Schiit Happened: The Story of the World's Most Improbable Start-Up
Jul 14, 2021 at 1:29 PM Post #79,397 of 148,519
O....M...G! The wait is over!!
Not for me. After reading Chapter 10 I now want Lokius Unobtaiablus:
True balanced
Parametric adjustment
Six bands
Remote control
Balance control ('cause it isn't elsewhere)
On-board power supply
[edit] Power switch on the front panel
[edit] non-blinding power indicating LED
In black
:)
 
Last edited:
Jul 14, 2021 at 1:46 PM Post #79,398 of 148,519
Not for me. After reading Chapter 10 I now want Lokius Unobtaiablus:
True balanced
Parametric adjustment
Six bands
Remote control
Balance control ('cause it isn't elsewhere)
On-board power supply
In black
:)
:scream: Haha We can dream.
 
Jul 14, 2021 at 1:56 PM Post #79,399 of 148,519
rokinz.jpg

So the Lokius is Rokinz huh? Totally intentional. :p
 
Jul 14, 2021 at 2:08 PM Post #79,401 of 148,519
Not for me. After reading Chapter 10 I now want Lokius Unobtaiablus:
True balanced
Parametric adjustment
Six bands
Remote control
Balance control ('cause it isn't elsewhere)
On-board power supply
In black
:)

It’s on a different timeline. And has a power switch on front.
 
Jul 14, 2021 at 2:35 PM Post #79,403 of 148,519
  • Matches the size of Modius and Magnius, and the footprint of Asgard, Jotunheim, and Lyr. This means it looks good in a stack—though be careful when stacking on top of components with internal transformers like Asgard, Jotunheim, and Lyr.

So hypothetically, if I stack my Valhalla on top of Lokius I shouldn't run into problems?
 
Jul 14, 2021 at 3:17 PM Post #79,405 of 148,519
Will schiit release more stuff tomorrow 15th?
 
Last edited:
Jul 14, 2021 at 3:24 PM Post #79,406 of 148,519
2021, Chapter 10
A Trickster You Didn’t Expect

Ha!

That’s all I can say.

You thought you knew what we were doing. Hell, I’ve been talking about Loki Max for a while now. I’ve shared a photo of a prototype board, I’ve said that it’ll be Freya-sized, I’ve told you it will be totally insane, with relay potentiometers, 67 relays, and remote control.

Except this ain’t Loki Max. It’s Lokius.

As in, Loki Maximus.

As in, it’s exactly the same size as with Magnius and Modius, and it’s also sized to stack with any of our “6x9” products (Asgard, Jotunheim, Lyr, Saga, etc).

And, as in, it has both balanced and single-ended inputs and outputs, so you can use it with a whole lot of different stuff that we make. And, with two more bands than Loki Mini, it’s offers a lot more options to control how your system sounds.

Aside: yeah, I know, I know, some people here undoubtedly think equalization is the devil, having absorbed far too much high-end propaganda from the 1980s and 1990s. We have a chapter or two for you on that. Have fun reading those if you want to dismiss equalization out of hand.

Aside to the aside: and yeah, I know, digital parametric eq, software, infinite control, draw curves on a screen, others are saying. Yeah, that’s an option too. This is an option for people who want to twist some knobs to compensate instantly for bad recordings or imperfect systems. There’s no need to fight. The existence of Loki Mini and Lokius in no way will result in the elimination of software EQ options. They are not alien pods or Borg cubes. There’s nothing to fear. No need to nuke the site from orbit. Cool?

I know some of you are confused. I mean, we have Loki Mini. Now we have Lokius. We’ve promised a Loki Max. How does all this work? Why are the names related-ish? How do we expect people to choose?

Aaanndddd…where the hell did Lokius come from, anyway?

lokius black front 1920.jpg

Loki Family Geneaology

Let’s get this out of the way first: Loki was intended to be a family of products from the start. That’s why we started with “Loki Mini.” Because we were also playing around with more complicated designs, and expected to shortly introduce a Loki (a 5-band remote EQ, Jotunheim sized) and Loki Max (a 6-band relay-based remote EQ, Freya sized).

But there was trouble in paradise.

Loki never really worked right, mainly due to size constraints—motorized potentiometers, inductors, and a power transformer don’t interact well, even when shielded, if you’re talking something the size of Jotunheim. And Loki Max had huge operational problems that were so intractable that it took several revs before we ever got it to function from input to output.

Because of these troubles, I figured, “Well, heck, maybe we just do a Loki Mini.”

But Loki Mini was popular…and it kept getting more popular…and people kept asking for more capable EQs…

…and so I kept trying.

First, I revived the Loki Max and actually worked through its huge problems. More on that when it’s introduced. But Loki Max is like 10x the price of Loki Mini+. It’s not for everyone. It takes a lot of crazy to do a real LC equalizer with relay potentiometers (NOT the same thing as relay volume ladders, not at all), Nexus, and remote control. So it won’t be cheap.

And reviving the previous mid-sized Loki wasn’t really in the cards. The motorized pots made it expensive. The transformer wouldn’t be able to be inside the chassis, either, so it would be the first Jotunheim-sized product that used a wall-wart. There was limited space for custom inductors, and the inductor cost also went crazy on us…which meant that we might end up with a $800 product that used a wall-wart, and still had two inductors that weren’t exactly ideal due to lack of space.

However, if I threw out the idea of remote control, and used gyrators for the two lowest frequency bands, we could conceivably do an affordable 6-band equalizer with balanced and SE I/O, especially if it shared the form factor of Magnius/Modius…

…and that’s how the idea for Lokius was born.


A Hard Truth

Some of you guys are probably super excited to see balanced inputs and outputs on Lokius, because you can finally achieve the One True and Perfect Balanced Path Through All Components.

Well, yeah, but no.

As in, yes, Lokius has balanced inputs and outputs, but internal processing is single-ended.

Oh gawd, cue the wailing. Cue the crying.

Well, relax. Because:
  1. It really doesn’t matter that much
  2. Balanced LC equalizers will cost a lot more
  3. Balanced LC equalizers won’t fit in a Lokius chassis
  4. Chasing balanced is sometimes really stupid, and this is one of those times
Here’s the thing: a true balanced LC or gyrator-capacitor EQ like this just ain’t gonna happen. It’s so different that literally every part is different—potentiometers, capacitors, inductors, gyrators, etc.

And yeah, I’m sure some of you are thinking you’ll just wait for Loki Max, that will do it, right?

Well, also again, no.

A true balanced relay-potentiometer LC EQ would be at least 2X the size of Loki Max, and would use something like 134 relays. With all the additional testing required, and a new chassis form factor, it might be well over 2x the cost, as well. This is getting mega-stupid.

Aside: but more on that later, when we intro Loki Max. Suffice to say, I have no problem putting it up against any equalizer from today or yesterday, even if it is converting from and to balanced.

And that’s why we concentrated on making Lokius a real step up from Loki Mini, rather than chasing a balanced internal signal path. Things like a much higher-current, lower-distortion discrete gain stage, a two-stage complementary “superbuffer,” and quality parts used throughout, including film capacitors even in the gyrator stage, as well as custom inductors. Add a super low-noise power supply based on a custom two-stage topology, and you have a truly insane equalizer that doesn’t need to make any excuses.

What’s more, adding two bands to Lokius allows us to offer more precise, better placed controls. The 6 bands we’ve chosen are very similar to the much-celebrated Cello Audio Palette equalizer, in fact, and very similar to Loki Max.

Now, I know some people are thinking that there are equalizers with more bands, and, yes, there are. Those are typically going to be constant-Q equalizers, implemented with multiple gain stages (op-amps), rather than a single discrete gain stage and LC filtering, like Lokius.

And I know there are others saying they like parametric EQs, and that’s fine, too, but again, it’s a different approach. Parametric also uses multiple gain stages (op-amp), unlike our more “purist” or “old-skool” approach. Lokius won’t be for everyone—but neither will software, constant-Q, or parametric equalizers be for everyone.


Other Development Notes

Lokius’ development was actually fairly straightforward. I really hate it when it works like that, because it usually means I’ll get boned bad on something else coming up. But hey. Take the win, right?

I really only hit snags in two areas:

1. Buffer performance. Getting a buffer to drive a reactive load (and, by “reactive load,” let’s call it what it is—a metric asston of inductors and capacitors, plus the relatively low impedance of the inverting input terminal of a current-feedback amp) is not trivial. Sure, you can use an op-amp with a billion dB of feedback around it, but that’s pretty boring, especially when you’re using a discrete gain stage. Because of this, I played with a bunch of different buffers:
a. Your typical buffer—say, a diamond buffer—falls to its knees and whimpers in pain. Well, not literally. But they don’t work well.​
b. Some buffer topologies have very poor DC output control, which is exactly what you don’t want to be feeding into a gain stage, even if it has a servo​
c. A lot of buffers are fine until loaded with 600 ohms or so, then they fall apart​
d. And yeah, I tried tons of stuff, from Broskie’s triadtron stuff to enhanced diamond buffers to error correction​
e. What I settled on was actually a combination of two buffers, which have DC offset problems in the opposite direction, so they are essentially DC correct, and, when used together, they are both low-noise and load-invariant. I don’t really feel like naming this “superbuffer,” so I’ve just been calling it “superbuffer.”​
2. Power supply performance. One drawback of discrete current-feedback stages is that the power supply noise gets through to the output much more than with, say, an op-amp.
a. Throwing a simple LM317/337 pair on the power supply and calling it done wasn’t going to be good enough​
b. We’d already improved Loki Mini with filtering after the LM317/337, so I decided to experiment a bit with other, more modern parts​
c. We eventually settled on a new TPS part from TI that does both positive and negative voltages at once, and offers much better ripple rejection, at the cost of still being fairly wideband-noisy​
d. So I added the filter from Loki Mini afterwards, and boom—our quietest power supply to date.​

Yeah. Kinda boring. Two prototypes, and we were good to go. It doesn’t hurt that Lokius uses the same bottom chassis as Magnius (exactly, down to the screen), so we only had to get specific top chassis.

What can go wrong, right?

LOL. Of course, the single glitch we had was related to the tops. When we received first articles of the Lokius boards (way back in March!), we tested them, made sure they made the same specs as the prototypes, and then put them in chassis…

…and instantly realized the potentiometers weren’t poking through the front holes far enough to use!

Yeah.

This is what happens when you don’t do a 3D print of the top chassis. Argh. We really need a larger 3D printer.

So what did I do?

Simple. I moved the potentiometer on the board, re-ordered a new run, and I told the PCB assemblers to throw away all the current boards.

Yep. Boom. Gone. All boards wasted, because a potentiometer was 0.050” off.

So maybe the development wasn’t so simple. After all, one mistake put us months back. But in the craziness of today, where parts are frequently delayed or unavailable, this is really what you’d call smooth sailing.


So How Do I Choose?

Now we have two equalizers. And soon, three. Furthermore, they all share a family name. How do you choose?

Well, it’s easy, especially if you don’t believe in EQ. If you don’t believe in EQ, you choose none of the above, and that’s the end of that.

But if you’re sold on wanting one of our single-gain-stage, LC-style equalizers, then it’s also pretty easy:
  • Loki Mini: least expensive, matches Magni/Modi sized products, 4 bands—easy to try and see if you like it!
  • Lokius: still pretty affordable, matches Magnius/Modius/Jotunheim/Saga sized products, 6 bands, balanced I/O means you can plug into anything we make
  • Loki Max: blindingly expensive (for us), offers features not found on any other equalizer, a true shot at the best of the best, matches Freya/Urd/Gungnir sized products, balanced I/O means you can plug into anything we make.
Loki Mini and Lokius are desktop friendly, Loki Max is more rack friendly. Loki Mini and Lokius are priced at $149 and $299, both in the sane realm. Loki Max will be $1499 or more.

And there you have it. Beyond that, it's perfectly fine if you don't want an EQ at all, or a different type of EQ, like a parametric, or constant-Q, or software. As I said in the last chapter, there's no "T" in engineering!
Thanks for the article and the "how to choose" section.
Gotta wait for the Loki Max in that case.
 
Jul 14, 2021 at 4:04 PM Post #79,409 of 148,519
Maybe I spoke too soon...

Bottom of the stack is where it will go, I guess. I've replaced the feet on BF2 so it has additional spacing for ventilation -- hopefully enough.
 
Jul 14, 2021 at 4:11 PM Post #79,410 of 148,519
Will this pair well with a Rag2?
Lokius will work if placed between an analog source (e.g. Mani, Modi, or Bifrost) and one of the Ragnarok 2 analog (RCA or XLR) inputs. Won't work if using internal DAC (dual 4490 or multibit) module via a USB source connection (there is no analog tape / effects / processor loop) on Rag 2.
 
Last edited:

Users who are viewing this thread

Back
Top