Accuton C95 and OW1 in about 4 hours that was on par with the 1801.
I don't intend to infer that obtaining an optimized crossover is something that could be done in a weekend by a rookie with some modeling software. My effort follows, oh, @ 6000 hours of reading and testing. Other folks are surely smarter and are able to climb the learning curve quicker. I do intend to infer that many folks are capable of good typology.
I believe this is true among speaker guys and amplifier guys.
The only caveat I will offer (and repeat) is that a few amplifier typology issues are VERY complex. As mentioned previously, the fully regulated power supply in the very expensive Levinson is profound.
After obtaining good typology, parts quality will provide that extra 10% of performance that is so difficult to obtain. It can also be detrimental. Many electrolytic capacitors are labeled with 10-20% tolerance because this is a reality for many components made in China. IME any decent capacitor will be @2% tolerance. And, this is on the obvious measurable differences.
Hererin I will offer some questions on the fringe of my knowledge that I believe impact manufacturing quality. I don't have the answers to these questions in the amplifier realm and some of them may be a non-issue. Nonetheless, I believe many of these have substance for amplifiers and loudspeaker crossover components.
What happens when the film used in a poly capacitor becomes a bit thin/porous? What happens when the chemistry of the film changes from 1 roll to the next. What happens when the metalized material crystallizes at a different rated due to ambient temperature variation? What happens when the spray nozzle depositing the metalized material becomes partially clogged. Why aren't all capacitors annealed. Is all end-spray the same? Are there more quality control problem areas????
In the speaker realm...
Are all voice coils perfectly centered in production run? Do all surrounds have the same chemistry? Are all surrounds the same thickness? Do all tweeter domes have ideally dispersed coating material? Are all paper cones the same mass? Are all voice coils concentric? Are driver motor tolerances always consistent? Are all top plate and pole piece assemblies precisely machined?
The answer to all of the speaker questions is mostly "no" - in varying degrees of extremity. Some folks do an extremely good job with these things. Other folks have, ahem, less than ideal results.
My point in rambling, and am rambling, is that loudspeaker/amplifier research & development and the brilliant implementation by engineers when building a component is most often NOT commensurate with the end result. The board members and accountants must realize a profit, and this often necessitates the use of inexpensive components and their commensurate quality.
Indeed, topology IS the important thing, as it is for speakers. I prefer average priced caps on a 1801 than silver/gold on...some other speakers
Primarily, the focus of this string is amplifiers, not speakers.
I believe amplifiers are much more sensitive to component quality than loudspeakers. This assertion has significant room for interpretation... so I'll clarify one aspect.
Upgrading from a Solen Capacitor to a Sonicap behind the tweeter of the 1801 is modestly audible. After living with them for several years, I will always use a Sonicap. However, I can easily understand that folks may wish to use the solen or something similar. It has a touch more grit, and some folks may prefer this. It adds splash and sizzle. Also, I think the use of a Mundorf Silver/Oil is completely unwarranted in a tweeter circuit.
Upgrading from a Solen Capacitor to a Sonicap in a coupling circuit of a tube amp is significantly audible. This upgrade was earth-moving for me when first accomplished in my CD player.
IMO, using a good quality coupling capacitor has 5x more impact than using a good quality tweeter capacitor.
I also feel it's somewhat fair "evidence" to use the results of the folks on the AKSA forum who upgraded their AKSA amps to Nirvana + versions and their results. All commentary on this matter is extremely positive. Better components DO matter.
In summary, I believe many folks can obtain a very good typology and this remains absolutely necessary. After this, the quality of the components used is decisive.
Dave
