Jim Audiomisc
pfm Member
Really? I though they were used:
A) on high power bipolars to help prevent secondary breakdown.
B) on low noise bipolars to make them act as a parallel array to engineer the noise behaviour.
Do you need to be discrete?
- Thread starter BlueEyes
- Start date
Really? I though they were used:
A) on high power bipolars to help prevent secondary breakdown.
B) on low noise bipolars to make them act as a parallel array to engineer the noise behaviour.
mansr
Objectionist
Well, I can't find any at Farnell. Maybe I'm looking in the wrong place.
But discrete are?Except that the people responsible for buying speakers for recording studios don't know far more than you when it comes to buying speakers.
Marketing trumps engineering excellence in terms of getting market share. Always has done, probably always will.
Do not ever use market share as a basis to rule in or rule out any audio component for your personal use. Especially not ones costing thousands of pounds.
Recording studios are just as prone to good marketing as anyone else.
And you only have to look at the ubiquitous use of over compression by the music industry over the last 25 years to realise that studios - as part of this industry - don't give a .... about sound quality.
In the 1950's to 1970's the vast majority of American recording studios used JBL or Altec speakers. Speakers like the Altec 604 are better in some key repsects: dynamics and midrange clarity (especially vocals) than ATC 100's.
Regarding op amps. In hi-fi less is more, when you can get away with less.
What's in an op-amp? Well that will depend on the op amp.
Going to the wiki page on op amps there's a circuit diagram for the 741 op amp:
https://en.wikipedia.org/wiki/Operational_amplifier
Count how many transistors the signal path goes through between the input and the output in this op amp. It's a lot.
Op amps are not compatible with the less is more approach to amplifier and crossover design.
"There is only one thing better than the best component money can buy, and that is no component at all."
You might not think it's a problem. I do. Because I subscribe to the less is more school of amplification.
https://www.pcmag.com/encyclopedia/term/discrete-transistor
Discrete solid state amplification designs where individual transistors are used do lend themselves more to a less-is-more philosophy than op amps.
With the signal being passed through fewer active amplification devices.
BTW, Active ATC 100's are pretty good speakers for low efficiency medium sized ported designs. I'd happily pay £700 for a pair.
How much did I pay for my speakers that you've heard? And how much did I say I'd gladly pay for active ATC 100's?
S-Man
StrivingON
After many 100s of hours of fiddling with circuits I have come to the conclusion that for signal level applications ICs are superior. For power amps I prefer designs based on dicrete transistors. I've tried low and zero feedback circuits at both signal level and power output level and concluded that feedback (used sensibly) is a good thing.
If you want some added flavour it's easier with low feedback and discrete circuits.
Someone mentioned the ETI discrete preamp. I also built this many years ago and realised that tiny changes to the implementation made rather obvious differences to the sound... easy to change the flavour.
I also built a "Curcio Daniel Vacuum Tube Preamp". It was easily beaten by a few NE5534s.
If you want some added flavour it's easier with low feedback and discrete circuits.
Someone mentioned the ETI discrete preamp. I also built this many years ago and realised that tiny changes to the implementation made rather obvious differences to the sound... easy to change the flavour.
I also built a "Curcio Daniel Vacuum Tube Preamp". It was easily beaten by a few NE5534s.
Paul R
pfm Member
Weren't we in the same thread with the same OP 10 years ago?
It's ATC, the 'discrete' will work at least as well as the standard, but will scratch the audiophile itch. One advantage I recall is that they turn off silently whereas the standard versions grumble a bit.
Opamps are NE5534 for the record. Detecting whether the speaker is powered on by ear requires a keener ear than mine.
It's ATC, the 'discrete' will work at least as well as the standard, but will scratch the audiophile itch. One advantage I recall is that they turn off silently whereas the standard versions grumble a bit.
Opamps are NE5534 for the record. Detecting whether the speaker is powered on by ear requires a keener ear than mine.
John Phillips
pfm Member
It seems to be axiomatic that for any Internet expert opinion there's always and equal and opposite Internet expert opinion. For example see "MYTH: DISCRETE IS BETTER" here.
So answering the original question through others' views reduces, I think, to who do you believe.
Or maybe it comes down, for me anyway (as I concluded years ago), to deciding what pleases me for any reason - not just how I perceive the sound. And believing primarily myself based on arranging to get my own experience. That may be the only valid way to answer the original question.
only once you've got your power cable sorted, mind.
S-Man
StrivingON
Jeff Rowland suggests AD746 as an upgrade to his earlier discrete circuits:
https://www.diyaudio.com/forums/sol...rence-1-schematic-modules-11.html#post5840159
https://www.diyaudio.com/forums/sol...rence-1-schematic-modules-11.html#post5840159
Jim Audiomisc
pfm Member
In general, the data lists for bipolars don't mention this. They just note the noise level or SOA, etc, you get as a result. Some fuss was made (mumble) decades ago when the technique started being used, but since then the actual internal geometry of devices tends to pass without comment.
For low noise devices the parallel (pun alert!) is obvious when you see people using paralleled devices to get improved noise performance. But people rarely think about secondary breakdown - particularly now the changes to power devices have dealt with it to a much better extent than in the distant past.
IIRC The last maker's data book I've seen which actually showed a drawing of the die geometry was the now-ancient 'Silconix' one on their excellent FETs.
mansr
Objectionist
If it looks the same from the outside, I don't care how it's made on the inside. Results are what matters.