|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
173.247.0.211
Some reading indicates that today's op amps are as good as discrete components for phono stages.
I find that hard to believe, what's the concensus?
Follow Ups:
Depends on whos's doing the design. There are fantastic opamps which can yield excellent results in the right hands or the same part can be part of crap.
Agreed - and when everything else is right, the design with the lower noise floor wins.-reub
Edits: 08/26/14 08/26/14
Noise is definitely a big part of it. I feel that minimizing distortion is probably equally important. Keeping CM distortion low seems to be a bit of a lost art though, not many designers going that extra mile these days.
Depends on your budget (as a designer) to ensure a commercially profitable design at your chosen price point - going discrete allows you to match and optimise each and every component to minimise errors but pushes the final price point up. If you are building your own, then go for broke and select the best components that you can afford.
The question is, if you were a commercial designer, how many people would appreciate the benefits of hand matched components at a lower target price point? Discerning users will already know what they are looking for.
In reality, the sound quality is actually more determined by the accuracy of the EQ circuitry and the quality of the signal path capacitors.
You also have to understand the purpose of the op-amp in context of the EQ topology. There is no reason why a well-designed op-amp based circuit shouldn't sound indistinguishable to a discrete stage. Regardless of discrete or otherwise, overload margin is probably the most important parameter next to EQ in designing a phono stage.
Commercial designs often use electrolytics in the signal path to remove DC offsets due to cheap opamps for convenience which is a significant source of colouration. I suspect that is why discrete versions at a particular price point may be preferred where the effect of gain on input offset voltage can be minimised. Personally I prefer designs to use a DC servo instead.
Distortions are dominated by the cartridge alignment parameters (tracking error, VTA, SRA etc etc) anyway.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
Op amps are designed to have very high gain so that the performance is dominated by the feedback resistors. This is best achieved using an integrated circuit. A 'discrete' op amp is, therefore, an oxymoron - it is a lot of effort for dubious benefit.
OTOH, a low-feedback or no-feedback circuit made of discrete components is a completely different beast. John Curl is mentioned elsewhere in this thread - his experience is much greater than mine and he has written that in any given situation using as little feedback as possible is best for sound. It is likely that equipment with discrete stages will have low/no feedback and will likely sound best. So, IMO, the question is really about high feedback vs. low feedback rather than integrated vs. discrete design.
OTOH again, a lot of audiophile equipment containing op amps gets good reviews. And, almost all the studio equipment that recorded our favorite albums contains opamps. And the opamp based gear will probably measure better than discretes!
Regards
13DoW
any gain stage is constructed using local feedback to reduce the open loop gain of the active component. The real issue is the source of the feedback and any potential phase shift between the input and the output, where the feedback is taken from. This can be a problem even within a single stage, if there is enough reactance between the input and output of the stage, At high frequencies one even has to consider the Miller capacitance of the active device to avoid oscillation and compensation has to be made in the design of the stage, but we are talking GHz designs where that is a real issue. No-Feedback is a marketing term that has nothing to do with reality :). When manufacturers talk about a zero feed back design what they really mean is ZERO GLOBAL FEEDBACK. One of the liveliest sounding preamp i had is a Sao Win design...it has 2 ic stages in it...He sanded the identification off of one of them so to keep his sauce a secret...unfortunately it makes it impossible to repair it. So far he has not given me another set of the sanded off ICs to fix my preamp that developed a strange oscillation problem.The biggest challenge with a phono preamp is the power supply design, no matter what the active components are.
P.S. Here is an interesting read from Nelson Pass
dee
;-D
True terror is to wake up one morning and discover that your high school class is running the country.
quote by Kurt Vonnegut
Edits: 08/26/14 08/26/14 08/26/14
"any gain stage is constructed using local feedback to reduce the open loop gain of the active component. The real issue is the source of the feedback and any potential phase shift between the input and the output, where the feedback is taken from."
Bah ha ha --- for the purpose of analog circuitry we don't call drain resistance "feedback" ... for all intent and purpose for analog circuits that's pedantic bunk
Never trust an Atom, they Make Up everything!
An op-amp is designed to work with GLOBAL feedback from output to input, likely encompassing more than one gain stage.
Of course you can design a single-stage amplifier without applying any feedback. It will depend upon the largest input signal you want to handle and the characteristic of the amplifying device. Bipolar transistors have higher gain but start distorting at smaller input voltages compared to JFETs, which have lower gain. If the input signal is large enough to create more distortion than you want then degeneration (local feedback) can be added to effectively lower the gain and increase the signal handling. There will always be some parasitic feedback due to capacitive coupling between output and input and the transconductance of any amplifying device will be degenerated by the resistance of emitter/source/cathode connection. But these will be negligible except at very high frequencies and very high currents.
It is worth noting that adding feedback, either globally or locally, will increase noise. Any amplifying device has an inherent dynamic range - the difference between the largest signal it can handle before distorting and its noise floor. Feedback can never improve that range, only move it around - if you add feedback to improve signal handling the noise floor will raise up too.
Global feedback improves linearity by reducing the input voltage seen by the first amplifying device to very small levels.
The stability of feedback amplifiers (and control systems in general) is a subject that books could be written about, and indeed have. Many times over!
I was not trying to be a pain...but it gets under my skin when people talk about zero feedback amps, and not realize what an unstable and un-reproducible that design is :). Zero global feedback is a common practice in industry.
dee
;-D
True terror is to wake up one morning and discover that your high school class is running the country.
quote by Kurt Vonnegut
G'day all, this is an interesting thread! Doing quite a lot of DIY stuff myself (including moving magnet cartridge phono stages) for many years now, I've come to regard op amps very highly indeed, something that I probably would not have said ten or so years ago!Op amps can have 'issues', but as with all circuit designs, it takes a very good design and implementation to show op amps at their best. At the moment my phono stage of listening choice is a simple single op amp based full feedback (kit) design with a few unremarkable component replacements and it sounds very nice indeed with nothing to complain about at all.
It delivers around 40 db of low distortion and (very) low noise RIAA equalisation voltage gain and it sounds great, which is precisely what it is supposed to do. Regards, Felix.
Edits: 08/28/14
For me it was witnessing what a GSP ERA Gold V phono stage could do. That has to be more than 10 years.
My monoblocks use very high power opAmps, a Blue Circle design of about a decade ago or more (same devices used in their longstanding series 200 power amps).
Never trust an Atom, they Make Up everything!
just sayin', not makin' it up either.dee
;-D
True terror is to wake up one morning and discover that your high school class is running the country.
quote by Kurt Vonnegut
Edits: 08/28/14 08/30/14
The first gain stage of my phono preamp consists of a pair of J-FETs, each running open loop into a linear drain resistance.Where's the feedback?
Mark Kelly
Edits: 08/27/14
if so there it is :) Called a parallel feedback. If you have none, then there is no feedback, but that makes the FETs running somewhat out of the mostl linear range, ie you have more distortion....some FETS tolerate it more than others. and there is still the feedback internal to the the device, but that is negligible compared to the open loop gain. if you use them in a cascade config then the lower one provides a variable parallel feedback :)dee
;-D
True terror is to wake up one morning and discover that your high school class is running the country.
quote by Kurt Vonnegut
Edits: 08/26/14 08/26/14
The J-FETS are running in balanced mode, there's no AC voltage at the source.
Mark Kelly
in absolute terms the answer is no, because at one frequency or another the model will contain residual resistances between gate, source and drain, and there is the Miller capacitance which kicks in big time at high frequencies...in practical terms at a given frequency band, where the fringe effects can be neglected, you can get to practically zero feedback. To make it work you need to select components very carefully. That may not be a viable option for a production design, so designers use some amount of localized feedback to ease component selection. It is all a compromise, and meeting specs. If you use no feedback your THD maybe somewhere .01% and .1% and your IM might be between 1% and 0.1% depending on the particular sample of a component. which is ok for a DIY project you just keep swapping them till you find one that meets your particular need. Try that when you have to make 10 units a day :)...having a local feedback say 10 db will give you a 3.2 fold reduction in the numbers ...so you can meet specs with any of the components. so you can publish a spec that say you device will be with a THD of 0.03% and IM less then 0.3%.
dee
;-D
True terror is to wake up one morning and discover that your high school class is running the country.
quote by Kurt Vonnegut
Edits: 08/27/14 08/27/14 08/27/14
integrated circuits are superior in this application. You can't achieve the S/N and uncolored response with discreet components, as with integrated circuits in the hands of a talented designer. All the wire, board traces and distance between components make it impossible.
The JC3 isn't the only good phono to use them. How about Walker?
The AHT Non Signature was the only 5 star component at TAS for 2 or 3 years running. These might not be your favorite phono stages, but ICs are limited to budget pieces the same way transistors were 40 years ago.
neo
BIRD LIVES
Quality components tend to sound good, as do good circuit designs. I've heard some very good op-amp stuff and would not rule out owning/using them. I still prefer using tubes.
I am glad you asked this question I have wondered this myself. I just cant get behind the idea of OP amps in anything. However, there are suppose to be some pretty good Op amp based phono stages out there. Graham Slee comes to mind. I would never buy any though. It's either tubes or J-fets or the combination of the two for MM/MC stages for me....
That said, I still prefer discreet designs like Michael Yee's. His design for the Musical Surroundings Phonomena 2 proves good discreet designs don't have to be ultra expensive.
Some excellent info at his company's website (link below).
Opus 33 1/3
I do not think the form-factor of an op-amp is any limitation to sound quality. In fact, they could potentially be outstanding because of low noise and short signal paths. But Design-wise, I doubt that any general-purpose op-amps are optimized for great sound (examples per the Michael Yee link).
IMHO the requirements for a business case for high volume production is the greatest limitation to op-amp quality for phono stages. Maybe this will change. Look what happened to Class D amps with ICE modules. Perhaps sooner than later we'll see really high-end designs with optimized components for phono stages?
"Knowing what you don't know is, in a sense, omniscience"
"I do not think the form-factor of an op-amp is any limitation to sound quality."
Actually there are issues. Big deal, there are issues with everything especially tubes.
But back to op-amps three pop to mind...
-Flexure sensitivity: The die could be used as a phono cartridge, stress alters offset voltages and generates internal voltage shifts.
-Thermal feedback: Clever die layout balances the thermal offset but not the loop gain variations due to carrier mobility changes.
-Noise/GBW tradeoff: Bigger geometries have less noise as it tends to average out, smaller geometries have wider bandwidth because they have fewer "strays" but individual carriers become a bigger part of the energy.
Perfection is a limit, not a possibility so you do the best you can with what you got. Good Engineers get good results. But that doesn't mean that tomorrow they won't do better or that things are bound to improve. That's just life...
One area where I do expect to see improvements is psychometrics. If we could really sort out how we hear and process music and our sensitivities to noise and distortion then the divergence between metrics and experience should substantially reduce making specifications a lot more useful. Actually without any additional knowledge components could easily be more fully specified which would really help audiophiles, but the former would do more for everyone.
Regards, Rick
We do try to fit solutions into existing paradigms. I like to think about alternative pickup mechanisms which do not involve moving a magnet through a coil. An optical cartridge with embedded IC comes to mind.
Interestingly, I think some manufacturers already have strong proprietary measurement tools that are highly tuned to assessing perceptual and experiential components of music experience via playback equipment. I've spent a part of my career doing perceptual testing and sensory testing, so I've seen some pieces of the process. These measuring tools are trade secrets. They are not widely shared with competitors (with the possible exception of the Canadian speaker consortium), much less the general public. But a great deal of perceptual testing has occurred and does impact on design and engineering goals within these organizations.
Unfortunately, since they are undisclosed, these measurements are not of use to us poor consumers trying to make decisions based on specifications.
"Knowing what you don't know is, in a sense, omniscience"
Regardless of the transducer principle, the 45/45 groove and the cutting procedure will dominate the final result.
I have an ELP (Laser Turntable) and the output is still fed to a phono preamp. This proved to my satisfaction that many of the distortions and blemishes that can be heard via conventional LP playback are often cut or embedded in the groove and not due to any stylus/cartridge design issues.
I guess the thread has deviated from the original point, but in my view the discussions around the phono stage ignore the more fundamental issues regarding the interaction of the transducers to the pre-amp electronics and other more dominant sources of distortion.
Regards Anthony
"Beauty is Truth, Truth Beauty.." Keats
"I like to think about alternative pickup mechanisms which do not involve moving a magnet through a coil. An optical cartridge with embedded IC comes to mind."
Well, you still would have a huge list of problems at the mastering and pressing end. Ultimately the solution to vinyl's problems is digital. No matter what you think of it's current state and implementation, the answer is still the same. It has it's own problems but theoretically they are all solvable to any desired limit while the mechanical ones are constrained by the structure and durability of the materials.
As far as the cartridge goes I recall that there already were capacitive or optical ones very long ago. I'm happy with my MM cartridge but think that a lot of it has to do with the blending of the various resonances in your turntable.
Regards, Rick
Good one--yes Michaels Design is excellent
Des
I'm with you plus a little, it's more than hard to believe.
ET
You must be kidding! :)
There are some good opamp based designs available today such as the John Curl designed Parasound JC-1 and Jeff Rowland's stuff.
As for me, I prefer tubes and discrete JFETs.
Johns Vendetta -which I own has them
However, the OP's used in the Vendetta are SERVO (or DC offset control) only, and work exclusively below 1HZ --The SCP-2T is both Direct Coupled and Complementary JFET's.
Des
I don't doubt, for instance, that using opamps is cheaper, and I suppose, if you consider that there are opamps and opamps it might be possible, but it appears to me the only reason for using opamps is to save cost and time.
But hey, I did not stay at a Holiday Inn last night.
Obviously, my post above is wrong-headed.
Maybe I shoulda tried the Holiday Inn.
I think E-Stat's point might be summarised by asking how one can come to such a conclusion, and on what evidence? E-Stat might also be implying that you would need a body of evidence on which to base such a conclusion - for the countering of which, E-Stat provided some well-known examples to the contrary.
But you probably know this.
big j.
"... only a very few individuals understand as yet that personal salvation is a contradiction in terms."
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: