Good points thanks - food for thought. So if this really is the main cause of disappointing SQ from many opamps that look good on paper, you could say we're stuck with needing to use RF resistant opamps (or discretes) as there's no convenient way to prevent these kinds of noise getting in completely. Perhaps use of additional LC filters in the power supply would help with mains noise, and using balanced interconnections should sort out noise picked up on cable shields, but what can you do about DAC OOB noise...
So did you measure any differences between the same circuit using the dual and the single versions? What were you measuring?
I have a digital 200MHz scope and an analogue 50MHz scope. Normally I can see what I want to see more quickly and easily on the analogue scope - eg ringing on square waves and signs of instability. Though the averaging can be handy on the digital scope. But rarely has the 200MHz revealed anything that wasn't visible on the 50MHz. What sort of things were you getting at?
Have you compared the SQ of your designs to others, perhaps with some unbiased fellow listeners and in a blind fashion? Are you experienced at listening to acoustic concerts and know what natural instruments should sound like? Sorry to ask, but I think these are reasonable questions given you're telling us how great your own designs sound...
My experience has been that the LM4562 sounds artificial, it is tonally "hard" and unpleasant to listen to. I found I wanted to switch the system off and stop listening. This is normally a good sign that something is wrong! I was unable to see any signs of instability in the circuits I tried it in, but who knows what else could have caused it. But since the two circuits were completely different designs I tend to think it's down to the opamp itself.
I am not making any comparative claims about how 'great' my designs sound over anyone else's designs - I am quite sure there are many other great designs out there. I simply stated that the 4562 sounds good and yes I have listened to a lot of equipment over the years so I think I am qualified to comment.
You ask if my stuff has been listened to by someone that is unbiased. I might ask you the same question about the sound of your construction.
I recently had two cases where I had HF oscillation up at over 100MHz that would have been visible with a low bandwidth scope. 50MHz is not bad, and clearly your digital scope would find a lot of this type of stuff as well.
You ask if my stuff has been listened to by someone that is unbiased. I might ask you the same question about the sound of your construction.
I recently had two cases where I had HF oscillation up at over 100MHz that would have been visible with a low bandwidth scope. 50MHz is not bad, and clearly your digital scope would find a lot of this type of stuff as well.
That's precisely what my latest blog entry is designed to ameliorate : http://www.diyaudio.com/forums/blogs/abraxalito/968-line-input-rf-reject-filter.html
I know its a while back, post #213
Would those pots you mention be the Omeg ECO types as here, or one of the other (metal enclosed) types Omeg produce ?
http://www.diyaudio.com/forums/part...nductive-plastic-vs-cost-other-questions.html
I've "studied" Doug Self's preamp and I enjoied it. It is quite interesting and there are a lot of "tricks".
There is one of them I'm not shure I have fully understood the way it works.
I'd like if Mr Self could confim (or correct) my guess
I'm not sure I can publish the schematic on this forum so I'll refer to the schematic published on Elektor.
The circuit I refer to is named "load synthesizer"
Self writes:
"A load-synthesis circuit around IC4 is used to make an electronic version of the required 47 KOhm resistor from the 1 MOhm resistor R16. The Johnson noise of this resistor is however not emulated"
with no further details of the circuit.
An analysis by inspection of the circuits tells:
RL= R16 / (1 - A)
where
RL= is the "synthesized" resistor
A= is the gain of IC4A (the gain is inverting so A=~ -16)
Vn= en / A (on the first order and neglecting OP-Amp voltage and current noise)
where
Vn= is the noise of the synthesized resistor
en= is the Johnson noise of R16
A= as above
As
- en rises as the square root of the R16 resistor value
- A rises linearly whith the rise of the R16 resistor value
hence the Vn noise decreases as R16 increases.
In our case we have that the ratio of the noise of a 47 KOhm "real" resistor and the "synthesized" one is SQRT(16)= 4 ie the noise is 4 times lower.
Is the math correct?
Any further clues?
Thanks
There is one of them I'm not shure I have fully understood the way it works.
I'd like if Mr Self could confim (or correct) my guess
I'm not sure I can publish the schematic on this forum so I'll refer to the schematic published on Elektor.
The circuit I refer to is named "load synthesizer"
Self writes:
"A load-synthesis circuit around IC4 is used to make an electronic version of the required 47 KOhm resistor from the 1 MOhm resistor R16. The Johnson noise of this resistor is however not emulated"
with no further details of the circuit.
An analysis by inspection of the circuits tells:
RL= R16 / (1 - A)
where
RL= is the "synthesized" resistor
A= is the gain of IC4A (the gain is inverting so A=~ -16)
Vn= en / A (on the first order and neglecting OP-Amp voltage and current noise)
where
Vn= is the noise of the synthesized resistor
en= is the Johnson noise of R16
A= as above
As
- en rises as the square root of the R16 resistor value
- A rises linearly whith the rise of the R16 resistor value
hence the Vn noise decreases as R16 increases.
In our case we have that the ratio of the noise of a 47 KOhm "real" resistor and the "synthesized" one is SQRT(16)= 4 ie the noise is 4 times lower.
Is the math correct?
Any further clues?
Thanks
The above formula con be simplified ad follows:
Vn is "proportional" to en / sqrt (R16)
Last edited:
New Doug Self Tone Control
As promised earlier in this thread Jan Didden has published Doug's "A low-noise preamplifier with variable-frequency tone control" in his Linear Audio magazine. My copy will soon be in my mail box!
See details here: Linear Audio magazine
As promised earlier in this thread Jan Didden has published Doug's "A low-noise preamplifier with variable-frequency tone control" in his Linear Audio magazine. My copy will soon be in my mail box!
See details here: Linear Audio magazine
"A low-noise preamplifier with variable-frequency tone control"
Cool, I guess I have to make an order for this one & the vinyl-trak phono design as well.
Actaually, i would like to automate the tone control, using something like a dac, vca, epot, any suggestions? I wonder how this would impact performance. The way I see it, a tone control is a necessary requirement for a pre-amp design. But it is only needed to make up for some deficiency, usually in the source material. In that, the source is probably compromised in some way, so adding a little distortion (assuming that the tone control has some) is of little consiquence.
Using a tone control such as this, are normal HP & LP filters also required in a pre-amp design? Maybe only a rumble filter is necessary!
Cool, I guess I have to make an order for this one & the vinyl-trak phono design as well.
Actaually, i would like to automate the tone control, using something like a dac, vca, epot, any suggestions? I wonder how this would impact performance. The way I see it, a tone control is a necessary requirement for a pre-amp design. But it is only needed to make up for some deficiency, usually in the source material. In that, the source is probably compromised in some way, so adding a little distortion (assuming that the tone control has some) is of little consiquence.
Using a tone control such as this, are normal HP & LP filters also required in a pre-amp design? Maybe only a rumble filter is necessary!
Is it something like Massenburg's parametric equalizer? If it is than I must admit that it is much more attractive proposition than ordinary tone controls with fixed turnover frequencies.
JFET op amps are generally more RF-resistant in the critical input stage. That may be why some prefer to use them (even though they typically do not have as low noise and distortion as the best bipolars).
Cheers,
Bob
I found this disappointing, because he skated over some critical areas in what is fundamentally a sales pitch for ESS. This made me distrust the other section of his presentation which was about perceived differences in SD DAC performances ESS claim to have documented in blind testing.
The 2 principal things that he neglected to mention are:-
1) that digital controlled IC analog volume controls probably can meet the performance of any discrete or mechanical attenuator, a point he was forced to concede when pressed at the end of the volume control presentation, and
2) that low-jitter clocks do NOT require to be locked to any 'transport clock' since it has become economic to store data purely in solid-state memory.
This is not that different from what I observe amongst the majority of participants here, talented people perverting their skills to provide less and less plausible justifications for performance differences which are not demonstrated to be audible.
You all proceed as though the audibility of these features can be taken for granted. You hang around here in your cozy little blinkered group effectively congratulating each other on the profundity of your expertise because you know that the likelihood of your being taken to task for it is small, but to all of us who retain our critical faculties with respect to the evaluation of evidence it's still just a load of redneck noise.
Thanks Bob. Seems plausible, though having tried the OPA2134 and the OPA2604 in the same design (a simple series feedback line amp), I thought the 2604 sounded very good, while the 2134 was very disappointing. Only my opinion of course
. From the datasheets there appears to be some significant differences in their topologies, so I wonder if there isn't more to it than just the extra RF resistance due to fet inputs.I am tempted to call this one of the great truths of audio electronics. From now on I shall call it Finch law after the first man who formulated it in such clear way.
I'm not sure about any credit there, ivanlukic. I obviously trod on toes with the comment but as you suggest, a lot of people would have arrived at a similar conclusion long before me.
I've been sucked back into this thread recently and have spent some time re-reading the posts and I'm only up to page 49! Beyond the basics I'm not knowledgeable at all about the electrons and how they travel and what affects them. I just know good sound when I hear it, without questioning how it's produced, or the minutia of detail that ultimately shapes it.
I will admit, that I always prefer to put the science first - as a starting point to understand what should affect the sound - or not. Since we can never re-create the original recording in terms of time and space, a mentor once told me more years ago than I care to remember, all we have is the ability to try and recreate the 'illusion' of the recording.
So having said that, I've never understood how any level of distortion - whether introduced into the circuit by design or not - is a good thing or helps to more accurately re-create that illusion. Perhaps that's just my naivete about how noise make things sound better. If one wants to introduce noise into their personal listening experience all you need to do is listen to any recently recorded Pop CD! I've found that some of my older CDs from the 80's sound soooooo much better than the current crop of available CDs. I know that is a broad-brush statement and I'm sure there are some 'good' recordings out there, but I've found it almost unbearable to listen to the last 2 CD's I've purchased. And I don't believe it's all because of my golden ear of yester-year is loosing some of its luster. But I will admit some of it is.
I've built the new DS pre-amp and what I can say is it has revealed the short-comings of my up-stream equipment. This project was probably over the top for my needs, but I looked at it more from the standpoint of a DIY challenge and to get my feet wet with one of Doug's designs. Needless to say, I picked a comprehensive project! All I can say is it far exceeds my expectations and provides me with wanting to upgrade the rest of my gear down the road to match it's level of performance.
That's my latest ramblings on this topic.
Beyond the basics I'm not knowledgeable at all about the electrons and how they travel and what affects them. I just know good sound when I hear it, without questioning how it's produced, or the minutia of detail that ultimately shapes it.I will admit, that I always prefer to put the science first - as a starting point to understand what should affect the sound - or not. Since we can never re-create the original recording in terms of time and space, a mentor once told me more years ago than I care to remember, all we have is the ability to try and recreate the 'illusion' of the recording.
So having said that, I've never understood how any level of distortion - whether introduced into the circuit by design or not - is a good thing or helps to more accurately re-create that illusion.
Perhaps that's just my naivete about how noise make things sound better. If one wants to introduce noise into their personal listening experience all you need to do is listen to any recently recorded Pop CD! I've found that some of my older CDs from the 80's sound soooooo much better than the current crop of available CDs. I know that is a broad-brush statement and I'm sure there are some 'good' recordings out there, but I've found it almost unbearable to listen to the last 2 CD's I've purchased. And I don't believe it's all because of my golden ear of yester-year is loosing some of its luster. But I will admit some of it is.I've built the new DS pre-amp and what I can say is it has revealed the short-comings of my up-stream equipment. This project was probably over the top for my needs, but I looked at it more from the standpoint of a DIY challenge and to get my feet wet with one of Doug's designs. Needless to say, I picked a comprehensive project! All I can say is it far exceeds my expectations and provides me with wanting to upgrade the rest of my gear down the road to match it's level of performance.
That's my latest ramblings on this topic.