[Ableza]

Indeed. On bike it's suicidal. Don't do that again.

Unless you are suicidal. And if you are, think of the coffee you'll miss.

 

[Pietro Cozzi Tinin]

Unless you are suicidal. And if you are, think of the coffee you'll miss.

Yes, think of this Italian barista hoping to make you a terrific espresso some time in one of your Euro-tours.

 

[scottbrown5271]

2019, Chapter 15:

...There was, however, another part we’d been playing with, the relatively new OPA1688 from TI. The OPA1688 could take the +/-12V easily—it was rated for +/-18V. It didn’t have quite the current output capability of the LMH6643, but it certainly measured a lot better.

Jason,

I'm pleased to see these products released, I hope they do well for you. I was also pleased to see you (rightfully so) bragging about the quality of the opamps you used in Hel. And you did it right in front of a bunch of self-admitted audiophiles!!

I decided to start studying just what an operational amplifier is a few days ago, and some of the answers I found were surprising: The "operational" part of the name refers to the amp operating mathematical functions, such as addition, subtraction, division, integration, etc... Originally designed as computing modules, c. 1930 or so, and after a decade or so of development the name "Operational Amplifier" was coined and a set of parameters for the perfect opamp were set down; Infinite input impedance, zero output impedance, infinite gain, and perfect linearity of output to inputs (this last one would have EE's working around transconductance droop almost 100 years ago...). They also must force both inputs to match each other through the use of feedback, which as a practical matter will result in clipping at the rail voltages when driven too hard, unless the designer didn't think that was a priority and didn't bother to address the exception conditions, in which case they might invert the output and probably release magic smoke. Whew!! That is just the beginning of an incredibly complex decision making process for someone designing circuits with opamps. Or not. The application notes for these particular opamps show a circuit that uses a +/- 5v supply and can sink 20mw at 30 ohms!! That should do it, right?

All of that to say that the OPA168x opamps are a seriously compromised design compared to the ideal, but those compromises and the high voltage rails are what makes them ideal for this application as a headphone amp (well, four each in series for each channel in this application). And of course this is about as non-standard of an implementation of these chips as you could do and still drive headphones with them. I love it when a venture into unexplored territory works out, and I am never disappointed by your designs; Always something new and different from Schiit!

Hel directly addresses one of my task-critical use cases, IP Telephony, but I needed it three years ago. So I fell back on what little I knew about sound production, bought a couple of cheap-ass mics, decided that the worst problem I had was a noisy environment, and bought a dbx 166S channel strip to process the input. Proper gain staging, compression, limiting, de-essing, expansion, mild eq, and suddenly I sounded good on the phone. I'm using a Focusrite Solo for I/O, and a Chromebook for the call software. Windows 7 couldn't stay connected with the Solo, or any of my other dacs, so I'm taking advantage of a Linux kernel in a user friendly environment in order to make my work flow more consistent. The biggest issue with all that junk is the balanced connections. Yeah, I said it, balanced sucks because the cables are big and heavy and expensive, the connectors take up a lot more real estate, and at least in the instant matter, NOT NEEDED. My production mic at the moment is a cardioid pattern face mic. It will tolerate a 12v bias, but really only needs a coupla few volts, so I feed an inline adapter with 48v phantom from the 166s which drops the bias voltage and buffers the output from the mic to get to 4v pp for the balanced connection. The big advantage of this setup is that all the pieces can be quickly configured to do something else if need be. Oh yeah, it doesn't pick up hum at all, either. Did I mention music? Modi Multibit -> Magni 2 Uber -> k702. Those components really cast a long shadow over the Focusrite Solo (2nd gen) which I really can't stand to listen to music through.

The major payoff with Hel is the utter simplicity of the setup. All single ended, cheap cabling, and industry standard consumer class microphones. And Hel will fit in my computer bag so that when I am inevitably called on to work while I'm on vacation in some far-flung corner of the planet again I will have all my tools with me. And, I'm willing to bet that it will sound better than the Solo (not hard to do, my original Fulla The Dongle is still way better, scratchy pot and all).

The only thing left to ask for is the much discussed digital preamp... I really need 32 bit floating-point dsp for the filters I want to run... I can get that for free, but I have to use a DAW running in Linux to get there. Talk about not user friendly, and the last thing I need is yet another computer in my setup.... And, as the banner on the advertisement above says at the moment, "KB EAR, Kiss Your Ear, Be Your Favorite." Words to live by...

 

[Charente]

They did my Gungnir upgrade from gen 2 to gen 5. I see no difference with the upgrade to Unison USB™. Unless the firmware upgrade, not involved with gen2 -> gen5 , is an issue here. Best way is to ask them. Maybe they already know what their position is.

I was referring to Jason's original statement ... that the UNISON firmware upgrade is indeed an issue with regards to the GMB ... but hopefully you are right about Intl. distributors being able to undertake this upgrade ... and happy to stand corrected.

I posed my question publicly as this may affect other internationally based owners considering this upgrade, when it becomes available. Hopefully Jason may make a further statement when he is in a position to do so.

 

[Scott Kramer]

...The "operational" part of the name refers to the amp operating mathematical functions, such as addition, subtraction, division, integration, etc... Originally designed as computing modules, c. 1930 or so, and after a decade or so of development the name "Operational Amplifier" was coined and a set of parameters for the perfect opamp were set down; Infinite input impedance, zero output impedance, infinite gain, and perfect linearity of output to inputs (this last one would have EE's working around transconductance droop almost 100 years ago...

Spoiler

Could not resist to mentioning this vid, in case you haven't run across it... Paul even demonstrates doing math! So cool how concepts progress, the origins of technologies, and how we iterate on them. Just a measly 600V rail to rail at that time.

 

[scottbrown5271]

Spoiler

Could not resist to mentioning this vid, in case you haven't run across it... Paul even demonstrates doing math! So cool how concepts progress, the origins of technologies, and how we iterate on them. Just a measly 600V rail to rail at that time.

I did see that video, that is is what got me to finally start figuring out what is an opamp. It takes him forever to explain something, but you will never misunderstand for lack of detail and clarity. I'm still a little foggy on the seeming paradox of the virtual ground at the inverting input in an inverting configuration with the non-inverting input grounded...

I like how the black-box amp concept started as a component in analog computers when everyone was sure that analog computing would win over digital. Now they have evolved into the ubiquitous answer to the question of small signal gain in almost any modern electronic design among all their other uses.

 

[AudioGal]

Indeed. On bike it's suicidal. Don't do that again.

Or while skiing.

 

[tincanear]

I did see that video, that is is what got me to finally start figuring out what is an opamp. It takes him forever to explain something, but you will never misunderstand for lack of detail and clarity. I'm still a little foggy on the seeming paradox of the virtual ground at the inverting input in an inverting configuration with the non-inverting input grounded...

I like how the black-box amp concept started as a component in analog computers when everyone was sure that analog computing would win over digital. Now they have evolved into the ubiquitous answer to the question of small signal gain in almost any modern electronic design among all their other uses.

regarding the virtual ground at the inverting input in inverting voltage gain mode here's how I see it:
if output is no infinitely high or low, then output must be proportional to difference between the inputs multiplied by the open-loop gain of the amp
for a case where gain is very very high, approaching infinity, then input difference is very very low, approaching zero. as the non-inverting input is a ground, then the inverting input must be infinitesimally close to the same potential as the non-inverting input at ground potential.

 

[Jason Stoddard]

I was referring to Jason's original statement ... that the UNISON firmware upgrade is indeed an issue with regards to the GMB ... but hopefully you are right about Intl. distributors being able to undertake this upgrade ... and happy to stand corrected.

I posed my question publicly as this may affect other internationally based owners considering this upgrade, when it becomes available. Hopefully Jason may make a further statement when he is in a position to do so.

Our distributors will be able to do the motherboard firmware upgrade required for Unison on Gungnir when the cards are available in January. They won't have to come back to the USA.

 

[darkarn]

Seems to me the USB hub is helping, or that problem decide not to show up so far

Meanwhile, I understand that there was a Modi Multibit firmware upgrade in mid 2018, what does that do exactly?

 

[RickB]

Seems to me the USB hub is helping, or that problem decide not to show up so far

Meanwhile, I understand that there was a Modi Multibit firmware upgrade in mid 2018, what does that do exactly?

Supposedly fixes a glitch that was inaudible anyway.

Many people (including me, who has owned two units of Modi Multibit with both firmware V1 and V2) think it leads to a marked improvement in sound quality.

I continue to be astonished at the sound quality I get with my "budget" system. No upgraditus for quite a while now.

 

[scottbrown5271]

regarding the virtual ground at the inverting input in inverting voltage gain mode here's how I see it:
if output is no infinitely high or low, then output must be proportional to difference between the inputs multiplied by the open-loop gain of the amp
for a case where gain is very very high, approaching infinity, then input difference is very very low, approaching zero. as the non-inverting input is a ground, then the inverting input must be infinitesimally close to the same potential as the non-inverting input at ground potential.

Thanks for that explanation. The more times I go over it, the more it makes some sense. The paradox I can't come to terms with just yet is the virtual ground that does not sink current, but measures at ground potential. Then the only remaining obvious sink is the feedback loop itself, so the input circuit is sinking its current directly into the output? Madness! This is way worse than trying to figure out a circlotron.

Regardless, I wrote my original post to point out that well designed opamp audio stages can sound really good, and there shouldn't be a stigma around them. Next point: Why is there a stigma about using the same, or equivalent, circuits for playback that were used to produce the music in the first place? Rupert Neve has used, and to this day is still using, transformers and opamps for most of his solid state designs. I'm not trying to start an argument, I've heard and read it all, over and over; I think all of us have. I would personally like to see some new blind testing between an opamp stage and a discrete stage to see how many of us can really tell the difference; the tube challenge was illuminating... Maybe just a simple, well designed buffer stage, preferably both versions designed and built for the test by the same person with the goal of making them both sound as good as possible.

Science fiction moment: Imagine a dystopian future where all electronic products are designed and made in Shenzhen. It wouldn't be good SF if it wasn't already partly true. iPhone, anyone? Try finding anyone who works directly for Apple that understands the comms radio software stack; That knowledge is extremely proprietary and reserved to Qualcomm and the like, as an example of where in the world all the IP is going.

I like the simplicity and elegance of a matched-pair jfet differential buffer, but the difference in cost is staggering. Try finding matched pairs of those fabulous old unobtanium LSK170-LSJ74 or equivalent for a reasonable price... It got so bad at one point that Nelson Pass finally added a trim pot to his J5 diy design that you adjust for minimum distortion using a signal generator and o-scope. My bet is that in ten years this will all be moot, the parts won't be available, and the ic solution will be all that is left (in fly-schiit size packages of course, with thanks to Mr. Stoddard for that adjective!). I propose that we all start getting over it now. And there will always be room for the (wealthy) purists in our hobby to spend as much as they want on whatever floats their boat, but I think the price of admission to that club is going to get even more exorbitant in the future.

 

[tincanear]

Jotunheim R + Raal SR1a or Mr Speakers Ether 2 + Asgard2 (or Mjolnir OG or Jotunheim or Ragnarok 2)? which combo would provide best instrument separation, soundstage depth?

 

[Torq]

Jotunheim R + Raal SR1a or Mr Speakers Ether 2 + Asgard2 (or Mjolnir OG or Jotunheim or Ragnarok 2)? which combo would provide best instrument separation, soundstage depth?

The Jotunheim R + SR1a.

It won't just have better depth, and separation, it'll have depth-wise imaging ... more like properly setup near-field speakers. There's literally no comparison on that front with any other headphones. The two closest competitors in that regard are the MySphere and the K1000.

 

[vladpetric]

May you make schiitloads of money from affordable gaming products as well!