Krell KMA-i800 monoblock power amplifier

Ever since I raved about Krell's K-300i integrated amplifier after it was released in early 2019, I've wanted to review other Krell products. After spending more than a year and a half (since its prerelease announcement) awaiting the opportunity to review Krell's new flagship mono power amplifier, the KMA-i800 ($73,000/pair), the time has come. Both Krell models utilize the company's proprietary iBias technology, albeit in different iterations, and both were designed by longtime Krell engineer Dave Goodman (footnote 1).

In the conclusion to my K-300i review, I wrote, "The Krell ... has the smoothest, most listenable, and most all-of-one-piece sonics of [any] integrated I've reviewed. ... There's a round edge to its images that some might equate with the gentlest sprinkling of warmth, but others would describe as listener friendly. It certainly leaves me smiling." I thought it so excellent, and such a good buy for a Roon-ready streaming integrated with an optional onboard PCM/DSD/ MQA DAC, that I urged a friend to buy one. More than three years later, he's still smiling.

Would Krell's new, far more powerful class-A monoblocks inspire the same excitement? As much as asking the question may sound like the first installment in a made-for-TV soap series, finding out made all the work I did carting these beasts around more than worth it.

Design and engineering
The product overview for the KMA-i800 monoblock amplifier on the Krell website (footnote 2) lists the following key technologies: differential output; iBias—see below—paired with XD technology, which lowers output impedance (footnote 3); Sym-Max (Symmetry Maximization), said to virtually eliminate second-order distortion; a power supply "consisting of 5400VA of transformers and 188,000µF of filter capacitance"; power delivery via gold-plated circuit boards and silver-plated solid-copper busbars "with a very short physical path for extremely low impedance"; "Current-Mode circuitry with advanced current mirroring that is balanced, differential, and direct-coupled," with each stage individually tuned; an output stage with "16 pairs of 200W audio power transistors and eight pairs of audio driver transistors," resulting in a wide bandwidth; protection against excess direct current and short circuits on the outputs, plus voltage sags ("brown outs"); and Ethernet connectivity that allows status monitoring and control via a webpage.

I discussed iBias in depth in the sidebar to my K-300i review. Krell's product overview offers a KMA-i800–specific explanation, fleshed out on its iBias technology page. In short, "iBias enables the [monoblock] to operate in class-A up to its full rated power of 800 watts with absolute accuracy but without the heat produced by a traditional class-A design. It [uses] a novel circuit to unobtrusively monitor the current flowing through the output stage. This is part of a closed-loop controller that ... [adjusts] the bias to maintain [a ] low preset current level as the signal and speaker load changes [regardless of ] what the signal is doing or how the speaker impedance changes with frequency. ... By maintaining that preset current level the amplifier is always operating in class-A."

In a recent phone chat with Goodman and Krell owner Rondi D'Agostino, Goodman said that Krell's current designs are mostly the culmination of an evolutionary process that improves upon his work and the work of other engineers. "On the other hand, iBias is revolutionary. It came out of the mind of one of our previous engineers, but he wasn't around long enough to put it into a production model amplifier. That was my job."

Dave said that implementation of iBias in the KMA-i800 monoblock includes new circuitry "that adds feed forward compensation to make it respond a little bit faster and more accurately to the amplifier's biasing needs." He also noted that iBias is what allows the monoblock to deliver 800W of class-A power regardless of the load (footnote 4).

"That's the real beauty of iBias," he declared in his customary low-key manner. "It doesn't depend on the speaker load, it doesn't depend on how it varies with frequency, it doesn't depend on what the input is doing. You have a fixed amount of class-A power to work with, and you get that no matter what. The class-A operation you get virtually all the time will give you that nice, warm, easy-to-listen-to sound you associate with class-A amplifiers."

I asked Dave what other aspects of the KMA-i800's technology he considered most unique and important. He replied, "XD technology lowers the output impedance significantly and results in a lot more control over the speaker drivers. This reduces the acoustic distortion you would otherwise get when the drivers are less controlled and flapping around a bit on their own outside of what the signal wants them to be doing. It produces a lot more front-to-back depth to the soundstage and a lot more black space around the performers and instruments.

"Sym-Max technology significantly reduces second-order harmonic distortion and brings out more detail. It really accentuates your perception of the pacing. You get a much more well-defined sense of timing and deliberateness from the musicians on the recording. It's also easier to separate the performers on the soundstage. Together, XD and Sym-Max deliver a much more detailed, realistic, and enveloping sound experience.

"Sym-Max technology is applied throughout the amplifier to reduce second-order harmonic distortion at every stage. By the time you get to the output, that distortion is almost imperceptible.

"The second and third harmonics are the dominant harmonics in the distortion content of most amplifiers. In tube amplifiers, the second order is the dominant harmonic. They'll tell you that distortion sounds good. But what's really happening in tube amplifiers is the third harmonic, which is primarily responsible for harsh, fatiguing sound, is much lower than the second harmonic. That's why tube amplifiers with less third-order harmonic distortion tend to sound somewhat sweet and pleasant. But the second is still high enough to end up masking a lot of the performance's detail and spaciousness. With the Sym-Max and several other design techniques, we really minimize both second- and third-order harmonic distortion and deliver the best of both worlds. The third is still quite low in the KMA-i800 due to some of the advanced current-mode techniques we utilize.

"One of the other key design features, compared to previous designs, is the addition of extra driver transistors in the output stage; this gives much better control and speed over the output devices they're connected to. It also contributes quite a bit to the overall performance of the amplifier. The configuration grew out of developing a purpose-built amplifier for the tweeter section of the Apogee Grand speaker system."

Goals and values
Given the KMA-i800's huge power output—Krell says it can output 800W RMS into 8 ohms, doubling twice to 1600W into 4 ohms and 3200W into 2 ohms, and can drive speakers of 1 ohm impedance—I wondered whom Dave and Rondi envisioned as buyers. "There are plenty of people out there who have rather large rooms and want commensurately large speakers they can play loudly," Dave replied. "No one is going to use this much power, but it makes average listening levels, which may go up to 100W, sound much better because that's a small percentage of the total possible power output. It's more linear than an amp that tops out at 100W, because as you approach the limit, your distortion goes up dramatically. A high-power amp gives you much more quality than an amp whose power exactly meets your needs." As someone acutely aware of the arbitrary and frequently over-exaggerated divide between the two viewpoints commonly labeled "subjectivist" and "objectivist," I asked Dave, "When you developed the KMA-i800, were you looking at measurements or at sound? What was the final arbiter that led you to declare, 'It's done!'?"

"Quantitative measurements—distortion and noise—are there to let you know you're on the right track," he said. "Certainly, if the distortion is high enough, you're going to notice it right away; below a certain level, you won't notice it. Some of the circuit-design techniques come into play when you decide how to lower that distortion. Once you get the design to operate within a reasonable range of the objective specifications, then you start your listening evaluations and make adjustments based on how it sounds."

Footnote 1: Goodman, who has been with Krell since 1987, designed Krell's Altair, KAS and KAS-2, KSA-200S and KSA-300S, KAV-500, and KAV-250A amplifiers. He also designed some of the circuitry in the FPB amplifiers.

Footnote 2: See

Footnote 3: See

Footnote 4: With a typical class-A amplifier, the power operating with class-A bias is halved when the load impedance is halved.

Krell Industries LLC.
45 Connair Rd.
CT 06477-3650
(203) 298-4000

georgehifi's picture

Without any consideration for the environment, and for the filthy rich.

Bet the Alexia's didn't know what hit them, and never will again.

Nice review
Cheers George

Ortofan's picture

... one of 75 pairs of the McIntosh MC2.1KW 75th anniversary edition monoblock power amps.
2kW into 8, 4 or 2 ohms and $100,000 per pair.

georgehifi's picture

Only to the ones with no idea and that love the glitz and glamour of "McIntosh Blue/Satinchrome" and that have no knowledge that an output transformer is complete and utter BS when tacked onto the output of a "good solid state amp". Just ask the likes of the great solid state designers Pass, D'Agostino, Curl, Flemming etc etc

Cheers George

Ortofan's picture

... the output autoformer equipped McIntosh MC462, you may observe that SM stated that the amplifier "will bring the breath of life to your music" and that "the senses of ease and literally quiet power created by the MC462 are palpable and most welcome. A first-round vote pick for induction in Class A of the next edition of our "Recommended Components." Highly recommended—insisted on, even." JA1 concludes his technical report with "summing up the McIntosh MC462's measured performance is easy: It is an extraordinarily well-engineered, exceptionally powerful amplifier."

Based upon the above, to what deleterious effect, if any, can you point that might serve to support your opinion that "an output transformer is complete and utter BS when tacked onto the output of a "good solid state amp"?"

georgehifi's picture

The only time it "may" be advantageous on a solid state amp is if the amp is an inherently unstable design into low/capacitive speaker impedance's, otherwise if you were to put one on these Krell Mono amps every parameter that makes them great amps would take a hit because of the transformer.

Cheers George

JohnnyThunder2.0's picture

more sophisticated equipment than McIntosh. A silly comparison. Filthy rich people who like to have parties in their mansions all the time may crave McIntosh though.

bhkat's picture

While I would never buy a pair, I'm glad audio equipment like this is being made.

georgehifi's picture
hiendmmoe's picture

Every time I read a review on a new design they are alway accompanied by the same old troll's negative comments. Either they reinforce each other’s negative stance or they argue over B.S. They want to tell you their negative opinions because some how they believe you should think like they do. There’s an old saying: opinions are like….everyone has one, but nobody gives a …… about anybody’s else’s except their own!

cognoscente's picture

That's more than 2x the price of an Audi A1 and even 2x the price of an Audi A3. Also more expensive than the Audi A6, which is an expensive excellent hell of a car. And that for only a power amplifier, then you still need all the rest.

funambulistic's picture

... a quick browse of Autotrader shows they are 1/13 the price of a used 2023 Ferrari SF90 Spider in my area. Just putting it in perspective...

cognoscente's picture

I'm just putting thise prices into perspective, if you still don't find such prices absurd, then you don't find such prices absurd. Simple!

georgehifi's picture

In anything the price goes up exponentially for what's seen as the very best

Cheers George

David Harper's picture

"In anything the price goes up exponentially for the very best". In many things. But not high end audio.

LTig's picture

Somehow the specs seem to be wrong:

- Output voltage: 245V peak–peak, 86V RMS.
- Output current: 56A peak (which is ~20A RMS)

86V * 20A = 1700W max power at 86V/20A = 4.4 Ohm load. This is clearly less than what JA measured: 2600W at a line voltage of 100V. At 110 V we could expect a power of 2600W * (110/100)² ~ 3150W which is very close to spec. Impressive.

I wonder though what kind of speaker would survive so much power.

Jason Victor Serinus's picture

Power of this nature, when instituted properly, shortens response time and allows super Ferrari amplifiers to turn on less than a dime.

windansea's picture

Perhaps the sliding bias will help extend the lives of these Krell amps versus the old Krells that often had to go back to Connecticut due to the extreme damaging the innards (shipping costs!). I had my Krell phase with Martin Logans and I won't go back to that. But I do miss the bottomless bass. I remember once running a frequency test CD and thought there was an earthquake rattling my windows.