The 709C was not an audio-specific component, so I'm not surprised a discrete design performs better. The NE5532 and other op amp chips specifically intended for audio applications are whole different category and, when designed in properly, are very difficult to beat with a discrete equivalent because it's hard -- often impossible -- to source discrete parts that have the sort of precision that's easily achievable by a decent silicon foundry. IOW, as Paul himself found out with his vintage Studer mixing desk, "op amp rolling" is a waste of time and money in an expertly engineered piece of audio gear that uses carefully selected, audio-specific ICs for the op-amp function.
Audiophile op-amps exist that beat many discrete circuits nowadays. Also, analog studios are using lots of op-amps. Your speaker setup is where things sound different when the electronics is properly engineered discrete or not.
With regard to speakers sometimes you want a different sound and you can't buy a new set of speaker but changing the op-amps can give you a different sound
@@daysofgrace2934 If I want a different sound, I would not bother to change op-amps as any difference between (good) op-amps will do near no audible difference compared to anything you can do with room acoustics incl. speaker placement. A typical audiophile speaker distort many many times more than a typical audiophile op-amp.
I've found that most problems associated with op amp circuits can be booked down to two major issues: 1. Budget-price op amps with low bandwidth and 2. Designer "cheaps out" by having too much gain in each stage. This result in high levels of slew-induced distortion (sounds absolutely awful). In extreme cases poor high frequency response and phase shifting as well. I have a guitar amp with both of those problems. Rather than replacing the op amps and adding 2 more gain stages, I replaced the entire preamp section with a two stage (zero feedback) circuit consisting of one 12AX7 tube. Sounds wonderful.
I guess why all jfet distortion sound better because you force to use stages while cheap amp just have one op amp stage for distortion for guitar amps.
Yes, the answer is "it depends". The 709 op-amp was among the first of such, and of course had many warts. The best of today's devices have few warts and take advantage of all manner of characteristics not possible with discretes, such as device matching. And keep in mind that recording media, such as mixer boards, run audio signals through sometimes a dozen or more op-amp circuits before they're plopped onto the recording media. The A/D converter also contains an amplifier chain to drive its digital circuits. So in my view, ignoring the distortions caused by the recording chain and focusing only on reproducing devices is naive..
As a teenager over 30 years ago I used some already ancient 741 op amps to build a couple of preamps for magnetic phono cartridges. If you ask me, they sounded great! Low noise and distortion, great equalization. Wish I could remember where I got the circuit diagram, some dusty/musty old electronics hobbyist magazine.
Hi, Paul, I think, we both are about the same age and for sure from that generation which started to use OPAmps the first time in the electronic new age. I remember well, the now vintage 709 (with external capacitor for frequency compensation), the the 741 (without it). I always considered, and still do, an OPAmp being a device, like a transistor or a tube is one, with which you design a circuit for a special purpose. Be it a unity gain buffer, a filter, a precision gain amplifier, an impedance matcher, you name it. What made and make this device unique is its insanely (this is how we felt about it) open loop gain, its extremely high input impedance, the ease of how you could precisely determine the gain you wanted or the input impedance for the same matter. And with mathematical precision a filter characteristic. Now I come full circle: it is a device, not a circuit (I am not looking inside the housing now) by itself. But now back to the question about sound: you will have to design the sound characteristics yourself, the device is not having one! That said: of course it does. But this wonderful device gives all the responsibility back to where it belongs: the designer. Cheers! Keep going strong. Hans.
Paul, you look a lot more relaxed and less stressed now! I think this Cov19 thingo worked well for you! Perhaps a 'one day per week' at the office would be in order? ..and Chris is doing a fine job in your absence, too! Thanks for keep these daily videos coming and stay safe out there!
I agree! The area where he lives is absolutely Gorgeous, but if I'm not mistaken the lot sizes in that area are run-of-the-mill 8-10k sq ft suburbia plots. We may have already seen most of what there is to see. Boulder doesn't have space for properties with a back 40! Nowhere close to town anyway
I would say that an op amp by definition has very high open loop gain, and is thus unusable without feedback. You can make a discrete amp with low or zero feedback, but it wouldn't contain an op amp.
A couple of years ago I bought an op amp from TI to improve sound quality from my hybrid phones amp. Already I had upgraded the Chinese valves with RCA and Russian valves. After swapping the op amps...the sound was much worse,muddy, less detail. Again I change op amp with some NE ....yeahhhhh.great improvement!
One of the problems with "upgrading" op amps is that newer chips may be much faster. They are more prone to HF oscillation. You have to look at the circuit with a scope to make sure it's behaving. You may have to partially re-design the circuit to optimize it for a new chip.
I could be wrong but I thought that one problem with chip op-amps is that they are more susceptible to internal capacitive losses / cross-talk as well.
When I purchased my Audio Technica moving coil phono cartridge back in the 1980s I knocked up a simple IC op amp based 20 dB gain circuit on a scrap of Veroboard. I still have it somewhere. It performed really well. No noticeable crosstalk. And it was as cheap as chips as they say. Certainly a lot less than a transformer.
In the 60's Electronics World had a construction article on building a tube based opamp. It looked interesting so I drew up a parts list and set out to see how much it would cost to build. This design had 4 or 5 dual triodes, power supply and all the discrete resistors and capacitors. By the time I was done it was goping to cost a hundred dollars with out any consideration for a chassis so that project was moved from the active to inactive list. Several years later after getting out of the army I worked for Analog Devices as an engineering tech and got to work on some cutting edge discreet designs that had very good characteristics. I built a phono stage with a pair that sounded very good for it's time but that would have been pretty expensive if you didn't have access to engineering samples.
As long as you dont overdrive them they should be pretty clean and very linear at audio frequencies nowadays. The old stuff is a far cry compared to what we have now, so its not really valid to use the 709C as an example. Modern op-amps are far better and made for precision amplification, and while not all are equal, there are plenty of low noise op-amps that will do just as good given the large amount of feedback available to clean up any remaining nonlinearities at the output. Fair enough to say that most aren't since many are low voltage now to cater to low voltage electronics, but there are plenty just for audio these days. Also since all the silicon is on one die there will be less issues with thermal effects since it will all track, and everything is matched. To say they "don't sound as good" kind of ignores the fact that a lot of the same topologies are found on an IC as you would find in a discrete version. In reality, there probably would be almost no discernable difference at audio frequencies, and with audio almost always the speakers/headphones will be the dominant source of distortion for sound reproduction. In an audio amplifier it makes sense since you need the voltage swing and the final stage has to supply amps to 10's of amps and not milliamps, but with that said there are audio power op-amps with THD in the parts per thousand for power outputs < 65W. Edit: Im assuming the "two device op-amp" was probably a pair of Darlington transistors since there was feedback being used.
Hi there Paul, I have been watching your video's recently, thank you. The reason I am posting a comment here today is because I believe we have met one in other back in the 80's, while I working in an Audio Shop located in Phoenix Arizona, owned by a guy named Buzz Jenson and his Audio Alternative store I was just curious. Back to audio and OpAmps. I worked in the semipro and HiEnd audio sales field for many years, and remember in the late 70's early 80's you could definitely hear the difference between the brand names of the day, most of the on a chip designed receivers, amplifiers had different musicality qualities directly associated with the design choices made, at that time I would always recommend to our customers that discrete was the way to go, these days maybe not so much. Anyways always a pleasure to check out your video's.
Lloyd Stout So is Nelson Pass and he has his own DIY site. Many designs over the years from Nelson for public use. Don't you think it's up to Paul if he would like to do that? I respect Paul and whatever he wants to do I would be behind him.
I loved the reference to the Quintessence Preamp, I had one before switching to a dB Systems preamp. The Quintessence was very elegant looking and felt like a quality product. All of these came after using tube Dynaco stuff and eventually went back to tubes (ARC) by 1976 only to go with First Watt stuff in 2011 and TBH I haven’t looked back yet.
its like tube rolling, all opamps chips has an unique flavor, some sounnd super crhistal clear but not musical, some others sound musical but no extra details, its up to your ears to chose what is "better"
Why does the audiophile community hate feedback? Do you like to make your job harder!? Theres nothing inherently wrong with feedback, and if the circuit is already close to linear, you will drive the final distortion products down below the noise floor if there is plenty of open loop gain available. A lot of gain is needed since the error correction is strictly proportional, so some portion of error will inevitably remain, and less feedback means more residual error at the output. More feedback, as topology and stability permits, will mean less error and better linearity. Almost every commercial amplifier uses quite a lot of feedback to avoid non-linearity and distortion, esp. in the cross-over region. Not using it almost certainly means more distortion unless you're using exceptional devices or a complex topology. Even then it would be _even better_ with feedback. Why leave that improvement on the table? Edit: Using the OPA134 as an example, it has a listed THD of 0.00008%. Thats pretty goddamn low! Granted it will inevitably climb a bit above 10kHz, but so will your discrete one. Maybe not as fast with less feedback, but thats offset by other issues that come up when less feedback is used.
I don't know... the LME49720 sounds pretty darn good. If a system is designed with discrete components, then there is no need for "discrete op-amps" because those are just functional replacements for op-amps. The topology should be different enough that when designing a discrete component system with minimal feedback, op-amps of any kind wouldn't be incorporated into the design. So "discrete op-amp" are really more for the OCD type who bought op-amp based systems and can't sleep well at night.
There is nothing wrong with feedback, if the open-loop gain-bandwidth product and slew rate are high enough. The problem only comes in, when there are multiple response poles too close to the audio range, requiring heavy compensation for stability, which then pulls the slew rate down. We need to get away from the myth that high feedback is uniformly bad. It becomes less and less true as time goes by and devices become faster.
I had a Quintessence preamp and Ugh those 709 op-amps were bad. I remember it as being kinda dark and vague sounding. I was going to build a better circuit in it and once used it as a test bed for op-amp but eventually took it apart and kept the faceplate and knobs but threw the rest out. I also had the SWTP preamp which had plug in discreet op-amp boards but I never got that to sound great so I kept the boards but threw the rest out. Then I bought a Rapport Pre-1 preamp and it had some pretty serious stability issues. I cleaned up the grounding and replaced the pot and it sounded pretty good. Then one day lightning hit outside and blew the output stages while I was listening to Pink Floyd (loud hum both channels). I found that the high level stages were sealed in epoxy. I broke one open and was surprised to find an op-amp inside. I was of the opinion then that discreet was the way to go. I built replacement circuits for the high level stages using the same op-amp. This has been my reference for solid state preamps since then. Sounding better than Mark Levinson and Classe and pretty much every high-end preamp I compared it to.
A good chip op-amp can sound great. The design of the rest of the piece of equipment matters enormously too. The power supply has to be high quality, the layout of the printed circuit board in terms of component placement and ground plane design has to be good and input and output circuits have to do things like impedance matching with good results in terms of connecting to other components. Is the chassis well shielded, etc? It's the complete package that counts.
That's true as Paul says. You can have an IC. oppamp which has incredibly good sound quality and very low noise. If you have two preamps from 2 brands that cost the same. IC. op amp can quickly have better sound quality than the sepparate transistors preamplifier has. But it can quickly be the opposite that separate transistors have better sound. It is too easy to say that separate transistors are always better. IC. op amp does not always have to be worse than separate teanistors. IC. up amp came the first time approx. in the late 1963
Sorry to disagree. It sounds like Paul is a few years behind (or maybe a few decades?). I totally agree that MOST op amps are not great at audio, but there are many that are designed specifically for audio and some, even better, for instrumentation. An integrated op amp has the advantage of size. This means lightning fact response (if it is well designed and has low internal compensation) which allows for very high open-loop gain thus heavy feedback. All distortion is eliminated (or at least negligible) except that of the input differential pair. All internal noise is shunted except that of the differential pair. The differential pair has a built-in advantage over discretes : it is paired by construction. Man parameters in semiconductors can vary wildly from one unit to another. For instance the Hfe (current gain) of transistors of the same make and model can vary from 500 to 10,000 due to small variations in the masks in the silicon. So, if you make a discrete op amp, you have to buy many and pair then individually. In an integrated op amp, the variations are the same for the two input transistors, thus greatly reducing parameter variations to a maximum of about ±10%. As an instrument and audio designer, I did use high quality op-amps (not something like the 324!!!) and preceded then with a discrete pair of carefully selected matched very low noise large bandwidth transistors (either bipolar or FET), whose collector or drain was connected to a low-capacitance current source. The outputs then went into a large bandwidth integrated op-amp. These transistors were placed as close as possible (1 or 2 mm) to the op-amp input. The results were outstanding since the input transistors were submitted to no base-to-emitter or gate-to-drain change, thus retaining fixed parameters even through changing signal conditions, providing for a wide distortion- and noise-free dynamic range. Frequency response was, of course, flat from DC (0Hz) to at least 50Kz.
Anybody else here get a Fosi v3 monoblock amp and received 2 Muses02 chips? I bought 2 amps so they sent 2 chips, does anyone know if the two Muses02 chips are both supposed to be installed into one amp? Or is one chip for each amp? The reason I'm asking is because when you open the amp, there's 2 chips that look like they can be replaced by the Muses02. I know this is probably stupid, but this is my first time using an opamp
I misunderstood the subject before clicking on the video, I was expecting a perspective on IC amplifiers that you would find in a car stereo, bookshelf stereo, or a Bluetooth speaker. There are even some popular car amplifiers that use IC amps. Any opinions on those? I realize that this type of equipment qualifies more as "consumer grade" than "hifi"
Hmm maybe last century this was correct. This century I've not seen a single equivalent item where discreet beats integrated. It's easy to see why when it's thought about too. Longer signal path, no thermal coupling, mismatched components not the result of billions of dollers of r+d.
Vertical power devices, at >>100 V. In particular the SiC superjunction NPN's ("GeneSiC", only really still made for high-temp applications because the chips themselves easily work at temperatures kitchen ovens are not willing to go to (350-500 C is usable depending on packaging), as the need for DC/bias base current had them loose against the JFETs) and the SiC JFETs from e.g. UnitedSiC (now part of Qorvo) and maybe soon others again (last competitor ceased about 10 years ago). Those are sadly only good to a few MHz (and that's with frequency-dependent distortion there that needs compensating!) due to their gate resistance, but they are extremely robust and very good (power) amplifiers. (The JFETs are designed for switching applications like power supplies, but they are safe to use in linear mode and the only real vulnerable part is excessive sustained forward current through the gate-source diode. Like, around 10+% of nominal TDP used for that diode. Everything else is just a "eventually it overheats and breaks" issue, afaik. I postponed my plans to get some proper measurements of the RF behavior, mostly due to some difficulties brought on by the high voltages (1700~2200 V avalanche, roughly, depending on if you get a 1200V or 1700V nominal one.).)
Depends on the ópamp or discrete transistor using. I like discreet JFET's . Very Tube Like sounding . Well one in particular I found just amazing quiet with high output. And somehow they were all exactly the same. If you know anything about JFET's that never happens. You have to bias them individually. And I'm not telling you the model number or anything there mine, all mine. I will tell you they're very similar to the ones that Marantz were using. To make their amplifiers sound more tube like.
You just need to try them to see.Personally i found amazing op amps which cost 2-3 pounds.The sound is just mind blowing.The bass and the richness of the sound is amazing.AD827SQ / 883B that's the op amp.
And to find an engineer that can make that better op amp, they undoubtedly work at a large chip manufacturer with Deep Pockets not at small audio firms. Anyways thanks for the history lesson. Today's op amps are not your grandfather's op amps
In my experience it depends a lot on the rest of the system. I have a normal highend consumer stereo and to make that sound at all special it requires (slightly DAC dependent) the best discrete opamps I have in my collection in order for the system to have decent dynamics and overall sound. I also have a home built/designed system that uses pro drivers and pro amps. With the same source into the DAC I use regular chip opamp (really good one...) in order to keep the dynamics in the realm of sanity. So, discrete opamps are way better for home hifi because home hifi needs all the help it can get, but with pro gear you have so much power and dynamics that you are just fine with regular top shelf opamps. No opamps were hurt in these experiments and the refereed to opamps in this story are Sparkos (discrete, from Colorado, about 10 miles from PS Audio) and OPA 2134 and Muses 02 chip opamps. Bryston, Muses 01, Muses 03 are all fine units, too)
Yesterday in my RIAA preamp I compared a pair OPA2111KP what it come with originally. To a high end specifically made for audio MUSES 01. The MUSES 01 was superior in many ways so no contest. Today I got time to compare a descret op-amp against the good MUSES 01. It is hard to say something definitely about witch sounded better than the other they were so good both of them. But the descret op-amp played lower down in frequency that the MUSES 01 didn't. So it won. So this "random" pair of descret op-amp sounded better in the end, than a high price high end IC op-amp made for audio. I mean "random" above because it is only the first and only that I tried and there is many other variants. (Name: moing hd-08) And I paid a lot for the IC op-amp in the same ballpark as the descret ones at $100 for the pair. One reason were that I realized that there is a lot of copies and to be sure that I got original op-amps I decided to get them from the manufacturer. So I know they are genuine. So if I never had heard the descret op-amp then I would also think that IC op-amp is as good as it gets. Now I know differently. And that kind of money for "simple things" as op-amps I would need to multiple many times that amount, to get that type of sonic improvement by trying to upgrade my cartridge that anyway would be fruitless when there is a hi end IC op-amp in the signal path that do not amplifying the lowest frequencies.. .. everything matters. So it would not help if I threw infinite amount of money on a cartridge.
What do you think of the THAT Op Amps that are specifically designed for audio use such as the THAT 1512 or THAT 1580? Are they leveling the battlefield so to say?
The THAT 1512 and 1580 (and similar earlier products, like the INA103) are specifically microphone preamplifiers, not general op amps. (Although I suppose people could use them as a gain make-up stage after a passive summing mixer). For that purpose, they have excellent specs.
Today, unless you are designing to special requirements such as optimal Noise Figure for a given source, high voltage or high power, it's usually not worth a designer's time to develop a discrete-component op amp. www.ti.com/amplifier-circuit/op-amps/audio/overview.html www.cycfi.com/projects/six-pack/op-amp-shootout/ (2017) tangentsoft.net/audio/opamps.html (last update 2016) Of course, incorporating a suitable op amp into a product and achieving its best performance still requires an experienced and knowledgeable designer. Using an IC doesn't guarantee freedom from all pitfalls. Too bad Paul didn't mention the story of how a competitor designed a preamp with the 709 that sounded much better than his - and why.
the discrete one will cost more - assuming other aspects are on a level pegging. might be an interesting 'shoot out' - a discrete at the same price point as a chip, though the chip with more sophisticated psu...(if such a 2 competitors van be found...)
Burson opamps are great value for money, especially V5i, that is not too big. I am using them in my Asus Xonar STXll, much better sound now than with the originals.
@@bergennorway No soldering needed. I know I'll be slagged for this by geniuses like Lloyd Stout, but here it goes, and for anyone who thinks this is shit, stop on by and I'll a/b this for you... I took out the not so great power supply (the computer's) and replaced it with a Seasonic PRIME 550 Platinum SSR-550PD2 550W 80+ (it was as far up the food chain as I was willing to afford ) and connected that to the wall with a Pangea AC 14SE MKII, (which costs somewhat less a few years ago), and lastly I put in an odd power filtering board that is designed to filter the power going to add in cards through many capacitors. The before and after difference of the whole package was dramatic. I had no interest in figuring out powering the card with an external power supply. While all that worked and is good, to be honest I have matched the output quality with a few other -not expensive- DAC options (which at the time I was not convinced, say, a Topping DAC for $100 or $130 or $200 etc would be any good, I was somewhat wrong about that, and with many of the Topping DACs you can swap opamps.) So what would I do now? Well, an STXII costs 220$, the power supply 140$, power cable, 90$, bus power filter $45, opamps, 200$= $700, my external DAC with all the extras is about $1000, and a Topping D30 (swappable buffer opamp) is about 130$, and a D50 (no replaceable opamps, unless you soldier) is 500$ plus, if you want to be equal, 90$ for a power cord and a USB cable and maybe a sibling power filter for the USB output from a computer -40$ so, all up, less money, all have good DAC chips... So if was the broke human I am and wanted a low cost DAC with a swappable opamp, which is a feature I rather like, I would look at a Topping D30... cheers
The problem with transistors, is that there are a lot of fake ones. sound-au.com/counterfeit.htm The problem with tubes, is that there are a lot of fake ones. jacmusic.com/nos/images/Fakes/index.htm The problem with resistors, is that there are a lot of fake ones. www.tsec.ltd.uk/resistor-obsolescence-and-how-to-spot-counterfeit-resistors/ The problem with capacitors, is that there are a lot of fake ones. www.dfrsolutions.com/counterfeit-capacitor That's life in this world overrun by criminals and frauds. Most of the phone calls I receive these days, are from criminals trying to scam me. But I have fun with them.
In my humble opinion, you have a better chance of improving the sound of your device by putting in a better chip than by changing the speaker cable from $10 to $100 per meter.
Pretty much every readily available op-amp IC out there is designed without any precision resistors or capacitors. Integrating these is not easy! (Anyone know of any exceptions? I could be wrong). Even with the best designer on the job, that's a severe hindrance.
Generally you don't need precise absolute values for R's and C's on the chip. The parts which set the overall gain, pole and zero locations, etc., are external to the IC. But on-chip resistors can be matched to precise ratios, as can transistor offsets and current-mirror ratios. In discrete designs, those things need to be done by hand, which greatly inflates labor cost. I should also point out that some of the best low-noise preamp designs today, are hybrids of discrete transistors and IC op amps.
I frequently use the LM833 It's VERY VERY good sounding chip opamp. It's made for audio and the distortion bandwidth is below 0.005 (datasheet says). We don't need morte than that if we are diying a home audio! Fits most uses, worth the test!
It really depends on what you want to call discrete... A chip with one single opamp on it is often also called a "discrete" component. Because it's one spare opamp instead of multiple that for example makes AND/OR ports. These discrete AND/OR ports can be grouped into things like J/K flipflops or D latches. These spare D latches can be grouped into a clock divider, which are often used as a counter.
@Dave Micolichek Yes exactly, "individual components". But when is something an individual component or built up from subcomponents? You can divide components up into subcomponents until you're on molecular level... The term "discrete" is just as undefined as "much", "large" or "good" and can only be defined with the necessary context required to do so.
@Dave Micolichek I know, the molecular example was a bit weak, since it has no explicit relevance to electrical properties of the complete component in most cases. However, for example a traditional PNP/NPN transistor is made of at least three components for the junction, some metal leads that extrude from the package and some internal wiring that connects these to the junction. For example a standard electrolytic capacitor is made of metal film, a dielectric film and some type of electrolyte liquid. Two metal wires are attached to the anode and cathode foils and protrude through the package. The amount of components in both packages of those examples are significantly less, that's about the only significant difference.
@Dave Micolichek One could also argue the other way around for components like transistors or diodes, since it's just one die cut from a wafer after all. Just like an opamp or microprocessor.
Context. It really depends on the strength of he rest of the system. I have discrete opamps that blow everything away, but they are no the best choice in all situations, I have a situation where IC opamps are the best sounding choice, and in another system of mine discrete is the only way to top sound.
Wierd subject. I do not know of any ADC or DAC that is high speed and does not either contain or use an op amp. Makes about as much sense as an amp that does not use electricity. Single stage mic amps in condenser microphones often do contain a small FET based class A amplifier to transition from the very high impedance microphone element to drive the low Z balanced output, but from there almost every EQ, Pre Amp, Mixer, DSP, etc are universally Op Amp designs. They became popular for low noise, high speed, very controlled gain, high linearity, low phase shift, etc. All instrumentation amplifiers are based on them.
In discreet vs integrated, do not overlook the actual advantages of the small package. Low noise, very high slew rates, very high CMMR, very high bandwidth, etc are commonly used for precision instrumentation applications. Compare any discreet op amp performance with for example the performance of a MAX4146 for example. The single die is thermally stable, low noise, slew rate of 1,000 V/uS compared to the obsolete part Paul mentioned with a slew rate of 0.25 V/uS. A bandwidth of 130 MHZ, is of a 3 op amp design for instruments with a CMMR of 70db at 10 MHZ. Flat response to 90 MHZ with 0.1 dB gain. Very low noise of 3.45 nV. Sorry Paul. Try some modern very precise parts. Discreet parts have problems due to the physical distances, inductance, capacitance, etc that are greatly reduced with the small dimensions inside a small SMT device. Cheaper and better is the advantages of modern instrumentation op amps.
My sparkos 3602 Opamps sounded way better than incremental upgrades; speaker cables, fuses, usb cables or even newer dacs. These are 5 dollar opamps are in most major rave reviewed dacs. What a mistake it was listening to audioscience, when i started this hobby . Why? they let measurement tools do the listening for them not ears and every known objective scientific study they push happens to be done with old farts that have moderate to severe hearing damage. Never once they use Youth aka golden ears for test/study. That's why r2r tubes opamps and cables all can give a more enjoyable better sound than these measure well DAC.
Agree about ASR but I replaced sparkos with the Weiss and it blew it out of the water. Andrew is a great guy but his op amp is muffled through a pillow by comparison. You get what you pay for.
@@xfloodcasual8124 One problem with these discrete amplifiers is their size, which too often is a total hindrance to their integration in a device. We are not talking about their current consumption, which could damage the power supply of the device. 😢
lol I also had the corona hairstyle (almost as long as that of a hipster) until a few weeks ago, luckily the hairdressers are back in business over here since 1st week of June
The 709C was not an audio-specific component, so I'm not surprised a discrete design performs better. The NE5532 and other op amp chips specifically intended for audio applications are whole different category and, when designed in properly, are very difficult to beat with a discrete equivalent because it's hard -- often impossible -- to source discrete parts that have the sort of precision that's easily achievable by a decent silicon foundry. IOW, as Paul himself found out with his vintage Studer mixing desk, "op amp rolling" is a waste of time and money in an expertly engineered piece of audio gear that uses carefully selected, audio-specific ICs for the op-amp function.
Chip op-amps can have extremely good performance now days. Especially for a small signal solution like a DAC.
Audiophile op-amps exist that beat many discrete circuits nowadays. Also, analog studios are using lots of op-amps. Your speaker setup is where things sound different when the electronics is properly engineered discrete or not.
With regard to speakers sometimes you want a different sound and you can't buy a new set of speaker but changing the op-amps can give you a different sound
@@daysofgrace2934 If I want a different sound, I would not bother to change op-amps as any difference between (good) op-amps will do near no audible difference compared to anything you can do with room acoustics incl. speaker placement. A typical audiophile speaker distort many many times more than a typical audiophile op-amp.
I've found that most problems associated with op amp circuits can be booked down to two major issues: 1. Budget-price op amps with low bandwidth and 2. Designer "cheaps out" by having too much gain in each stage. This result in high levels of slew-induced distortion (sounds absolutely awful). In extreme cases poor high frequency response and phase shifting as well. I have a guitar amp with both of those problems. Rather than replacing the op amps and adding 2 more gain stages, I replaced the entire preamp section with a two stage (zero feedback) circuit consisting of one 12AX7 tube. Sounds wonderful.
I guess why all jfet distortion sound better because you force to use stages while cheap amp just have one op amp stage for distortion for guitar amps.
Yes, the answer is "it depends". The 709 op-amp was among the first of such, and of course had many warts. The best of today's devices have few warts and take advantage of all manner of characteristics not possible with discretes, such as device matching. And keep in mind that recording media, such as mixer boards, run audio signals through sometimes a dozen or more op-amp circuits before they're plopped onto the recording media. The A/D converter also contains an amplifier chain to drive its digital circuits. So in my view, ignoring the distortions caused by the recording chain and focusing only on reproducing devices is naive..
As a teenager over 30 years ago I used some already ancient 741 op amps to build a couple of preamps for magnetic phono cartridges. If you ask me, they sounded great! Low noise and distortion, great equalization. Wish I could remember where I got the circuit diagram, some dusty/musty old electronics hobbyist magazine.
Hi, Paul, I think, we both are about the same age and for sure from that generation which started to use OPAmps the first time in the electronic new age. I remember well, the now vintage 709 (with external capacitor for frequency compensation), the the 741 (without it). I always considered, and still do, an OPAmp being a device, like a transistor or a tube is one, with which you design a circuit for a special purpose. Be it a unity gain buffer, a filter, a precision gain amplifier, an impedance matcher, you name it. What made and make this device unique is its insanely (this is how we felt about it) open loop gain, its extremely high input impedance, the ease of how you could precisely determine the gain you wanted or the input impedance for the same matter. And with mathematical precision a filter characteristic. Now I come full circle: it is a device, not a circuit (I am not looking inside the housing now) by itself. But now back to the question about sound: you will have to design the sound characteristics yourself, the device is not having one! That said: of course it does. But this wonderful device gives all the responsibility back to where it belongs: the designer. Cheers! Keep going strong. Hans.
Paul, you look a lot more relaxed and less stressed now!
I think this Cov19 thingo worked well for you! Perhaps a 'one day per week' at the office would be in order?
..and Chris is doing a fine job in your absence, too!
Thanks for keep these daily videos coming and stay safe out there!
Thanks and I do now spend a few half days in the office. Chris always does a fine job!
Ok Paul....we're gonna need a full tour of the McGowan estate....including the lake and the back 40.......
I agree! The area where he lives is absolutely Gorgeous, but if I'm not mistaken the lot sizes in that area are run-of-the-mill 8-10k sq ft suburbia plots. We may have already seen most of what there is to see. Boulder doesn't have space for properties with a back 40! Nowhere close to town anyway
I would say that an op amp by definition has very high open loop gain, and is thus unusable without feedback. You can make a discrete amp with low or zero feedback, but it wouldn't contain an op amp.
A couple of years ago I bought an op amp from TI to improve sound quality from my hybrid phones amp. Already I had upgraded the Chinese valves with RCA and Russian valves.
After swapping the op amps...the sound was much worse,muddy, less detail.
Again I change op amp with some NE ....yeahhhhh.great improvement!
One of the problems with "upgrading" op amps is that newer chips may be much faster. They are more prone to HF oscillation. You have to look at the circuit with a scope to make sure it's behaving. You may have to partially re-design the circuit to optimize it for a new chip.
I could be wrong but I thought that one problem with chip op-amps is that they are more susceptible to internal capacitive losses / cross-talk as well.
When I purchased my Audio Technica moving coil phono cartridge back in the 1980s I knocked up a simple IC op amp based 20 dB gain circuit on a scrap of Veroboard. I still have it somewhere. It performed really well. No noticeable crosstalk. And it was as cheap as chips as they say. Certainly a lot less than a transformer.
In the 60's Electronics World had a construction article on building a tube based opamp. It looked interesting so I drew up a parts list and set out to see how much it would cost to build. This design had 4 or 5 dual triodes, power supply and all the discrete resistors and capacitors. By the time I was done it was goping to cost a hundred dollars with out any consideration for a chassis so that project was moved from the active to inactive list.
Several years later after getting out of the army I worked for Analog Devices as an engineering tech and got to work on some cutting edge discreet designs that had very good characteristics. I built a phono stage with a pair that sounded very good for it's time but that would have been pretty expensive if you didn't have access to engineering samples.
What a beautiful deck you got there!!!!!
As long as you dont overdrive them they should be pretty clean and very linear at audio frequencies nowadays. The old stuff is a far cry compared to what we have now, so its not really valid to use the 709C as an example. Modern op-amps are far better and made for precision amplification, and while not all are equal, there are plenty of low noise op-amps that will do just as good given the large amount of feedback available to clean up any remaining nonlinearities at the output. Fair enough to say that most aren't since many are low voltage now to cater to low voltage electronics, but there are plenty just for audio these days. Also since all the silicon is on one die there will be less issues with thermal effects since it will all track, and everything is matched. To say they "don't sound as good" kind of ignores the fact that a lot of the same topologies are found on an IC as you would find in a discrete version. In reality, there probably would be almost no discernable difference at audio frequencies, and with audio almost always the speakers/headphones will be the dominant source of distortion for sound reproduction. In an audio amplifier it makes sense since you need the voltage swing and the final stage has to supply amps to 10's of amps and not milliamps, but with that said there are audio power op-amps with THD in the parts per thousand for power outputs < 65W. Edit: Im assuming the "two device op-amp" was probably a pair of Darlington transistors since there was feedback being used.
Hi there Paul, I have been watching your video's recently, thank you. The reason I am posting a comment here today is because I believe we have met one in other back in the 80's, while I working in an Audio Shop located in Phoenix Arizona, owned by a guy named Buzz Jenson and his Audio Alternative store I was just curious.
Back to audio and OpAmps. I worked in the semipro and HiEnd audio sales field for many years, and remember in the late 70's early 80's you could definitely hear the difference between the brand names of the day, most of the on a chip designed receivers, amplifiers had different musicality qualities directly associated with the design choices made, at that time I would always recommend to our customers that discrete was the way to go, these days maybe not so much.
Anyways always a pleasure to check out your video's.
Really enjoying the last few of your videos. Would really like to see a step-by-step DIY project in the future.
Lloyd Stout So is Nelson Pass and he has his own DIY site. Many designs over the years from Nelson for public use. Don't you think it's up to Paul if he would like to do that? I respect Paul and whatever he wants to do I would be behind him.
I loved the reference to the Quintessence Preamp, I had one before switching to a dB Systems preamp. The Quintessence was very elegant looking and felt like a quality product. All of these came after using tube Dynaco stuff and eventually went back to tubes (ARC) by 1976 only to go with First Watt stuff in 2011 and TBH I haven’t looked back yet.
Great answers, one of Pauls best episodes👍👍👍
As always a great discussion. With op-amp's, the circuit topography and power supply all have a influence - not the chip solely. Its an accumulative.
its like tube rolling, all opamps chips has an unique flavor, some sounnd super crhistal clear but not musical, some others sound musical but no extra details, its up to your ears to chose what is "better"
Why does the audiophile community hate feedback? Do you like to make your job harder!? Theres nothing inherently wrong with feedback, and if the circuit is already close to linear, you will drive the final distortion products down below the noise floor if there is plenty of open loop gain available. A lot of gain is needed since the error correction is strictly proportional, so some portion of error will inevitably remain, and less feedback means more residual error at the output. More feedback, as topology and stability permits, will mean less error and better linearity. Almost every commercial amplifier uses quite a lot of feedback to avoid non-linearity and distortion, esp. in the cross-over region. Not using it almost certainly means more distortion unless you're using exceptional devices or a complex topology. Even then it would be _even better_ with feedback. Why leave that improvement on the table? Edit: Using the OPA134 as an example, it has a listed THD of 0.00008%. Thats pretty goddamn low! Granted it will inevitably climb a bit above 10kHz, but so will your discrete one. Maybe not as fast with less feedback, but thats offset by other issues that come up when less feedback is used.
Wow! Learned a lot today. Thanks for a great explanation.
I don't know... the LME49720 sounds pretty darn good.
If a system is designed with discrete components, then there is no need for "discrete op-amps" because those are just functional replacements for op-amps. The topology should be different enough that when designing a discrete component system with minimal feedback, op-amps of any kind wouldn't be incorporated into the design.
So "discrete op-amp" are really more for the OCD type who bought op-amp based systems and can't sleep well at night.
There is nothing wrong with feedback, if the open-loop gain-bandwidth product and slew rate are high enough. The problem only comes in, when there are multiple response poles too close to the audio range, requiring heavy compensation for stability, which then pulls the slew rate down.
We need to get away from the myth that high feedback is uniformly bad. It becomes less and less true as time goes by and devices become faster.
You're talking about op amps in the 70s. Effectively noiseless, wide bandwidth, high slew rate op amps are dirt cheap now.
I had a Quintessence preamp and Ugh those 709 op-amps were bad. I remember it as being kinda dark and vague sounding. I was going to build a better circuit in it and once used it as a test bed for op-amp but eventually took it apart and kept the faceplate and knobs but threw the rest out. I also had the SWTP preamp which had plug in discreet op-amp boards but I never got that to sound great so I kept the boards but threw the rest out. Then I bought a Rapport Pre-1 preamp and it had some pretty serious stability issues. I cleaned up the grounding and replaced the pot and it sounded pretty good. Then one day lightning hit outside and blew the output stages while I was listening to Pink Floyd (loud hum both channels). I found that the high level stages were sealed in epoxy. I broke one open and was surprised to find an op-amp inside. I was of the opinion then that discreet was the way to go. I built replacement circuits for the high level stages using the same op-amp. This has been my reference for solid state preamps since then. Sounding better than Mark Levinson and Classe and pretty much every high-end preamp I compared it to.
A good chip op-amp can sound great. The design of the rest of the piece of equipment matters enormously too. The power supply has to be high quality, the layout of the printed circuit board in terms of component placement and ground plane design has to be good and input and output circuits have to do things like impedance matching with good results in terms of connecting to other components. Is the chassis well shielded, etc? It's the complete package that counts.
That's true as Paul says. You can have an IC. oppamp which has incredibly good sound quality and very low noise.
If you have two preamps from 2 brands that cost the same.
IC. op amp can quickly have better sound quality than the sepparate transistors preamplifier has.
But it can quickly be the opposite that separate transistors have better sound.
It is too easy to say that separate transistors are always better. IC. op amp does not always have to be worse than separate teanistors.
IC. up amp came the first time approx. in the late 1963
The operational amplifier was invented in the 1940s. The first IC op amp was made by Bob Widlar at Fairchild Semiconductor in 1963.
As always,A great video. PS I can see on left video side that electricety poll - great design ...
Nice one! 👍
I didn't even notice that before...
Sorry to disagree. It sounds like Paul is a few years behind (or maybe a few decades?).
I totally agree that MOST op amps are not great at audio, but there are many that are designed specifically for audio and some, even better, for instrumentation. An integrated op amp has the advantage of size. This means lightning fact response (if it is well designed and has low internal compensation) which allows for very high open-loop gain thus heavy feedback. All distortion is eliminated (or at least negligible) except that of the input differential pair. All internal noise is shunted except that of the differential pair.
The differential pair has a built-in advantage over discretes : it is paired by construction. Man parameters in semiconductors can vary wildly from one unit to another. For instance the Hfe (current gain) of transistors of the same make and model can vary from 500 to 10,000 due to small variations in the masks in the silicon. So, if you make a discrete op amp, you have to buy many and pair then individually. In an integrated op amp, the variations are the same for the two input transistors, thus greatly reducing parameter variations to a maximum of about ±10%.
As an instrument and audio designer, I did use high quality op-amps (not something like the 324!!!) and preceded then with a discrete pair of carefully selected matched very low noise large bandwidth transistors (either bipolar or FET), whose collector or drain was connected to a low-capacitance current source. The outputs then went into a large bandwidth integrated op-amp. These transistors were placed as close as possible (1 or 2 mm) to the op-amp input. The results were outstanding since the input transistors were submitted to no base-to-emitter or gate-to-drain change, thus retaining fixed parameters even through changing signal conditions, providing for a wide distortion- and noise-free dynamic range. Frequency response was, of course, flat from DC (0Hz) to at least 50Kz.
old amps have no ship and sound good ..so good
Anybody else here get a Fosi v3 monoblock amp and received 2 Muses02 chips? I bought 2 amps so they sent 2 chips, does anyone know if the two Muses02 chips are both supposed to be installed into one amp? Or is one chip for each amp? The reason I'm asking is because when you open the amp, there's 2 chips that look like they can be replaced by the Muses02. I know this is probably stupid, but this is my first time using an opamp
Weve yet to see the stables
I misunderstood the subject before clicking on the video, I was expecting a perspective on IC amplifiers that you would find in a car stereo, bookshelf stereo, or a Bluetooth speaker. There are even some popular car amplifiers that use IC amps. Any opinions on those? I realize that this type of equipment qualifies more as "consumer grade" than "hifi"
Hmm maybe last century this was correct. This century I've not seen a single equivalent item where discreet beats integrated. It's easy to see why when it's thought about too. Longer signal path, no thermal coupling, mismatched components not the result of billions of dollers of r+d.
Vertical power devices, at >>100 V.
In particular the SiC superjunction NPN's ("GeneSiC", only really still made for high-temp applications because the chips themselves easily work at temperatures kitchen ovens are not willing to go to (350-500 C is usable depending on packaging), as the need for DC/bias base current had them loose against the JFETs) and the SiC JFETs from e.g. UnitedSiC (now part of Qorvo) and maybe soon others again (last competitor ceased about 10 years ago).
Those are sadly only good to a few MHz (and that's with frequency-dependent distortion there that needs compensating!) due to their gate resistance, but they are extremely robust and very good (power) amplifiers.
(The JFETs are designed for switching applications like power supplies, but they are safe to use in linear mode and the only real vulnerable part is excessive sustained forward current through the gate-source diode. Like, around 10+% of nominal TDP used for that diode. Everything else is just a "eventually it overheats and breaks" issue, afaik.
I postponed my plans to get some proper measurements of the RF behavior, mostly due to some difficulties brought on by the high voltages (1700~2200 V avalanche, roughly, depending on if you get a 1200V or 1700V nominal one.).)
Depends on the ópamp or discrete transistor using. I like discreet JFET's . Very Tube Like sounding . Well one in particular I found just amazing quiet with high output. And somehow they were all exactly the same. If you know anything about JFET's that never happens. You have to bias them individually. And I'm not telling you the model number or anything there mine, all mine. I will tell you they're very similar to the ones that Marantz were using. To make their amplifiers sound more tube like.
Great answer!
You just need to try them to see.Personally i found amazing op amps which cost 2-3 pounds.The sound is just mind blowing.The bass and the richness of the sound is amazing.AD827SQ / 883B that's the op amp.
Really good info
And to find an engineer that can make that better op amp, they undoubtedly work at a large chip manufacturer with Deep Pockets not at small audio firms. Anyways thanks for the history lesson. Today's op amps are not your grandfather's op amps
how do I make this channel stop appearing in my feed?
In my experience it depends a lot on the rest of the system. I have a normal highend consumer stereo and to make that sound at all special it requires (slightly DAC dependent) the best discrete opamps I have in my collection in order for the system to have decent dynamics and overall sound. I also have a home built/designed system that uses pro drivers and pro amps. With the same source into the DAC I use regular chip opamp (really good one...) in order to keep the dynamics in the realm of sanity.
So, discrete opamps are way better for home hifi because home hifi needs all the help it can get, but with pro gear you have so much power and dynamics that you are just fine with regular top shelf opamps.
No opamps were hurt in these experiments and the refereed to opamps in this story are Sparkos (discrete, from Colorado, about 10 miles from PS Audio) and OPA 2134 and Muses 02 chip opamps. Bryston, Muses 01, Muses 03 are all fine units, too)
@Lloyd Stout Sure, Ace, tell me all about how smart you are...
Yesterday in my RIAA preamp I compared a pair OPA2111KP what it come with originally. To a high end specifically made for audio MUSES 01.
The MUSES 01 was superior in many ways so no contest. Today I got time to compare a descret op-amp against the good MUSES 01.
It is hard to say something definitely about witch sounded better than the other they were so good both of them. But the descret op-amp played lower down in frequency that the MUSES 01 didn't. So it won.
So this "random" pair of descret op-amp sounded better in the end, than a high price high end IC op-amp made for audio.
I mean "random" above because it is only the first and only that I tried and there is many other variants. (Name: moing hd-08)
And I paid a lot for the IC op-amp in the same ballpark as the descret ones at $100 for the pair.
One reason were that I realized that there is a lot of copies and to be sure that I got original op-amps I decided to get them from the manufacturer. So I know they are genuine.
So if I never had heard the descret op-amp then I would also think that IC op-amp is as good as it gets.
Now I know differently.
And that kind of money for "simple things" as op-amps I would need to multiple many times that amount, to get that type of sonic improvement by trying to upgrade my cartridge that anyway would be fruitless when there is a hi end IC op-amp in the signal path that do not amplifying the lowest frequencies.. .. everything matters.
So it would not help if I threw infinite amount of money on a cartridge.
What do you think of the THAT Op Amps that are specifically designed for audio use such as the THAT 1512 or THAT 1580? Are they leveling the battlefield so to say?
The THAT 1512 and 1580 (and similar earlier products, like the INA103) are specifically microphone preamplifiers, not general op amps. (Although I suppose people could use them as a gain make-up stage after a passive summing mixer). For that purpose, they have excellent specs.
Today, unless you are designing to special requirements such as optimal Noise Figure for a given source, high voltage or high power, it's usually not worth a designer's time to develop a discrete-component op amp.
www.ti.com/amplifier-circuit/op-amps/audio/overview.html
www.cycfi.com/projects/six-pack/op-amp-shootout/ (2017)
tangentsoft.net/audio/opamps.html (last update 2016)
Of course, incorporating a suitable op amp into a product and achieving its best performance still requires an experienced and knowledgeable designer. Using an IC doesn't guarantee freedom from all pitfalls.
Too bad Paul didn't mention the story of how a competitor designed a preamp with the 709 that sounded much better than his - and why.
the discrete one will cost more - assuming other aspects are on a level pegging.
might be an interesting 'shoot out' - a discrete at the same price point as a chip, though the chip with more sophisticated psu...(if such a 2 competitors van be found...)
Maybe $5 - $10 in COGS? And likely a large differentiation in sound. No brainer in high end audio. Inconceivable for consumer grade.
Burson opamps are great value for money, especially V5i, that is not too big.
I am using them in my Asus Xonar STXll, much better sound now than with the originals.
If you have not done things to improve the power delivered to the STX... there are things that make a big difference, for not toooo much cash.
bvocal What can I do to improve the power to the card? Do you have any specific recommendations?
I have soldering iron and know how to use it. :)
@@bergennorway No soldering needed. I know I'll be slagged for this by geniuses like Lloyd Stout, but here it goes, and for anyone who thinks this is shit, stop on by and I'll a/b this for you... I took out the not so great power supply (the computer's) and replaced it with a Seasonic PRIME 550 Platinum SSR-550PD2 550W 80+ (it was as far up the food chain as I was willing to afford ) and connected that to the wall with a Pangea AC 14SE MKII, (which costs somewhat less a few years ago), and lastly I put in an odd power filtering board that is designed to filter the power going to add in cards through many capacitors. The before and after difference of the whole package was dramatic. I had no interest in figuring out powering the card with an external power supply.
While all that worked and is good, to be honest I have matched the output quality with a few other -not expensive- DAC options (which at the time I was not convinced, say, a Topping DAC for $100 or $130 or $200 etc would be any good, I was somewhat wrong about that, and with many of the Topping DACs you can swap opamps.)
So what would I do now? Well, an STXII costs 220$, the power supply 140$, power cable, 90$, bus power filter $45, opamps, 200$= $700, my external DAC with all the extras is about $1000, and a Topping D30 (swappable buffer opamp) is about 130$, and a D50 (no replaceable opamps, unless you soldier) is 500$ plus, if you want to be equal, 90$ for a power cord and a USB cable and maybe a sibling power filter for the USB output from a computer -40$
so, all up, less money, all have good DAC chips...
So if was the broke human I am and wanted a low cost DAC with a swappable opamp, which is a feature I rather like, I would look at a Topping D30...
cheers
Not unless you want to individually tune each physical unit. Put that labor cost into better chips, and it's no contest.
The problem with chips is that there are a lot of fake ones.
The problem with transistors, is that there are a lot of fake ones.
sound-au.com/counterfeit.htm
The problem with tubes, is that there are a lot of fake ones.
jacmusic.com/nos/images/Fakes/index.htm
The problem with resistors, is that there are a lot of fake ones.
www.tsec.ltd.uk/resistor-obsolescence-and-how-to-spot-counterfeit-resistors/
The problem with capacitors, is that there are a lot of fake ones.
www.dfrsolutions.com/counterfeit-capacitor
That's life in this world overrun by criminals and frauds. Most of the phone calls I receive these days, are from criminals trying to scam me. But I have fun with them.
In my humble opinion, you have a better chance of improving the sound of your device by putting in a better chip than by changing the speaker cable from $10 to $100 per meter.
Pretty much every readily available op-amp IC out there is designed without any precision resistors or capacitors. Integrating these is not easy! (Anyone know of any exceptions? I could be wrong). Even with the best designer on the job, that's a severe hindrance.
Generally you don't need precise absolute values for R's and C's on the chip. The parts which set the overall gain, pole and zero locations, etc., are external to the IC.
But on-chip resistors can be matched to precise ratios, as can transistor offsets and current-mirror ratios. In discrete designs, those things need to be done by hand, which greatly inflates labor cost.
I should also point out that some of the best low-noise preamp designs today, are hybrids of discrete transistors and IC op amps.
I frequently use the LM833 It's VERY VERY good sounding chip opamp. It's made for audio and the distortion bandwidth is below 0.005 (datasheet says).
We don't need morte than that if we are diying a home audio! Fits most uses, worth the test!
You should try LM4562NA and OPA1656 or OPA1612. These chips are incredible.
It really depends on what you want to call discrete...
A chip with one single opamp on it is often also called a "discrete" component.
Because it's one spare opamp instead of multiple that for example makes AND/OR ports.
These discrete AND/OR ports can be grouped into things like J/K flipflops or D latches.
These spare D latches can be grouped into a clock divider, which are often used as a counter.
@Dave Micolichek Yes exactly, "individual components".
But when is something an individual component or built up from subcomponents?
You can divide components up into subcomponents until you're on molecular level...
The term "discrete" is just as undefined as "much", "large" or "good" and can only be defined with the necessary context required to do so.
@Dave Micolichek I know, the molecular example was a bit weak, since it has no explicit relevance to electrical properties of the complete component in most cases.
However, for example a traditional PNP/NPN transistor is made of at least three components for the junction, some metal leads that extrude from the package and some internal wiring that connects these to the junction.
For example a standard electrolytic capacitor is made of metal film, a dielectric film and some type of electrolyte liquid. Two metal wires are attached to the anode and cathode foils and protrude through the package.
The amount of components in both packages of those examples are significantly less, that's about the only significant difference.
@Dave Micolichek One could also argue the other way around for components like transistors or diodes, since it's just one die cut from a wafer after all.
Just like an opamp or microprocessor.
@Dave Micolichek Yes, with some words it's very hard to come to a concrete definition; as simple as it might seem at a first glance...
First you say generally better then at the end you say always better. Wowzers.
Context. It really depends on the strength of he rest of the system. I have discrete opamps that blow everything away, but they are no the best choice in all situations, I have a situation where IC opamps are the best sounding choice, and in another system of mine discrete is the only way to top sound.
IC op amps have way tighter tolerances than discrete ones.
Wierd subject. I do not know of any ADC or DAC that is high speed and does not either contain or use an op amp. Makes about as much sense as an amp that does not use electricity. Single stage mic amps in condenser microphones often do contain a small FET based class A amplifier to transition from the very high impedance microphone element to drive the low Z balanced output, but from there almost every EQ, Pre Amp, Mixer, DSP, etc are universally Op Amp designs. They became popular for low noise, high speed, very controlled gain, high linearity, low phase shift, etc. All instrumentation amplifiers are based on them.
In discreet vs integrated, do not overlook the actual advantages of the small package. Low noise, very high slew rates, very high CMMR, very high bandwidth, etc are commonly used for precision instrumentation applications. Compare any discreet op amp performance with for example the performance of a MAX4146 for example. The single die is thermally stable, low noise, slew rate of 1,000 V/uS compared to the obsolete part Paul mentioned with a slew rate of 0.25 V/uS. A bandwidth of 130 MHZ, is of a 3 op amp design for instruments with a CMMR of 70db at 10 MHZ. Flat response to 90 MHZ with 0.1 dB gain. Very low noise of 3.45 nV. Sorry Paul. Try some modern very precise parts. Discreet parts have problems due to the physical distances, inductance, capacitance, etc that are greatly reduced with the small dimensions inside a small SMT device.
Cheaper and better is the advantages of modern instrumentation op amps.
Chip amps sound harsh and brittle to me. Discrete all the way.
Gullible fool
My wife’s got an input that causes her to invert if I use it...
I have half a mind to tell her what you said.
Marianne Oelund hey so do i, but I also have half a mind play my synthesizers so loud that people in my apartments will think aliens are invading...
@@graxjpg
Do it when the comet is visible, for best effect.
Paul has nice legs #yeahiwentthere
I designed and built several add-on components for my audio system using commercially available high-end audio op-amps with excellent results.
My sparkos 3602 Opamps sounded way better than incremental upgrades; speaker cables, fuses, usb cables or even newer dacs. These are 5 dollar opamps are in most major rave reviewed dacs. What a mistake it was listening to audioscience, when i started this hobby . Why? they let measurement tools do the listening for them not ears and every known objective scientific study they push happens to be done with old farts that have moderate to severe hearing damage. Never once they use Youth aka golden ears for test/study. That's why r2r tubes opamps and cables all can give a more enjoyable better sound than these measure well DAC.
Agree about ASR but I replaced sparkos with the Weiss and it blew it out of the water. Andrew is a great guy but his op amp is muffled through a pillow by comparison. You get what you pay for.
@@xfloodcasual8124 One problem with these discrete amplifiers is their size, which too often is a total hindrance to their integration in a device. We are not talking about their current consumption, which could damage the power supply of the device. 😢
Paul has COVID hair (long) like me. lol...
lol I also had the corona hairstyle (almost as long as that of a hipster) until a few weeks ago, luckily the hairdressers are back in business over here since 1st week of June
4:32
Thank me later
i was never First on the comment section, Hope this time i succeed
What did you gain from it
@@BijBijTCG you replied to him..
So apparently...your attention.