Schiit Happened: The Story of the World's Most Improbable Start-Up
May 13, 2021 at 10:31 AM Post #76,561 of 145,766
Magni IEMagni has a nice addition, look at the volume knob...it's got a highlighted dot for volume

Puuleease, an HP-48SX. Nothing so plebian as a HP41C

Griff
hp42.jpg
 

Attachments

  • hp42a.jpg
    hp42a.jpg
    222.2 KB · Views: 0
May 13, 2021 at 10:31 AM Post #76,562 of 145,766
I have had these in for about a year, keep planning to change them up, has not happened. Works well with HiFiMan Sundara and Beyerdynamic T1.
6N23P Voskhod Rocket - 1971 - matched pair -
I will look for some of those. I have used the Reflector silver shield 6n23P's for a while but they are getting pricey. 1975 being the sought after year but I do not mind the 73's and 74's.
 
May 13, 2021 at 10:36 AM Post #76,563 of 145,766
I will look for some of those. I have used the Reflector silver shield 6n23P's for a while but they are getting pricey. 1975 being the sought after year but I do not mind the 73's and 74's.
sideways.jpg
 
May 13, 2021 at 10:38 AM Post #76,564 of 145,766
Ok I am confused, the Gungnir I have is full sized but it is a DAC. Are you talking about Mjolnir 2 perhaps? :)

Speaking of which, any tube favorites in the Mjolnir? I have tried some pairs but I always welcome the input of others.
I recently got a pair of 1958 Brimar CV4033 triple micas that had bases added them and used in conjunction with 12Axx to ECC88 adapters, they sound absolutely fantastic!
 
Last edited:
May 13, 2021 at 10:45 AM Post #76,565 of 145,766
+1 on the CV-4033s -- at least on my Valhalla 2. As much as I really, really like the CV4033, they have not been in my amp since I rolled in a pair of Raytheon Unline 7728s. These are by-far my new favorite in VH2 and have are in for the long-haul.
 
May 13, 2021 at 10:54 AM Post #76,566 of 145,766
2021, Chapter 7:
The Thunderdome Returneth



So now we have three Magnis.

Yeah. In case you missed it, we have a new model in the Magni line, called IEMagni. This one is black and gray, rather than silver and gray or black and red, like 3+ and Heresy.

“So it’s a different color? Who cares!” you say.

Oh yeah, and IEMagni also has three gain modes, including negative gain for IEMs, and advanced protection, like DC sense and mute, to protect those same IEMs.

“Ah, I get it, it’s an IEM-Magni!” you exclaim.

Well, not exactly. Because it also has 0dB gain and +15dB gain and it still has 2400mW RMS output into 32 ohms for hard-to-drive headphones, so it’s the most versatile Magni, ever.

“So it’s a do-all Magni?” you ask.

Again, not exactly. It has a lot of cool stuff we couldn’t put in Heresy, like the TI OPA1656 throughout (7 gain stages per channel, 1 for voltage gain and 6 for output, in a unique no-overall-feedback topology), plus a 4-layer board that even has DC sense and protect on it. It’s, in short, the Heresy we’d build if we had more budget. Even more heresy-er, as it says on the board.

“Aha, then it costs more,” you declare.

Yes. It does.

Brace yourself.

IEMagni is…$119.

“Um, that’s only $20 more than 3+ and Heresy,” you tell me, after consulting your trusty HP-41C calculator. “How are people gonna feel about that?”

And that’s when I chuckle and point up at the headline.


IEMagni-front-1920.jpg

Three Magnis Enter…

Is this the most interesting Thunderdome ever?

I think it is.

I think we don’t really know a lot about what will happen when you start tweaking things at the $99 price point. Does stepping over the two-figure limit mean sudden death? Or does it not matter, since you still have to factor in shipping and tax when you buy? Will it confuse the hell out of people and drive Magni sales to zero? Will it excite everyone so much that Magni sales triple? Will IEMagni kill Heresy, or vice-versa? Will all three find a place?

Now, if we were a normal company, we’d do research and put together focus groups and puzzle over the results and wring our hands and distill what we think is the meaning from the nuance and then we’d introduce a replacement product that was still a WAG and might fall on its face and then we’d lose market share and have to scramble to replace it.

Dare: if you are in a business making consumer products, read that and tell me it ain’t true.

Because that is how it works. You research and ask and project…but when the rubber hits the road, all your beloved intelligence may be worth squat.

So. Thunderdomes.

That’s how we do it. Introduce a new product, see how it does with real data, adjust accordingly.

And that’s why this is so interesting. We can learn a lot from a Magni Thunderdome. All the stuff we’ve been asking what-if, all the guesses we’re making…they can be confirmed or denied instantly. Or we might learn something completely unexpected.

“Oh hell, does that mean you’re gonna maybe kill my favorite Magni?” you ask.

Short answer: doubtful, but I don’t know.

Longer answer: we’ll see how things work out. I personally think 3+ is safe, because that’s “my” Magni, as in, it’s fully discrete and a lot more interesting than IC-based amps. However, adding a negative gain mode to 3+ is more problematic and less functional…as in, it wouldn’t significantly reduce the noise floor, and it would result in an input impedance that could go as low as 2kOhms or so…but I’m really getting ahead of myself. I think we really need to see how it goes. Because, despite my pledge not to raise prices just because we can, the specter of price increases is looming—steel now costs 4x what it did a few years ago, many suppliers are raising component prices, heck, even labor is going up significantly. I expect you’ll see some small bumps on more expensive products before the end of the year, which really won’t have that much effect on sales. But when you talk about the inexpensive products, it’s a finer line. The question becomes, “Should we hold the line at all costs, or should we go for more features or performance to make a modest increase more palatable?”

“But what if you drop IEMagni after a run or two?” you ask. “What then?”

What then indeed? It’s not like we won’t support it, or repair it, if the need ever arises. You’ll be stuck with the most versatile Magni ever built.

“Well, what if this kills Heresy?” you continue to press.

Then you get an amp that performs even better and is even more versatile for a bit more money (and, if I am to be immodest, one that still costs less than amps built in—ahem—places with considerable cost and labor advantages.)

“Well what if you come out with a Magni 4 that’s like $79 and is even better than this, maybe even a discrete one?” you posit.

LOLLOLLLLLLLLOLOLL! Yeah, that ain’t happening. Notice that we increased the price of this one to cover better parts and more complexity? Any future Magni that performs even better than this one, or has more versatility than this one, and is discrete, will likely cost more. And that’s when we need to decide how important that $99 price point is. The main point is, that’s not something we can decide on now.

Because we need more info.

Because we need to let this Thunderdome play out.


Why Always More

Great question, actually. Why always more? Why aren’t we taught to say, “enough,” at some point in our lives? I mean, no matter your wallet, and no matter your interests, there will always be something beyond your means. Oh, you can afford a $15,000 pool table? Well, what about the auto-leveling $300,000 model for your superyacht? Oh and your yacht is just a $15,000,000 model? Well hell ya ain’t playin if it costs less than nine figures! You have to learn to say “enough,” even if only to preserve your own sanity.

“I was talking about the third Magni model, bub,” you tell me.

Aha. Well, ignore my wandering thoughts and proceed on your way.

“As in, I don’t know why you need yet a third Magni,” you continue. “The old ones seemed just fine.”

Oh. Aha. Gotcha. Yeah, I hear you. I personally think both Magni 3+ and Heresy are absolutely fine for me. But I use full-size headphones. We keep some IEMs around for testing, but we aren’t exactly IEM hounds.

Which is the rub. Some IEMs are sensitive.

Some are very sensitive.

Some are insanely sensitive.

Hook up some of the insanely sensitive ones to Magni 3+, connect it to a source with standard 2V RMS line out, and play some modern (dynamically range compressed) music, and you might find yourself with not enough usable volume control range. Crack the volume, and blast yourself out of the chair, even at low gain. Or the residual hiss (background noise) might be audible—not an ideal situation.

And it gets worse if you have a “hot” source (some DACs provide 3V RMS or higher from their single-ended outputs, as a way to fool unwary listeners into thinking they sound better—the louder source in an A/B test usually sounds better).
Aside: paradoxically, people who listen to older music masters, or classical, may not have a problem at all—I have tons of CDs from the 80s and 90s that are subjectively 10dB quieter than a typical current-generation track. This is why we get lambasted for providing gain in preamps, and not providing gain in preamps—because people listen to music from different eras of different types at different levels, and it’s not always a “hurr durr, 2V is more than enough to drive any speaker amp into clipping, so there’s no need for gain” type of thing.

Anyway, back to IEMagni. IEMagni was intended to solve “the IEM problem,” and give us an amp that would be totally comfortable driving the most super-sensitive IEMs out there. That’s “why always more” in this case.

“Wouldn’t it be best just to add that feature to all Magnis?” you persist.

Oh yeah, if features were free and more complexity was free and better parts were free, it would be best just to just add it on. But they aren’t. So IEMagni is more expensive. A bit. And we’ll see how that goes.

Or, if I was addressing a corporate board, I’d say, “It’s time to say goodbye to the traditional Magni 2-year product cycle…welcome to wildcards, new ideas, and continuous evolution that doesn’t follow a schedule.”

But we don’t have a corporate board. So I won’t say that.

But we’ll be closely watching how this one goes.


New Places, New Ideas

Here’s a fun fact: IEMagni is the first amp that was conceived, designed, and built in Corpus Christi, Texas.

“I thought you were already building amps in Texas,” continues my invisible foil.

Yes. We’ve been building Magnis and Modis there for a couple of months now, and we’re steadily moving models over from Valencia (the goal is all small and small-medium products by June).

But IEMagni is the first amp I first conceived of in Texas.

“Why does place matter?” you ask.

Hell, I don’t know. I can’t answer that with 100% certainty. But it does matter to some people, and I do think there’s an x-factor involved. Hence, “new places, new ideas.”

I never really considered an IEM-dedicated amp until one of our early trips to Texas—as in, we’re talking before we bought our house, hell, before we bought our business facility. Lisa and Tyler and I were out, most likely looking at candidate locations for Schiit in Texas, most likely at a brewery, and Tyler started talking about how people were using IEMs with desktop amps.

I thought he was a bit crazy, but I did say, “Well, heck, we could do a really good IEM amp. It wouldn’t even have to be big. Heck, it could just be a superbuffer. It doesn’t need gain.”

“It actually needs negative gain,” Tyler said. “0dB isn’t low enough.”

“Really?” I said, suspicious.

Long story short, Tyler showed me that yeah, people were using IEMs with desktop amps, and that there were some IEMs that were so insanely sensitive that they actually needed negative gain (AKA, a divider—as in, the output of the amp is lower than the input, but a passive divider would be problematic because multi-driver IEMs also have crossovers that would be affected by a high output impedance…but again, I’m getting ahead of myself.)

“So we could do it?” Tyler asked.

“We could do it in a whole bunch of different ways,” I told him. “The big problem is that you need a divider in it somewhere. Which means two superbuffers, if we do a buffer topology, and then you don’t have gain if you want to use the amp for other things.”

“You can’t do a negative gain amplifier?” Tyler had been learning about engineering stuff. He even bought Bob Cordell’s book.

“Ish,” I said. “Sure, you can do an inverting negative gain amplifier, as long as you are using a topology that’s stable at less than unity gain.”

Aside: a negative gain amplifier can be inverting or non-inverting. Negative gain refers to having a gain of less than 1, or less than 0dB. An amplifier with a negative gain (or loss) of -10dB gives about 1/3 the voltage out as what goes in.

Aside to the aside: if this sounds like a resistor divider, it can be. Other than the output impedance problem.

“What’s wrong with an inverting amplifier?” Tyler asked.

"Nothing, other than you need another inverting amplifier to make it non-inverting. That’s super-complex if you’re talking discrete, and if you’re talking current feedback, it’s also a very low-input-impedance amp, because you’re driving the negative node.”

Tyler waved his hands. “Okay, okay, you’ve exceeded my engineeringese. But you’re saying we could do it?”

“Yeah, in lots of ways,” I said. “The problem is finding the best one.”

Aside: by now, you may have realized that I’ve been talking about a dedicated IEM amplifier, with IEMagni is not. The concept had to evolve before we got to the current IEMagni.

Still, I was intrigued. A super low-noise, low-output-impedance, negative gain amplifier might be kinda fun. That was an interesting new challenge.

So I sketched up some ideas, usually based on various buffer topologies. But the fact that we were leaving a whole raft of headphones behind—inefficient planars and high-impedance designs—continued to bother me.

I mean, I told myself, we could just put a divider in front of a Magni 3+ and have a negative gain amp. Add another switch, we could have three gain levels and cover all kinds of headphones. Except that was a half-answer. We’d have to accept a low input impedance at some volume pot positions, and, worse, it wouldn’t reduce the inherent noise of the amplifier stage. It would divide any source noise down, but it wouldn’t change the Magni 3+’s gain stage noise. Magni 3+ was very quiet, but we were shooting for extremely quiet.

To get where we needed, we’d need to divide down after the Magni 3+’s gain stage, and add a buffer stage. But that would mean the 3+ gain stage would shrink, and the output buffer would get pretty big (and linearizing an open-loop buffer is fuuuuuunnnn…even our two-stage “superbuffer” would be hard-pressed with a 32 ohm load). What we really needed was something with a simple gain stage up front, and a stout buffer on the other side…

…aaaaahhhhh, like Magni Heresy! I realized.

And just like that, the whole thing just fell into place.

You see, Magni Heresy has a unique topology amongst measurement-focused amps: it uses a single IC gain stage for voltage gain, and multiple paralleled IC output stages for buffers, with no overall feedback loop. It’s less focused on “numbers at all costs via 237dB of feedback” and more focused on what sounds good to us (yeah, we’re crazy, we know, whatever).

And that unique topology meant that all we needed to do was to add a divider between the input voltage gain and output current gain stages, and boom, negative gain amp! Not only that, the divider would then reduce the inherent noise of the voltage gain stage, resulting in a stunningly quiet amp.

What’s more, we could use a three-position switch for both the feedback and divider switching, giving us three gain levels, including negative gain, so we could use the same Magni metal.

That was actually getting exciting.


Special Care and Feeding

“So why didn’t you just do that,” you say. “Sounds cheap and easy.”

Sure does. And we could have done that. Except for one inconvenient fact: IEMs require even more special care and feeding, so if you’re doing an IEM amp, you should really do an IEM amp right.

Which, in the case of IEMagni, meant adding DC sense and protect, so that the amp lifts the output in the presence of excessive DC output. This is especially critical for IEM amps, because they are just insanely sensitive. We didn’t want to take any chances, so we added the extra circuitry found in amps like Lyr 3 and above.

And it also meant moving to a new gain and output IC. The TI OPA1656 provides even higher performance and lower noise than the OPA1688 found in Heresy…but it also costs a bit more. We even added a thermal shutdown for hard-to-drive headphones.

So all of this stuff adds up to an amp that costs a bit more…so, yes, it costs a bit more. Like I said, we’ll see how it goes.

“So how many prototypes did you do, and did they catch on fire?” you ask.

Ah, no. This was a boring development cycle. The first prototype worked fine, as you’d expect for something based on a standard board size and chassis form factor—well, except the thermal protection worked, ah, exactly one time.

And then the amp didn’t come out of mute. At all. Ever again.

Turns out the thermostats we were using (really cool things, not thermal sensors but actual IC thermostats that don’t need analog or digital computing to tell when you’ve exceeded their limits) could only tolerate a small amount of current when they triggered. We added a resistor to limit the current that passed through the device when it triggered, and then it worked fine. We tortured it with a hot air gun all day to make sure it was fine.

Beyond that, it was just “order boards, get parts, run ‘em.”

Yeah. Boring.

Which is usually how you want it to go.


Beyond Thunderdome

“So what does the future hold?” my imaginary foil asks. “Are you gonna run just one run of these and then leave everyone hanging? Are you gonna go out of stock right away?”

Well, most likely no and yes. If we run out of stock right away, that’s a good sign that IEMagni will be around for a while. If we miscalculated and these things sit around on the shelf, it’s easy enough to wait around till they’re gone, then move on with the traditional two Magnis.

“Well for a while it was the traditional one Magni, can we get back to that?” you ask.

Great question. And I don’t know.

I mean, if you want it all—an all-discrete Magni with great measurements, low noise floor, plus three gain levels including negative gain…oooh, that’s a very complex device. As in, it might not even fit on the current board. It is certainly not a $99 product.

If you go somewhere in-between, what do you give up? The discretosity? I mean, that’s our hallmark. The three gain levels? The measurements?

In short, I don’t know.

I don't know a lot of things.

But I do know one thing: when this Thunderdome concludes, we’ll know a lot more.
@Jason Stoddard
''But I do know one thing: when this Thunderdome concludes, we’ll know a lot more.''

You'll know a lot more about colour preferences afterwards.

PS
Check the difference between your line up in 2014 and now. My advice: keep it slim as a Freya S :wink:
 
May 13, 2021 at 11:00 AM Post #76,567 of 145,766
Yea, people who "can't handle" the "f-word" are welcome to drop by for a beer so I can help desensitize them.

What kind of BBQ will you be serving? :beerchug:
 
May 13, 2021 at 11:05 AM Post #76,569 of 145,766
Magni IEMagni has a nice addition, look at the volume knob...it's got a highlighted dot for volume

Puuleease, an HP-48SX. Nothing so plebian as a HP41C

Griff
Screen Shot 2021-05-13 at 10.59.48 AM.jpg
Haha, my college calculator... no one noticed (or thought it was super nerdy) but every once in a bluemoon someone would walk by, and say "Excalibur" (or similar). Really like RPN!
 
Last edited:
May 13, 2021 at 11:32 AM Post #76,570 of 145,766
Screen Shot 2021-05-13 at 10.59.48 AM.jpg
Haha, my college calculator... no one noticed (or thought it was super nerdy) but every once in a bluemoon someone would walk by, and say "Excalibur" (or similar). Really like RPN!
Poindexter.
 
May 13, 2021 at 11:43 AM Post #76,571 of 145,766
Magni IEMagni has a nice addition, look at the volume knob...it's got a highlighted dot for volume

Puuleease, an HP-48SX. Nothing so plebian as a HP41C

Griff
I used an HP-71, but nobody knows what those are anymore.
 
Schiit Audio Stay updated on Schiit Audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/Schiit/ http://www.schiit.com/
May 13, 2021 at 11:49 AM Post #76,573 of 145,766
@Jason Stoddard

Check the difference between your line up in 2014 and now. My advice: keep it slim as a Freya S :wink:

As I've said before, there is zero evidence of customer "analysis paralysis" with our current line--most people want more options. Like Freya S. Also some people want Camaros, but they're not making those at all, so Chevy can make more trucks.

This is a weird year--the only certainty is we'll have what we have when we have it. There will be new stuff, there will be some big wacky interesting stuff, there will be some more experiments...and we'll see what happens next year. If supply chains start normalizing, then we'll see where we end up. Certainly with some streamlining, but maybe less than you think.
 
Schiit Audio Stay updated on Schiit Audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/Schiit/ http://www.schiit.com/
May 13, 2021 at 11:51 AM Post #76,575 of 145,766
Schiit Audio Stay updated on Schiit Audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/Schiit/ http://www.schiit.com/

Users who are viewing this thread

Back
Top