Sound Quality Vs. Measurements

Status
Not open for further replies.
However, while improvents will happen, this cannot turn a so-so device into a High End one no matter what we stick in. In the end, the circuits will be the limiting factor.
Personally, I don't see that. Improvements in the PS side of things have always given me great gains, and this doesn't necessarily require the supply itself be modified - simply ensuring the mains in is much cleaner allows a so so power supply to lift its game sufficiently so then the fed circuit performs well above what one would normally expect.

If I had make a choice, either between excellent or mediocre PS, or the same options in the amplification area, I wouldn't hesitate ...
 
Those were the days when speaker equirements were thought of only in terms of avalable power in Volts. Few thought of the possible current requirements as well.

Look over the SAE schematics of the day and you'll see the same thing - single ended outputs, sometimes with stacking, for power outputs as high as 200 Wrms/8 Ohms (model 2200, 2400, etc). And that was no novice on the block, that was James Bongiorno.

Such were the times.

By the early 80ies, this changed for the better.
The trouble with the ESL-57 is that it is almost pure capacitance above 10 kHz. High power into 8R is completely irrelevent and actually damaged the ESL-57. There were a few >8R speakers around in the UK, were there many in the USA needing these high voltage swings?
 
Personally, I don't see that. Improvements in the PS side of things have always given me great gains, and this doesn't necessarily require the supply itself be modified - simply ensuring the mains in is much cleaner allows a so so power supply to lift its game sufficiently so then the fed circuit performs well above what one would normally expect.

If I had make a choice, either between excellent or mediocre PS, or the same options in the amplification area, I wouldn't hesitate ...

And you talk to me, who makes and sells power line filters, sorry, "conditioners" in newspeak, about input power cleaning? :D :D :D

But that notwithstanding, Frank, if a modest PSU upgrade produces large benefits, then I must conclude that the PSU you started out with was junk. If the PSU was initially good, you'd have to make a big upgrade to hear any significant difference.
 
The trouble with the ESL-57 is that it is almost pure capacitance above 10 kHz. High power into 8R is completely irrelevent and actually damaged the ESL-57. There were a few >8R speakers around in the UK, were there many in the USA needing these high voltage swings?

I've been told by quite a few Americans that the rooms with speakers were larger to much larger than those we generally have in Europe. Consequently, more power is need by default.

I think it no coincidence that some of the most efficient speakers ever made, from people like JBL, Altec Lansing, Klipsch, Cervin Vega et al., delivering 104 dB/1W/1m, originated in USA, admittedly mostly during the times of low powered tube amps. If one's living room has the same area as my entire apartment, they needed them.

A rather radical friend from Boston blames AR for dropping speaker efficiency to sub-90 dB. Indeed, my old AR5 speakers were rated at 87 dB/1W/1m, while my current speaker do 92 dB/1W/1m, which is 3.16 TIMES more efficient.

That, in conjunction with my small room, is quite happy with a nominal output of 20Vrms into 8 Ohms, never even breaks sweat. At the volume levels I use, no amp had ever even become moderately hot in my room.
 
But that notwithstanding, Frank, if a modest PSU upgrade produces large benefits, then I must conclude that the PSU you started out with was junk. If the PSU was initially good, you'd have to make a big upgrade to hear any significant difference.
Well, thinking it over, most of the supplies I've played with in commercial units have been pretty unimpressive. Very conventional, poor energy storage ... the unit I did for the gainclone was the nicest, ;) ...

Then again, I see defects in the sound which I have learnt to link with PS problems, more so than most people ... :)

One surprise: my current Aldi TV experiment has shown that progress has been made - would be using a cheap switcher, yet is able to produce good sound at decent levels with no mod's - biggest weakness is that it's very sensitive to interference effects.
 
dvv said:
In my experience, swapping some non descripit caps of say 4,700 or 6,800 uF, as are mostly used in cheap gear, for quality caps from a reputable source will improve the sound.
maybe, but its not 'designing' is it? you keep on pushing my posts back at me as if I said these things make no difference, i'm a bit of a PSU fetishist, but as you said, such an upgrade (presuming it actually was an upgrade) can make a difference to an old tired amp, or a crap amp, which would leave you with a crap amp with new caps. Changing the caps in a circuit with reasonably high PSRR its not going to change it into a world beating amp is it?

i'd rather not carry on with the back and forth, because I think the 2 of us are basically on the same page, but i'm not sure you understand what you are really up against here with Aldi man ;)

Aldi is a supermarket that sells offbrand, or overstock food at discount prices, along with flavors from known brands that didnt make it past the taste testing. its actually not a bad place to do some of the shopping, but some of it is just plain wrong. not somewhere I would expect to find hardware to use in my audio system, no matter how many mods. modded crappy hardware is never going to measure up to hardware designed properly in the first place.

I like Siemens, Kemet, or RIFA, but i'm building amps with fully regulated linear and switchmode supplies nowdays, so such large caps are not so important, or even desirable.
 
I tend to agree with your approach qusp. I don't fault those that do, but this whole idea of over modding every circuit gets carried to extreme. And how does one truly measure the objective advantage of one part over another and how it relates to sound quality? How would one ever switch in one lonely part (say a capacitor), then do a double blind test with unbiased listeners to get any kind of reasonably scientific results?! I argue it simply can't be done without a large amount of subjective bias. "I changed a component, so it must sound different." Then repeat the same whole process with another component, and another, etc. You get the point. By then, any PCB would look like a mangled mess after soldering and un-soldering multiple components! It simply is an unruly exercise/process that is fraught with too many variables. And in the end, it comes down to a subjective evaluation by the listener. And how do know which component actually played the most significant part in 'changing' the sonic characteristics of whatever you are modding? My contention is most hobbyists are not going to do that simply because of the time and trouble factor involved (the whole scientific testing approach). I look at most mods with a very skeptical eye and hence don't often engage in the practice. There are a few known improvements that can be made, but OTOH, there are so many opinions here on DIY, with very few hard, substantive facts to base reliable and effective mods on. :) You have to wade through pages and pages of posts to often find one gem that is credible. Much more snake oil!

Like you said qusp if I can paraphrase a bit, "if it's designed properly, chances are it will sound pretty damn good to start with". :D As soon as you go and change one component, you've effectively changed the design and the designers intention. I guess I'll take some heat now. :)

Rick
 
@qusp

Of course changing low cost, low quality parts for better ones is not designing, it's simply upgrading, nothing else. A design is actually made only once, in the initial stage, that's when you set the ground rules once and for all. In most cases I wanted a change, I never made it because I would have done it differently from the start, i.e. choosing the topology.

I did say that no matter what you change for what, you cannot transform a low quality design into a top flight one, ever. At best, you can let it release all of its potentials, but they could be limited by the nature of the design (and most often are).

And yes, as far as I can see, we have been on the same page right from the start, at least that's my take on it. Worst case, our semantics are a bit off. Or we just express ourselves differently.

Just to make sure, let me say this: if I go for "tuning" a device, this means chucking out low quality caps and inserting what I see as quality caps and possibly replacing critical (input stage and NFB line) standard 5% carbon resistors with quality (e.g. Dale) 1% metal film resistors. And DEFINITELY changing usually crappy local adjustment pots with better quality multiturns. And installing an ALPS Blue pot for volume and hopefully balance, that never fails to improve things at least a bit, often more than just a bit.

100% parts replacement, hence tuning. 0% circuit changes, so 0% designing.
 
The really amusing aspect of this is that the sound of the 606 is completely dictated by relatively ordinary opamps; it is essentially a chip amp with power booster stage. It was used by Martin Colloms as a reference amp for some time, he liked it a lot ... ;)

The sound of the 606 isn't dependent on an op amp: the sole op amp is used for DC bias. The 405 was different.

http://www.keith-snook.info/Schematics/QUAD 606 Schematic circuit.pdf

A question:
The TLC271 op amp is 'programmable' and has a selectable 'bias mode' which changes its characteristics. I notice that in the Quad 606 this bias select input is linked to the op amp's output... Any idea why?

http://www.ti.com/lit/ds/symlink/tlc271.pdf
 
diyAudio Member RIP
Joined 2005
A question:
The TLC271 op amp is 'programmable' and has a selectable 'bias mode' which changes its characteristics. I notice that in the Quad 606 this bias select input is linked to the op amp's output... Any idea why?
That is a trifle bizarre. Since higher bias corresponds to a more positive voltage on pin 8, which in turn increases gain/bandwidth and reduces voltage noise, this would mean that the integrator formed with that part would change slightly in response for required positive correction outputs versus negative ones. Overall, with the large amount of loop gain, the differences shouldn't be that substantial. But if the schematic is accurate there must have been a good reason to do it. Perhaps it pertains to a power cycling transient control.
 
The sound of the 606 isn't dependent on an op amp: the sole op amp is used for DC bias. The 405 was different.
Bloody hell!! The years are catching up on me - you're dead right! I was thinking of the Meridian 605 monoblocks - my memory triggered on the almost right number, and being Brit ... got the important bit wrong!

As someone used to say -- missed it by thaat much ... :)
 
Poor qusp still doesn't get it - the point of the exercise is not whether my design is bigger than your design, that sort of thing should have been left in the playground a long time ago ...

The goal is to achieve satisfying sound, everything else is subservient to that, including original design. Massive metal boxes sparkling with bling will be instantly sacrified, thrown into the crusher if they don't perform the task required, irrespective of cost, and number of output devices - all that is irrelevant, only required by those who have a need to compensate ... ;)

I'm prepared to look at anything, use any technique, provided it helps me in my goal of getting sound that is good to listen to. And that especially includes "forcing" very low cost gear to perform, becauses that tells me a lot, teaches me many things. I do note that not a single person has apparently tried the "Aldi" exercise, that tells me a lot about the level of curiosity, the desire to learn more, of people in this forum ... ;)

My drive is to explore, to understand the factors that contribute to good sound, and everything is a tool towards that end: my explorations, other people's experiences, press reports, Spice simulations, large scale modifications, even completely original design. If tomorrow I thought of an exercise to try and destroy a persons' house by setting up powerful bass drivers, and monitors, to find the resonant frequencies of the panels of the structure and then driving each to the point of failure, I'd probably give it a go - that's the sort of chap I am ... :D
 
save your pity for yourself.

lovely story you made up about my outlook, based on nothing at all, but thats true to form. I dont care to learn about crappy designs built to last a few months, or torture myself with the results to test for 'transparency' of the type that somehow makes bad recordings more listenable, without adding anything. Its quite a feat that one, but apparently very common.. I could make something more suitable for music playback just from parts in my parts bin, not terribly exclusive thinking I know... sorry to disappoint.
 
Last edited:
Aldi is a supermarket that sells offbrand, or overstock food at discount prices
Do they actually sell electronics at all, or just reject food?
(i.e. is it even possible for Frank to own an Aldi TV, or has he been listening to some kind of vegetable all along?)
:confused:

p.s. Admittedly I'd rather watch a vegetable than most of what's on TV these days (which is why I don't own one (a TV that is, I have several vegetables)).
 
Don't have Aldi over there? I though they were a global 'disease', :D.

Like Ikea, they specialise in buying huge job lots of output from factories, etc, these days typically Chinese, like everyone else. Put out as weekly 'specials', one week will be gardening thingies, next will be bedding, including beds(!), then myriads of cooking nick nacks. It's the way, for example, to buy a cheap Android tablet, just have to wait until the electronics turn cycles around ...
 
Last edited:
Status
Not open for further replies.