Linn Klimax Solo 800 monoblock power amplifier

For reasons that have as much to do with why the creek does or doesn't rise as anything, almost all the UK-manufactured electronics I've reviewed over the years were from dCS of Cambridge. Through 2023, no electronics from Linn or a host of other UK-based companies have crossed the threshold of my music room.

That situation changed when, soon after New Year's, a pair of Linn's 60lb Klimax Solo 800 monoblock amplifiers ($90,000/pair) arrived from Scotland. Right away, they delighted me with their ease of maneuverability, handsome, uncluttered look, and relative compactness. Given their impressive power output—400W into 8 ohms, 800W into 4 ohms, and a whopping 1.2kW into 2 ohms—the Klimax Solo 800s set a record for highest price per watt and per pound among class-AB monoblocks I've reviewed.

The Klimax Solo 800 had not yet been released when an early pair arrived for review. Information about it began to appear online only well after the review period began (footnote 1). There is no manual, because every Linn product is dealer-installed: It's the dealer's responsibility to make sure you're well-informed. Besides, "the amp is designed to plug-and-play, as it were," Brand Manager Joe Rodger told me in an email. For those who want something to read, a "Linn Docs" wiki for the Solo 800 can serve as what Rodger termed a "no-frills encyclopedia." (footnote 2)

To learn about the Solo 800, I Zoomed with Murray Smith, the electronics team leader who initiated the project that led to the Klimax Solo 800 and who worked on its Adaptive Bias Control system, and Linn Senior Electronic Design Engineer Nina Roscoe, the design lead for the Klimax Solo 800. Years ago, Linn sponsored Roscoe's attendance at Cambridge University, providing sufficient industrial electronics manufacturing experience to allow her to take her second- and third-year exams. When she finished her studies, she spent a year or two at Linn, doing some work on the long-discontinued Klout amplifier before moving on.

"It was a really good experience," she said. "It's a good educational system—one that actually gets a design engineer to spend lots of time in manufacturing. You understand how the products that you make are built and learn the implications of the design decisions you make. The process creates reliable products and makes for long-term customers."

New product, new design goals
At Linn, the Klout was followed by the Solo 500. Intended for the smaller power market where the majority of loudspeakers present 4–8 ohm loads, it remains in production after almost 25 years. Eventually, acceding to requests for a higher power monoblock that could drive larger speakers with lower impedances, Linn decided to create a new flagship monoblock that would utilize recent improvements in electronic and component design and architecture. By this time, Roscoe had returned to Linn.

"I think probably the biggest performance advance we were aiming for in the Solo 800—the thing we felt makes a huge difference to a power amplifier—was lower distortion," she said. "We went all out to bring distortion down from every source and choose topologies and components that give you the lowest possible distortion. We were really aiming for the purest signal reproduction. While distortion levels in the older Solo 500 are really quite good, there is room for improvement, especially when amps are driven really hard by large, lower-impedance speakers.

"We were able to do more with more modern architectures and more modern components. Previously, we hadn't really done a lot of the design work necessary to drive difficult speaker loads. Mostly, we'd just aimed at 8 ohm speaker loads rather than at much bigger loudspeakers that sometimes present really difficult loads. We wanted the amp to perform with excellence even when it was working really hard." Many engineers initially focus on measurements as they build prototype after prototype. Once their measurements are the best they can achieve, they turn their attention to sound quality, "tuning" the sound. Roscoe and her team proceeded differently.

"We didn't really feel we needed to tune the amplifier sonically," she said. "We designed for very, very low distortion levels, very low noise levels, wide bandwidth, and low output impedance, all of which are obviously measurable. Then, when we listened to the design, we felt that it sounded so clear that there didn't seem to be reason to do more. I don't know what we would have tuned really. We didn't feel like we could improve it by listening to it and changing something empirically. Instead, we heard what you would expect from an amplifier that measures as well as the Solo 800 does. It really seemed to pick out music better.

"Ultimately, when 'tuning' the sound by listening and modifying the design, designers usually select which distortion is least offensive to the ear. However, when there is so little distortion there, this becomes a more meaningless exercise."

Short of resorting to class-D (which obviously they did not do), how could Linn get so much power out of an amplifier that weighs a teeny bit less than 60lb, is relatively small, and manages to run at a stable, cool temperature without the use of a fan?

"The amp weighs enough and is big enough to ensure it runs at a nice, cool temperature with maximum volume levels and average kind of music," (footnote 3) Roscoe said. "The fins, which are machined from a solid block of aluminum, are vertical; this design allows a lot of air to move up through them. We've also got vents underneath the bottom plate and in the groove around the lid. We tried different heights and sizes of the heatsink until we perfected a design that sucks air up from underneath to help cool the transformers in the switching power supply and the driver transistors on top inside of the board."

A switching power supply
"We also use an efficient switching power supply. We at Linn feel quite strongly that switching power supplies are a superior way of powering an amplifier because all the power supply noise moves out of the audio band. Our switching power supplies maintain very high switching frequencies that can't come near the audio band. The 50Hz, 60Hz, and 120Hz [supply-related spuriae] are very low in these amps, and the power supplies are very efficient."

Roscoe said that linear power supplies require very large transformers, which are very inefficient. "The first problem with a linear power supply is that it's working at 50 or 60Hz, which is right in the audio band. When you try to filter the noise and get ready to power an audio amplifier, generally what happens is you get a certain amount of mains noise and frequency leakage. ... It delivers DC voltage to the output, which then needs to be converted/regulated down to the lower voltages needed to power a lot of the circuitry in the amp. Doing this throws away quite a bit of power. ... Switching power supplies ... upconvert the power frequency to create a very high-frequency AC signal that requires a much smaller transformer. With a lot less copper and much smaller cores, switching transformers generate less heat and operate faster, at higher frequencies.

"Some people in the audio world don't like switching power supplies because they switch very fast and, in their opinion, create a lot of switching noise. If that noise does end up in the audio band, it can make things worse. That's why we make sure the switching frequency is way beyond anything that the human ear is ever going to pick up, even when we deliver 1.2kW into a 2 ohm load."

"I think the reason switch mode power supplies have gotten a bad rep in the audio world is that they were originally used mainly for computers and other applications where their kind of electrical noise levels weren't that important," Smith said. "You can't just stick a switching power supply from a computer into an amplifier and achieve a good result. ... You really have to understand all the principles behind them before they can meet all the myriad audio requirements of being silent and noise free.

"Linn has been using switching power supplies for a number of years. Designing them properly for audio requires a lot of hard-to-find, specialized skills. ... We take the money that other manufacturers usually spend on the iron and copper in their big transformers and spend it on sophisticated electronics instead."

A problem with linear supplies, Smith told me, is that because they operate at mains frequencies of 50 or 60Hz, their noise falls right in the audible range: 60Hz and its lower-order harmonics. This is often heard as a low-level hum or buzz. Finally, "traditional linear transformer supplies also lack any form of regulation. As a result, their output voltage is at the mercy of the mains voltage in your house.

"Short of buying a large mains generator to keep voltage constant, a good audio switch-mode power supply can keep voltage constant without dumping audible noise into the audio band. Instead of taking power in short pulses, switching power supplies can be designed using a Power Factor Correction circuit, which draws power in an even and consistent way, creating no more disturbance than an old incandescent light bulb (albeit quite a big one). That makes the amplifier a much better citizen on your mains network. It's far less likely to interfere and interact with household appliances or other components in a hi-fi system. Switch mode power supplies also feed the output they generate back into a control circuit that maintains constant and steady output voltage, thus protecting your audio output from any disturbance on your mains supply."

Adaptive Bias Control
While developing the Klimax Solo 800, Smith helped develop Adaptive Bias Control to address the crossover distortion created in class-AB amplifiers when the set of transistors that generates the positive half of the musical waveform hands over the signal to the set that generates the negative half. As Smith explained during an extremely cogent discussion, "We have this handover problem as the signal crosses through zero while it goes from positive to negative."

All linear amplifiers apply a bias voltage across the output transistors to set bias current during crossover. The applied bias must constantly adapt to compensate for the changes in temperature, voltage, and current that transistors undergo as they react to the signal passing through them, while also accommodating aging effects.

"It's almost impossible to do that with exact correctness," Smith said. "If you have too much correction, then you can end up with thermal runaway, where the devices just get hotter and hotter until they destroy themselves. That's why most compensation schemes are deliberately 'detuned': to make them safe. In addition, as the amplifier warms up, distortion characteristics can change. Quite often, that's the reason for long warm-up times in amplifiers: They may need a certain amount of time to settle to a good performance characteristic.

"Because no two transistors are the same, transistor circuits need adjustment from amplifier to amplifier in order to maintain correct bias. ... With Adaptive Bias, we measure the bias condition of the transistor and then use a digital control algorithm to maintain bias at a completely constant level regardless of temperature or transistor type. ... It's not an easy thing to do in a purely analog sense because the signals that you need to measure aren't easy to extract. The bias signal that we want is mixed in with the music signal within the amplifier."

Linn monitors the current flowing in the output transistors and uses a digital algorithm to extract the bias signal from the music signal. "That allows us to control that parameter on its own," Smith explained. "It's something that you can only achieve by having good knowledge of traditional analog amplifier design, digital electronics, and digital signal processing. It's a kind of marriage of two different worlds. Digital signal processing is my specialty, so that's the part of the design I was involved in."

He emphasized, though, that the audio signal is not digitized. "The signal path remains completely analog. But we have digital assistance to ensure the amplifier performs consistently and runs at its best. It's almost like the engine-management system in a car. ... We have a completely linear transistor amplifier with a digital engine-management system that is keeping operating conditions constant and ensuring that performance is always optimal."

Footnote 1: See

Footnote 2: See

Footnote 3: When asked to define "average," Roscoe said, "What I meant by 'average' is really the standard 12dB 'crest factor' which is usually applied to music. It dictates the average power handled by a power amplifier when operating just below clipping. Some tracks might push the amp a bit harder, but so far nothing has really managed to get them really hot, even while driving challenging 'difficult' speakers."

Linn Products Ltd.
Glasgow Rd.
Waterfoot, Glasgow
Scotland, UK G76 0EQ
+44 (0) 141 307 7777

georgehifi's picture

Nice to see a ClassA/B amp without Class-D switching noise filtering needed when it's being measured.

"how could Linn get so much power out of an amplifier that weighs a teeny bit less than 60lb, is relatively small, and manages to run at a stable, cool temperature without the use of a fan?"

SMPS (money and weight saver) and Class-A/B (low bias=cooler)

(not a lover of the SMPS especially in source equipment)
Go near one with a old AM portable radio that has no auto mute if it tuned in off channel, tune it in down around 600khz off channel, turn up the volume and hear what horrors come out of it.

Cheers George

hb72's picture

am in the camp of those who try to eliminate all SMPSs from the A/C circuit that feeds the hifi system. But, if I understand correctly, provided all components, sources, pre & power-amp are powered by SMPS, that noise on A/C would not be an issue to the hifi system.

If I understand correctly, it is the mix of Lin and SM PSUs that is problematic, as lin PSUs are less good in filtering noise on A/C from entering the DC section. High Quality SMPS provided.


georgehifi's picture

They radiate not only through wiring but also airborne radiation.

Cheers George

hb72's picture

Some suppliers of external SMPSUs allegedly provide useful filtering on DC & AC side and also usefully grounded housings (ifi); here is where a low (HF) impedance power cable comes handy :). I would assume the same for companies like Linn, wow have the SMPSUs integrated in their source and amp components.
Point being: (aside of their way faster voltage control) SMPSUs are said to be much better in isolating the DC side from AC side noise, while LinPSUs apparently are far less so, and thus more prone to AC side noise passing through to DC side; OTOH they are less noisy on AC side so, one would think this warrants limitation to linPSUs, or, otherwise, to SMPSU driven components only.
Is there anyone who can confirm such considerations from real experience?

SNI's picture

The reasons why you have to use filters for measuring class D amplifiers, maybe somewhat different from what most people think.
Using an AP analyzer for Class D measurements, several issues has to be considered.
First thing is, that you don´t want to put illeagal signals into an amplifier with limited bandwith. That makes no sense.
A class D amplifier is intended to amplify audio signals, not signals made for measuring purposes, hence i.e. square signals will not be of any use.
Second thing is, that you need to filter out the carrier wave. The reason for this has nothing to do with class D, but with the way he analyzer works.
If you want to look for very low noise and distortion levels, the AP analyzer will adjust its sensitivity for that purpose. That would not be possible with a carrier wave @ several V at the output of the class D amplifier @ ~500 KHz. The analyzers input stage would be damaged.
And it is worth noting, that a carrier does not affect the audioband at all what so ever.
Also worth noting is, that most class D amplifiers are actually real DC amplifiers without any kind of servos or like. They can just do frequencies down to DC by nature.
As a small extra, class D amplifiers has no crossover distortion.

Regarding SMPS there is at least 17 different types in existence.
Some of those are relevant for audio, and a lot of them are not.
On top they can be regulated or not, as well as they can be PFC SMPS, where PFC stands for Power Factor Correction, meaning less noise on the AC side, and increased efficiency.

Anyways what I´m trying to say is, that IMO nothin can be said generally about SMPS. There are to many different types, which easily gets mixed up.
Needed filters for class D amplifiers, has nothing to do with performance or sound quality, it is the analyzer, that is in need for filtering, not the amplifier.

But still the Linn amplifier presents very good results.
Almost as good as i.e. Toppings 800USD 56W Pch amp.

georgehifi's picture

SNI: "Needed filters for class D amplifiers, has nothing to do with performance or sound quality, it is the analyzer, that is in need for filtering, not the amplifier."

JA: was measuring these Class-D's way back with the AP analyzers without filtering at first and showed what was coming out of the amp for real which was ugly!! Fig-2 and then started using the AP filtering to make them more acceptable to look at Fig-3 even though it was not representative of what the speaker was really seeing which is what's in Fig-2

It is possible that these HF switching frequency oscillations riding on that square wave "can" cause an amp to become unstable, and cause it to ring/oscillate and punish tweeter voice coil over time, has happened a friends Watt/Puppy 7's tweeters which sounded "off" with his Nu-Force Ref Mono's and the reason when I dismantled them for new voice coil replacement, the old voice coils (both) got severely "blued" with heat and he's not a loud listener

Cheers George

CG's picture

Two questions...

Do the in-spec production units sound different than the not in-spec pre-production units?

What kind of notch filter is used to measure the distortion waveform? With distortion products below -100 dBc, that implies that you need a notch depth of well better than 100 db. (I'm guessing that it's part an analog notch in the APx555 and a lot calculated in the '555, but I want to see what I can do to come close at home.)

JRT's picture

Thank you for another interesting review and measurements.

At $90k/pair, I simply wouldn't want to accept the opportunity costs on these.

I would very much like to see review and measurement of monoblock amplifiers providing good implementation of Purifi's 2nd gen Eigentakt 1ET9040BA module. I have seen comments from Buckeye Amplifiers that they will be bringing their new 1ET9040BA based monoblocks to demonstrate at the MWAV (midwest audio visual) trade show in Kansas City in late June. Their target price is $1050/amplifier shipped direct, (so $2100/pair, which is $87,900 less than the amplifiers that are the subject of this review).

Note that -120dB is 0.0001%

My current understanding of the model number is:


1: single channel
E: Eigentakt
T: ?
90: = 90 volts peak at maximum output
40: = 40 amps peak at maximum output
B: bridged
A: amplifier

edit 05/14: Purifi has posted a link to the datasheet, now openly available to consumers at their website. The module is also now available (EUR 598.00 VAT excluded) at their online store, as well as an engineering eval/demo kit (EUR 696.00 VAT excluded).


I have no profit motive in any of this, nothing to personally gain or lose by it. merely exhibiting some enthusiasm.

Anton's picture

I can tell you before they even get built how the subjective review would go, but seeing JA hit those with his measuring tools would be way cool.

David Harper's picture

Like you I, to, can predict how the subjective review will go. It will feature imaginative and evocative descriptions of sound which will challenge the intellect and cause the reader to abandon all reason in his pursuit of audiophile bliss.

supamark's picture

Including prior Eigentakt, Ncore, and Ice power amps - but not any of the GaN FET amps, it will sound grey and harsh in the treble. They all sound like the worst of 80's digital. The closest analog would be the vinyl version of Eric Clapton's Crossroads anthology, which was edited/mastered digitally and sounds like grade A (or gray) dog crap. I also don't get why people like DSD, but different strokes.

I would like to hear one of the better GaN FET amps (and their higher switching frequency) to see if it "fixes" class D.

Bias alert - I prefer class A amps with minimal/no global feedback. And MOSFETs over BJT's in the amps I've heard.

Yes, the puns were intended.

JohnnyThunder2.0's picture

get you all hot and bothered instead of the love of music and subjective listening tests tells us all we need to know about you. If I were you, I'd take the $87,000 you've saved buy buying those black boxes you like and travel to NY or Chicago or Vienna and listen to real orchestras and music. PS - are you getting kickbacks from Buckeye or Purifi ? Very interesting that at Apoxna there were 100x more tube amps demonstrating music than little black boxes w class D modules.

Anton's picture

Can't crap on it unless you've heard it, right?

Put this with cheapie with a pair of DeVores, Klipsch, and have a comparo!

I would love to see that.

JRT's picture

I suspect that Andrew Jones' new MoFi Electronics SourcePoint 888 ($5k/pair) 3-way floorstanding loudspeaker might work well in combination with well engineered amplifiers utilizing Purifi's 1ET9040BA module, depending much on the surrounding components (input buffer, SMPS, ventilated case, etc.).

I recall some mention of Erin Hardison ordering/buying a pair of SourcePoint 888, so I expect to see, eventually, a subjective review and a set objective measurements (he uses Klippel nearfield scanner) with associated commentary on his Erin's Audio Corner website and YouTube channel.

edit: Stereophile posted a video on their YouTube channel, with Andrew Jones describing his new SourcePoint 888.
To watch the video, copy/paste the following link into your web browser address window.

Ortofan's picture

... the live sound of a "real orchestra and music" as heard in a concert hall in "NY or Chicago or Vienna" as a reference point.
Have you not read J. Gordon Holt's writings in that regard?

JohnnyThunder2.0's picture

and orchestral reproduction. I was suggesting that with all that money he would save by buying inexpensive equipment to travel to different cities to experience an orchestra live - in an audience, as an enriching life experience, not as a point of comparison with home audio. I was making a point about the love of music and life vs. equipment and measurements. That's all.

JohnnyThunder2.0's picture

with your point....That would be interesting. M

georgehifi's picture

Nice to see an amp again that almost doubles it's "tested" wattage from 8 to 4 to 2ohms again which means it's got a pair.

Cheers George

supamark's picture

I would like the sound of *first* amp JA tested a fair bit more than the second amp. The difference is in the amount of global feedback - which is almost certainly why the first amp had more ouput power; way less feedback. This is clearly visible in the power vs distortion curves.

If the first amp tested is also the one of the ones JVS listened to, there needs to be a follow up to see how the sound is changed.

georgehifi's picture

"amount of global feedback - which is almost certainly why the first amp had more output power;"

Amount of feedback, local or global doesn't influence output power or current ability much at all. But it does influence distortions and damping factor.

Cheers George

supamark's picture

That's the approximate difference in power. If removing global negative feedback from a 1 Watt headphone amp can add 6 dB (4x) gain (Schiit Asgard 2), then I think an approximately 4% drop in power is quite plausible in a 400 Watt amp.

Also, the way the feedback is added fitfully is probably audible with the right playback material... mfg's need to stop chasing SINAD, anything over ~100 dB s/n is weiner waving since microphones top out in the mid 80's s/n wise. The mic is just the start of the signal (to noise) chain, plenty more noise is added along the way (and some of it intentional).

georgehifi's picture

supamark: "If removing global negative feedback from a 1 Watt headphone amp can add 6 dB (4x) gain (Schiit Asgard 2), then I think an approximately 4% drop in power is quite plausible in a 400 Watt amp."

Yes increasing or reducing feedback will effect "the gain" (sensitivity) of any amplification device be it a buffer/preamp/or poweramp, not the wattage or current ability of it.
It will not give the device the ability to give any more or less maximum wattage/current (power) than it had originally, it will just have "less gain", and the need for more or less voltage from the sources before it to drive it to those maximum wattages/currents "powers"

Cheers George