KIH #38 – Chips Ahoy!

38_header

ICky. That’s a pervasive high-endish reaction to ICs, short for integrated circuits. Of course most D/A converters plain wouldn’t work without ‘em. Even a €20’000 Gryphon relies on an ESS Tech Sabre chip. And what is a fashionable FPGA but a large-scale integrated chip? For this component category, hi-end’s chip on the shoulder makes far-reaching allowances. ICs are not only okay.  They’re hi-tech and studded with amazing specs which has the tech-obsessed drool over 140dB noise floors and imaginary 32-bit resolution. It’s once we move away from converters that the hate kicks in like a clunky Harley Davidson clutch.

38_goldnote_volume

47lab’s infamous GainCard amp might explain why. Based on a truly minimalist circuit ending in op-amp outputs, pictures of its innards soon incensed various forum posters. Converting the few parts into a BOM (bill of materials) exploded into outrage over the machine’s actual sell price. The War of the GainClones has raged ever since. Like Ahnuld’s Terminator, the clones keep being back, albeit at the far outer fringes of snobbish respectability. Nobody who’s actually heard one denies its wonderful sound. But the most popular National Semi chips for these applications deliver 25wpc or less, making most GainClones transistorized low-power SET in the minds of people who consider such power ratings unfit for real speakers.

38_goldnote_outputs

As they always do, exceptions exist. Blue Circle has used undisclosed ‘military’ or ‘industrial’ high-power gain chips in multi parallel to rewrite chip-amp power figures. In general however, opposition to the genre, from an engineer’s perspective, seems to rest on two arguments: lack of creativity; and low monetizing. Regardless of actual performance, op-amp outputs based on a part’s basic application notes don’t seem terribly creative. They appear fit more for a first semester project of an engineering 101 class than desirable to any audiophile designer who is keen to show off his/her  proprietary circuit design brilliance. Because standard op-amps usually aren’t as costly as doing the  same job with many more discrete audiophile-approved boutique bits, the finished product can’t sell for as much. In short, GainClones are entry level, not upscale. They’re fi but not high fi.

38_goldnote_hero

What happens if we broaden our definition to include any gain or attenuation stage—voltage, current, preamp, driver, buffer, output, volume—that uses opamps? Suddenly we spot the many-legged critters in unexpected places. Costly Pass preamps embrace a high-zoot Muses chip for volume control. The $2’850 Resonessence Veritas DAC runs fully balanced opamp outputs via ADA 4898/2 types. The $2’500 COS Engineering H1 drives headphones via a balanced output stage of two Texas Instruments LME49600. The €5’700 Gold Note CD-1000 runs a fully balanced Burr Brown OPA2228P output stage and an analog TI LM1972M on-chip stereo volume control. Our €10’000 when new Esoteric C-03 preamp revels in opamp gain. None of it is entry level by any stretch.

38_cos

As is true for GainClones, opamp presence and excellent sonics can coexist, comfortably. Plenty of propaganda to the contrary, from Ayre to Burson, would suggest against it of course but an open mind attached to two open ears can find sufficient evidence that opamps needn’t be bad per se. Cell phones wouldn’t work without ‘em. Neither would portable players. Open those up. The by far biggest thing inside is the battery. Everything else to make actual sound is chip-based. Since those industries dwarf hi-end audio, IC developers can afford to spend resources and supply them with better performing chips. Not all of these efforts are about cost savings. Many are about superior performance with better specs.

38_cos_hero2

The moral of today’s little ditty? Don’t issue a blanket diss on opamps everywhere. Like switch-mode power supplies, what is said about them and what they can actually sound like routinely doesn’t gel. It may not be as sexy to read propaganda about opamp circuitry than advertising that promotes discrete boutique parts from the usual suspects. But should that really stop a smart shopper? Next time you catch yourself disqualifying a hifi machine just because it uses opamps other than D/A chips, hit ‘pause’, then ‘back’.  In my own case, one of the finest headfi/DAC decks in recent memory was the COS Engineering with its forbidden SMPS and uncouth chippy outputs. That’s exactly why I acquired it. It relegated a trio of AURALiC Vega and two Questyle CMA800R mono amps to the unused hifi rack, doing the same thing better in a single box and for decisively less money. Rather than feel ICky, opamps can give you a hICky. Smooch!

38_cos_hero

Written by Srajan Ebaen

Srajan Ebaen

Srajan is the owner and publisher of 6moons. He used to play clarinet at the conservatory. Later he worked in audio retail, then marketing for three different hifi manufacturers. Writing about hifi and music came next, then launching his own mag. Today he lives with his wife Ivette and Blondie the cat in a very small village on Ireland’s west coast, between the holy mountain Croagh Patrick and the Atlantic ocean of Clew Bay in County Mayo’s Westport area. Srajan derives his income from the ad revenues of 6moons but contributes to DAR pro bono.

9 Comments

Leave a Reply
  1. This tradeoff and the accompanying heated debate is not limited to audiophiles. In software development we call this “use a framework vs code by hand.”

    In this case, the IC manufacturer is Facebook, Google, or Microsoft who offer development packages that have solved 90% of the job of writing a piece of software, have theoretically fixed any flaws along the way, and have designed these frameworks such that they can be figuratively plugged in like an IC.

    But as with an engineer who designs a complete audio amplification component, one has to ask “Did the original framework/IC designer have the same needs, values, and use case that I do? And does this solve my problems or will I have to do more work to wire it up than I would if I had started from scratch?”

    The answer is never completely clear but I appreciate your approach to ask “How does it sound?”

  2. In the spirit of kih – Harley Davidsons do not have clunky clutches.
    What you are hearing is the uneducated putting the bike into gear before the clutch is 100% disengaged.
    That is the clunk.

  3. Ian: Point taken. Being an ex motorcycle fan (including a BMW Enduro and a Moto Guzzi California Highway Patrol model), I heard my fair share of what you described so I thought I’d use it as an example people might relate to -:)

    Geoffrey: Ditto for point well taken. Just so, opamps per se still need surrounding circuitry to work. That would seem to leave plenty of leeway for an engineer to get busy and optimize his chosen off-the-rack chip/s. My main point really was to show that non-discrete circuits can sound a lot better than ideology would have it; and in some cases will outperform dearer discrete implementations; or match them for less money.

    • Maaaaate, how do you become an ex-motorcycle fan?
      Moto Guzzi Calis – sensational, I wish I’d kept the 1100 Sport I had in the mid ’90s. When the planets aligned, there was no better fun to be had.
      By the way, (and off topic so I’ll keep it brief) the Peachtree iNova works a treat with my Zu Soul Superflys. I’d like to give the new 150 an audition. One day.
      regards, Ian

  4. Why are people saying that Kimura San’s amps are a ripp-off? He got the idea and made it work. No 47 Lab means that there wouldn’t have been a new path in audio. I really do not think that 47 Lab have ripped off people by asking too much money for their product. Just listen to it and you easily find out that it sounds great. So the price is in order for the enjoyment you get. Yes, it can be made for less money but who got the idea to build the amp? None of the guys who say it is too expensive. Is a Thonet chair too expensive and a ripp off? No, it is the original and a timeless design, that’s what you pay for. Same with 47 Labs. So if it can be made cheaper, ok. But who got the idea?
    No, I don’t own their product, but I do respect the brainpower that went into the development and the finished product. It is simple and it works brilliant.

  5. Thanks for championing the obvious. It’s not the package, it’s the innards. My Ray Samuels Nighthawk is a case in point. I’ve nearly been swayed several times by the doubt inflicted on the humble op-amp based phono pre-amp only to have a great listening session that validates the wisdom of a great design using hand chosen humble parts.
    If ever there was a time to enjoy the marvelous differences of the products offered to the listening public, this is the time.
    The more things change, the better it gets.
    thanks, be well
    rick

  6. Value is in the ears of the listener. If op-amps are used, then my only request is that they not be soldered in place so I can swap / upgrade them. This is the case with my Eastern Electric DAC Supreme and upgrading the op-amps has proven to be a very cost-effective upgrade to an already high value DAC.

  7. It seems that every so often one has to be reminded that “it’s all in the implementation”.
    The whole is greater than the sum of the parts. Maybe the next line of opamp based
    whatever can be named Aristotle Audio.

    All the best,
    Tim P.

  8. Rainer: Actually, 47lab didn’t ‘invent’ it. If you remember the GainCard’s genesis, its designer needed a known-reference amp whilst working on his Pitracer transport. He didn’t have time to design an amp from the ground up. So he quickly wired up some opamps for output devices pretty much like the application notes would tell you to do it. That was purely for his own design working out the transport’s intricacies. The impetus wasn’t making a stupendous amp, the impetus was a quick’n’dirty solution which he knew could serve duty in his lab. Apparently enough people heard his one-up and thought it sounded fab so eventually it got tweaked up and into production. At least that’s the story I remember from my review of it.

    I’m not taking anything away from his daring move to market it for high-end purposes. Again, I reviewed it and it sounded very good. I reviewed a number of GainClones afterwards and ditto. I still have a pair of monos. The ‘problem’ with the GainCard only arose later when folks published opened-up pix on the Internet. What was the ‘problem’? The GainCard was squarely priced based on what it competed with sonically rather than what it cost to make. That’s perfectly valid and happens a lot. In this instance the delta between parts cost and retail simply incensed the peanut galleries; *and* because of the circuit’s simplicity, compelled them to roll their own.

    Once DIY had their hands on the concept, doing the same thing commercially got a lot harder. After all, commercial implies making a profit. Once it’s open knowledge what it costs to make, opportunities for expansive markups shrink. Just so, Clones Audio from HongKong today remain fans of the concept. They keep making GainClones whilst selling them for much less than the original. Even their name is an open hommage at 47lab -:)

Leave a Reply

Your email address will not be published. Required fields are marked *