I did, Scott, you know me, I'm all Greased Lightning so I tried e-mailing the schematics yesterday, but all I got is two bounces. One of our servers is not happy with file sizes.
Not to worry, I'll do it again in a few minutes, but in several messages, to ease off on the per message attachment sizes.
That puts us slap bang on the same page, Frank.
And that page is in a book called "Kama Sutra".
I think the truth is out there somewhere between these two comments, Frank's and Wayne's.
The fact is that measurements tell me nothing about how something actually sounds, so "metrics" in the classical sense mean nothing to me. However, some more involved measurements, such as harmoncs decay behavior, rise ans settling times, square wave performance and such would be most welcome.
Having heard that "small" Krell, I must say I was hardly flabbergasted by it; nice, warm sound, but nothing to write home about.
On the other hand, I have heard many implementations of chip amps, mostly based on the LM offerings, and they performed within a wide variety, from dull to surprisingly good. The common points of those which sounded really good were generally two items: most had another FET input op amp before them installed as a buffer, and all had rather good power supplies, including large toroids (you'd be surprised at the wide variety of choice regarding locally manufactured toroidal transformers).
Frank, if memory serves that small Krell's toroidal transformer uses sintered materials for its core, rathen than classic plates; this is easily the better solution, but also hard to come by and expensive. In return, it is capable of delivering something like twice its nominal power under short term conditions, something no classic toroid can even dream about without serious core saturation and hence very distorted output signal.
So, however you look at it, in the end, it always comes down to having to audition it yourself, in your own room and with your own system to REALLY know how it sounds.
The fact is that measurements tell me nothing about how something actually sounds, so "metrics" in the classical sense mean nothing to me. However, some more involved measurements, such as harmoncs decay behavior, rise ans settling times, square wave performance and such would be most welcome.
Having heard that "small" Krell, I must say I was hardly flabbergasted by it; nice, warm sound, but nothing to write home about.
On the other hand, I have heard many implementations of chip amps, mostly based on the LM offerings, and they performed within a wide variety, from dull to surprisingly good. The common points of those which sounded really good were generally two items: most had another FET input op amp before them installed as a buffer, and all had rather good power supplies, including large toroids (you'd be surprised at the wide variety of choice regarding locally manufactured toroidal transformers).
Frank, if memory serves that small Krell's toroidal transformer uses sintered materials for its core, rathen than classic plates; this is easily the better solution, but also hard to come by and expensive. In return, it is capable of delivering something like twice its nominal power under short term conditions, something no classic toroid can even dream about without serious core saturation and hence very distorted output signal.
So, however you look at it, in the end, it always comes down to having to audition it yourself, in your own room and with your own system to REALLY know how it sounds.
This is the heart of the matter ... "what it sounds like" is the only thing, the only thing I'm concerned with. Back in the 80's I went the rounds to realise that expensive means nothing, highly spec'ed means nothing, what it actually does in your listening space is the only metric that's meaningful.
And it's the little things that matter, the one flaw in the 1/2 million setup drags the sound down into the pits, makes it impossible to live with - its only value is as a show off bit o' kit to one's friends. Hence my 'shtick' has been "refinement", extracting the optimum from whatever's at hand.
I've done the 'drag the monster amp back from the dealer as a trial' thing a couple of times, decades ago - and the results have been abysmal: disturbing the optimised setup to stick in an unknown throws the sound backward so far that I lost interest almost immediately.
Expensive, exotic means nothing; cheap, even 'nasty', means maybe ... . Real, genuine, 60W power means superb sound in your room, irrespective of it being a chip amp or some bling encrusted monstrosity. I've got a great deal of satisfaction pushing nothing gear to deliver highly satisfying sound - it's proving a point, at least to me.
My goal is for the sound to always "work", irrespective of the supposed quality of the recording, and irrespective of the volume level. So, it should be a pleasure to listen to when it's so quiet that you barely notice it's working, right up to window rattling SPLs - having achieved this regularly, when all the right efforts are made, this is a very realistic goal.
The "unknown", the challenge in all this, for a particular system not behaving this well, is tracking down, identifying what's undermining the sound quality - and rectifying those issues ...
Last edited:
It depends on what your aim is, Pavel. I appreciate that you are concerned with highly competent technical performance being a key factor, and that obviously figures very strongly in many areas. However, my take is that using "lesser" equipment parallels the situation of playing a relatively lower quality recording - and my intent is for that recording to deliver all that it is capable of, for the musical message to get across cleanly. So, for 'mid fi' gear the aim is to eliminate every handicap that worsens its chance of making a good showing, and don't do anything which highlights areas of weakness.
A 'plain' girl and a standard kit of makeup material: that person could end up looking highly attractive; or a somewhat offputting trollop - it's all about the skill and knowledge brought to bear ...
A 'plain' girl and a standard kit of makeup material: that person could end up looking highly attractive; or a somewhat offputting trollop - it's all about the skill and knowledge brought to bear ...
Sounds like you should get out a bit more, a.wayne ...
Can you put on, say, AC/DC at a volume where you have to shout in the ear of the person next to you for them to pick up what you're saying ... but the sound is clean as a whistle, the drums and cymbals are "there", clear as the real thing ...?
That's what the LM38xxs can do, if you treat them right ...
Can you put on, say, AC/DC at a volume where you have to shout in the ear of the person next to you for them to pick up what you're saying ... but the sound is clean as a whistle, the drums and cymbals are "there", clear as the real thing ...?
That's what the LM38xxs can do, if you treat them right ...
I don't think there's a componente ever made, by anyone, which couldn't be made to sound good - it's not the components so much as it is HOW you use them.
Pavel, let me remind you how op amps were regarded as low fi in the early 80ies - and then the Burmester preamp arrived. And EVERYBODY agreed that this product, built around the lowly, unexciting specs OP27 op amp sounded divine. And it did, too bad they never made anything like it again.
Personally, I much prefer to use discrete components - it's a lot more hard work, to be sure, but it does let one be in full control of the sound all the way.
Full integration does have its benefits, just as it has its shortcomings. With them, the thermal behavior is incomparably better than with any discrete circuit simply because its all on one die in one package - don't tell me this is of no consequence.
The key problem with chip amps, in my view, is that they are usally incomplete, so to speak. The manufacturers are to blame here, as they advertise them as quick and simple solution, while practice shows them to be not quite so simple as they would have us believe in their best interest.
Here's a small, easy and simple test - throw in a FET input op amp, even a simple one like the LF 356 as an input buffer in front of the LM chips, make the 1,000 uF caps near to the chip amp not your only caps but your last in line caps, preceeded by say 10,000 uF caps, and be prepared to be surprised at what you hear. Ah, but adding two 10,000 uF caps cost money, and if you're doing that, you will have probably used not one but two say 300VA toroidal transformers, and while the sound you get will be rather good, it is no longer a simple and cheap solution - so nobody does it.
Using them in balanced mode, and as far as I am aware only Jeff Rowland did that, is even more rare - now tell me Jeff Rowland is low fi or mid fi.
For the very best, you probably still have to rely on discrete components, at least I do, but let's be honest - for most people, a well implemented chip amp will probably sound better than the rest of their system. Their bang for buck is wiping the floor with the so-called high end, which has let me down oh so many times. "Most people" have no idea of what system synergy means. Frank does.
Pavel, let me remind you how op amps were regarded as low fi in the early 80ies - and then the Burmester preamp arrived. And EVERYBODY agreed that this product, built around the lowly, unexciting specs OP27 op amp sounded divine. And it did, too bad they never made anything like it again.
Personally, I much prefer to use discrete components - it's a lot more hard work, to be sure, but it does let one be in full control of the sound all the way.
Full integration does have its benefits, just as it has its shortcomings. With them, the thermal behavior is incomparably better than with any discrete circuit simply because its all on one die in one package - don't tell me this is of no consequence.
The key problem with chip amps, in my view, is that they are usally incomplete, so to speak. The manufacturers are to blame here, as they advertise them as quick and simple solution, while practice shows them to be not quite so simple as they would have us believe in their best interest.
Here's a small, easy and simple test - throw in a FET input op amp, even a simple one like the LF 356 as an input buffer in front of the LM chips, make the 1,000 uF caps near to the chip amp not your only caps but your last in line caps, preceeded by say 10,000 uF caps, and be prepared to be surprised at what you hear. Ah, but adding two 10,000 uF caps cost money, and if you're doing that, you will have probably used not one but two say 300VA toroidal transformers, and while the sound you get will be rather good, it is no longer a simple and cheap solution - so nobody does it.
Using them in balanced mode, and as far as I am aware only Jeff Rowland did that, is even more rare - now tell me Jeff Rowland is low fi or mid fi.
For the very best, you probably still have to rely on discrete components, at least I do, but let's be honest - for most people, a well implemented chip amp will probably sound better than the rest of their system. Their bang for buck is wiping the floor with the so-called high end, which has let me down oh so many times. "Most people" have no idea of what system synergy means. Frank does.
LM 38XX are horrible, regardless how you treat them. It's a poor man's low highend. You will never get rid off their limited current output capabilities, quite poor stability and the residuals of cross-over distortion. For some, it may be satisfactory solution, though.
Heat dissipation problem in the power chip is a physical reality.
Heat dissipation problem in the power chip is a physical reality.
You must have had some bad experiences, Pavel. Asking them to drive some hideous XO may have elicited poor behaviour, but in reasonable, average sensitivity 2-ways there was no untoward behaviour - sustained operatic notes by sopranos at maximum volume caused internal protection to kick in ... but they were the only audible glitches.
Heat dissipation? I thought that was the idea behind decently rated heatsinks ...
Heat dissipation? I thought that was the idea behind decently rated heatsinks ...
Hmmm, in my experience when a system is sitting 'right', and paint peeling loud, no shouting is required.Sounds like you should get out a bit more, a.wayne ...
Can you put on, say, AC/DC at a volume where you have to shout in the ear of the person next to you for them to pick up what you're saying ... but the sound is clean as a whistle, the drums and cymbals are "there", clear as the real thing ...?
That's what the LM38xxs can do, if you treat them right ...
Very loud talking is required, but no shouting.
Dan.
We had this one before, Dan, ... depending upon the material, the average sound intensity is such that the human automatic ear protection mechanism cuts in, and attenuates the hearing sensitivity - the upper midrange seems to be most relevant here, IME.
With driving rock this is most noticeable, the same volume setting for classical is barely perceived as being 'loud', the sound is just "normal" - and some CDs are mastered so softly that max. is no more than conventional TV volume ...
Some of the OTT modern stuff, say Foo Fighters, 3/4 volume is the most one would want to cope with - any more is asking for bad ear ringing ..
... depending upon the material, the average sound intensity is such that the human automatic ear protection mechanism cuts in, and attenuates the hearing sensitivity - the upper midrange seems to be most relevant here, IME. With driving rock this is most noticeable, the same volume setting for classical is barely perceived as being 'loud', the sound is just "normal" - and some CDs are mastered so softly that max. is no more than conventional TV volume ...
Some of the OTT modern stuff, say Foo Fighters, 3/4 volume is the most one would want to cope with - any more is asking for bad ear ringing ..
Last edited:
You just hit the nail on the head, Pavel, by saying "a poor man's low highend".
If one believes the manufacturer, and takes it that way, in my view one is making two mistakes: 1) they were never intended to be any kind of "high end", but a relatively clean and cheap solution, and 2) the dichotomy surfaces - people want a CHEAP "high end", which does not exist, and never really implement the chips properly.
Remember the endless pages on forums written about them, which claimed that they sound the best with just 1,000 uF caps? What kind of filtering do people expect from 1,000 uF? What kind of energy, so needed for low impedances, can 1,000 uF caps store?
While hardly perfect, their problem is not so much their design as are the fact that most electronic dummies out there believe it's a very classy solution on a par with seriously designed electronics and the fact that just about every square head idiot out there started to believe that they were actually ingenious designers who "saw through the game".
Their key problem is that the level of intelligence of potential users is dropping sharply, but then the same could be said of audio in general. It's now an old game and has been replaced with home theatre, PC, CD, Blue Ray, mobile phones and whatnot, so everybody thinks they can deal with such simplicities as two caps. All the knowledge they need is on Wickipedia.
We live in a world in which fast food has been elevated to the level of haute cuisine. The chip amps have thus been perverted like most other things, and the economic crisis didn't help any.
Last edited:
Higher order harmonics (esp the 5th) are the key to ear shelf protection I think.
Dead clean audio and that ear shutdown does not occur nearly so readily.
Also one can then communicate quite easily over the top of very loud music, AC/DC included, and without the shouting factor.
Discos are typically the worst.... body language becomes the only reliable form of communication
.Dan.
Sorry, there are some forms of, and moments in, loud music which is totally acoustic, where it is completely impossible to hear yourself, let alone anyone else. I recall standing 10 feet away from a decent sized big band in full cry, and the sound was completely overwhelming at certain times - it filled the aural universe and completely snuffed out all other sound ...
- Status
- Not open for further replies.