2021, Chapter 3
Even Trickier
The new Loki Mini+ is just like Loki Mini, but quieter and lower distortion. That’s it. The end.
No, really.
No, totally totally serious. I mean, to create the Loki Mini+ manual, I opened the Loki Mini manual in Indesign, added a “+” after Loki Mini, and saved it again. Done.
Of course, I also had to completely re-measure Loki Mini+, and I did re-write the copy a bit.
And of course, I had to do the press release and press letter.
Oh yeah, and it is a completely different PC board, with a lot of tricks stolen from Magni 3+…
…so, hmm. Maybe there is a story in here.
Let’s look a bit deeper.
Starting With "Why a New Loki Mini?"
Since its introduction, Loki Mini has become a really popular product—so popular, it’s one of the ones we try like heck to keep in stock.
Aside: Loki Mini has been in backorder a while, so if you’re waiting for one, you get a Loki Mini+ instead! Free upgrade!
This popularity is a bit weird, though, because you don’t really hear much about Loki Mini. You hear about Magni and Modi and Hel and Fulla, heck, even Mani and Vali. But not so much about Loki Mini. It seems that a whole lot of people are out there, simply enjoying what Loki Mini does…despite the lack of chatter, despite the dearth of reviews.
And this is one reason why we decided to take a look at it again, and see if we could make it better.
Another reason we decided to take a look at it was that it was, well, a bit of a pain in the rear end to build. Loki Mini used a “girdle” to help align its 4 knobs. This was slow and fiddly to assemble, and even then sometimes the alignment wasn’t perfect. We’d changed Magni and Vali to use “pocketed knobs,” for better alignment, so it would be ideal to take it across to Loki Mini.
And Magni 3+ provided us with the final reason: the driver stage improved performance significantly. That, together with a new two-stage power supply we were using on products as diverse as Aegir and Saga S, meant we could deliver a much lower noise floor…and lower distortion, too.
Together with Magni 3+’s matched, paired parts, that also meant that even with higher performance, Loki Mini+ would use fewer parts.
So, simpler product, higher performance, easier to assemble…win, win, win.
Why Do I Need an Equalizer, Pop?
Yeah, cool. Call me old. I don’t care. I’m Gen X, not a boomer.
And I get it…equalizers have quite a stench of hair gel and fake bark, the dying gasp of the go-go neon ‘80s, expressed in the number of LED-lit sliders they had.
But here’s the deal: they had a point.
- No system is perfect.
- No room is perfect.
- No recording is perfect.
Seriously. Review the above. Can you dispute any statement?
Is your system perfect? Do your speakers or headphones have perfect frequency response? If you say “yes,” please pause while I laugh.
Is your room perfect? Does it have asymmetrical walls and ceilings to reduce room modes? Do you have thousands of dollars in acoustic treatment? Applied by someone who knows acoustics? Agreed to by a consensus of industry experts that it is “perfect?” Again, yeah. Thought so.
Are your recordings perfect? All of them? Including the ping-pong stereo stuff from the 60s? The supahdry no-bass recordings from the 80s? The 0.5-db dynamic range stuff from the 00s? The overproduced and over-sonic-maximized-and-computer-harmonized stuff of today? Uh-huh. Right.
Here’s the thing. An equalizer won’t solve all ills. But it can help.
Especially if it’s a quiet, good-sounding, high-performance equalizer like Loki Mini+.
Want more reasons? Take a look at the
original Loki Mini chapter.
Another Boring Development Story
Loki Mini+ is not scintillating material when it comes to development. Once I got the idea to improve the product, it really came down to how many things I was going to tweak:
- Add the driver stage from Magni 3+?
- Do the matched parts to reduce parts count?
- Add the two-stage power supply for lower noise?
- Play around with the bands a bit?
- Do a top to eliminate the girdle?
In the end, I decided on “all of the above.” I laid out a PC board, ran a prototype, and it worked just fine. Other than a rev to re-position the potentiometers so they worked better with the non-girdle top, that was it.
Aside: the non-girdle top was a thing I thought about rolling in on Loki Mini. We actually got prototypes of the pocketed front. Then we ordered parts with the pocketed front. When I realized that the current boards wouldn’t work with a pocketed front, we shelved the idea until we had time to do a proper Loki Mini+.
I say “well, that was it,” except for a bit of tweaking.
Loki Mini, you see, was a 4-band equalizer. Which is not a ton of bands. We chose them based on where we thought they’d be most useful, which was arrived at by a lot of listening to music. The broadness or narrowness of some of the bands was a bit up for debate, though, so we decided to have a look at the bands again with Loki Mini+.
The result? We didn’t change the center frequencies significantly, but we did tweak the “Q” (broadness) of some of the bands, and also went to one larger inductor to enhance performance.
Aside: Loki Mini+, like Loki Mini, remains a single-discrete-gain-stage with LC (inductor-capacitor) filtering for 3 of the 4 bands, and gyrator-capacitor filtering for the 20Hz band. This means it is super old school. Some recording engineers may be very excited about this. Some people who are 100% into measurements may not be very impressed. It also means, however, like all products that use inductors, it may pick up hum from other components that have, say, giant transformers inside. Stick a Loki Mini+ on top of a Vidar and it will probably hum.
Aside to the aside: a single-gain-stage variable-Q LC equalizer is one approach amongst many, as I noted in the original Loki Mini chapter. I understand if you prefer constant-Q or parametric or software EQ. It’s just not Loki Mini+.
With these tweaks, we have an equalizer that is super-quiet (literally 10-15dB more than Loki Mini) and has lower distortion, thanks to the added driver in the current-feedback gain stage. Not a bad result—but, even more importantly, we think it also sounds better. (But then again, we may be crazy. I really need to do a chapter on, “How We Fool Ourselves.”)
So is that it? Just a better product for the same price, as usual around here?
Nah.
The Coming Loki(s?)
When we introduced the original Loki Mini, I hinted that there may end up being a whole line of Lokis. And, when I said that, we were deep into development on two additional Lokis. Both had remote control, and one was an all-out assault on the state of the art in equalizers (well, LC equalizers, anyway.)
But some funny things happened on the way to a Loki lineup.
- The mid-sized Loki, called, unimaginatively, “Loki,” was too noisy to go into production. Big inductors, a bunch of remote control potentiometer motors, and a power supply transformer didn’t coexist well in a desktop-sized chassis. Shielding the transformer didn’t help. Moving the inductors all the way to the other side of the chassis didn’t help. Plus, the size restriction meant that two of the inductors weren’t all that great, even with fancy 80% nickel cores. Plus, Loki was pretty damn expensive, once you got all those pots, inductors, knobs, and etc accounted for.
- The big Loki, which went by the unfortunate moniker of “Loki Max,” never worked right. First, the gain stage didn’t work well enough. Second, its ambitious relay-ladder potentiometers (NOT a relay-ladder volume control…very different) didn’t work 100%, either. Heck, even the crazy two-stage complementary megabuffers decided to go up in smoke from time to time. Despite the insane complexity of the boards, I built up four separate prototypes, none of which succeeded in passing signal.
So. Yeah. “Line.” LOL.
Well, we’re still a ways away from a lineup, but I did share a photo of what we’re calling “Loki Max,” or “Lokiest.” (We still don’t quite know what to call it…but don’t worry, that’s easy enough to change.)
Better yet, it is working.
Once the development was done on Loki Mini+, I decided to take a look at Loki Max again. And the difference between then and now was laughably extreme. The Max was so old, it used Pivot Point, not Nexus. The non-paired devices we used in the superbuffer could easily be replaced by modern parts used in any Magni 3+…parts that were much more rugged than their predecessors. The relay potentiometers didn’t work because I screwed up the truth table, and forgot I had to lose one step to make them work (seriously, this is completely nuts stuff…there are no guides out there as to how to do it.) Heck, even the shift-register/relay-driver arrangement we were using could be replaced with power logic, substantially simplifying the product.
So I did one more board…and this one, with some hacks, actually worked a bit.
It still wasn’t there, though. Because when you’re talking 4VRMS balanced signals that you can apply 15-16dB of boost to, you need a ton of headroom. Especially with Nexus, which almost requires a stacked supply in order to swing anywhere near the output rails.
So we needed another prototype. This one used stacked rails, up to +/-32V on the voltage gain stage.
Aside: Yes, 64V total. Yes, in a line-level product.
This prototype worked. Of course, it still had some hilarious issues, such as not having enough buttons for all of its functions (not kidding) and needing more regulation on the +5V rail (driving up to 67 relays takes a bit of current).
And that’s where we are today. One more proto to clean up things a bit, and make sure everything works, and we’re ready to roll.
Loki Max is completely insane. It’s really designed to take on the legendary equalizers of the past, up to and including the Cello Audio Palette, which cost $25,000…in 1989. It’s our shot at doing an equalizer that is comfortably part of the highest-end systems…for, say, about 10X Loki Mini’s cost.
“Cello Audio Palette,” some people will cry, aghast. “How can you possibly compare with something so storied, so pedigreed, so expensive?”
Cool. Yeah. Let’s see.
- Audio Palette used custom-made switches for their settings, allowing a precise center. We are using custom-designed, microprocessor-controlled relay potentiometers for our settings, allowing a precise center.
- Palette wins this one maybe because each channel could be individually adjusted.
- Or maybe we win this one because we have remote control with presets.
- Palette used all-discrete amp stages (at least the early ones), but I’m not sure if it’s constant Q or if its passive LC-filtered. We use all-discrete amp stages throughout, with passive LC filtering, including heroic 1.5 and 0.5 H 80% nickel-core inductors (yes, that’s Henries, and yes, those are completely custom.)
- Palette wins this one maybe if you like nonconstant Q? Not sure, though.
- We win this one if you like old skool LC
- Palette is the size of a Ford F-250, ours is the size of Freya.
- Palette is seriously like 9 RU or something, so if you think bigger is better, they win.
- If you want to match Freya, we win.
You get the picture, though. A bigger EQ is coming. It’s gonna be completely nuts. I’ll save all the detail of the development (and more photos, and a real timeline) for the chapter on Loki Max/Lokiest/Whatever The Heck We Call It when it hits.
In the meantime, we have Loki Mini+. A seriously old-skool equalizer for anyone who wants fingertip control of their system…
…even if you don’t know what an inductor-capacitor equalizer is.
