High-Tech Meets Tubes
Posted on February 21, 2020 by Vade Forrester
Your car is full of computers that make it easier to operate and maintain; why shouldn’t your hi-fi follow suit? Increasingly, that’s what’s happening. One of the more recent computerized hi-fi circuits sets the bias current of the output tubes on tube amplifiers. Those circuits make it feasible to switch easily between different types of tubes. So if an amplifier is shipped with EL34s, you can change them to KT88s and the automatic-bias circuit will adjust the bias to suit the KT88s with no adjustment from the user. Just insert the new tubes and the bias circuitry changes the necessary parameters. Rogers High Fidelity’s new Corona KWM-88 integrated amplifier’s automatic bias circuitry makes it possible to switch between KT88s and KT150s. It’s your choice. Just specify when you buy the amp which output tubes you want. (Both types produce the same power.)
The Corona deliberately leaves out features that can be found on several other integrated amps. There’s no DAC, no phono preamp, no headphone amp. This makes sense to me; if you’re willing to shell out $15,000 for the Corona, you’re probably a very serious listener who wants an advanced stand-alone DAC, phono preamp, and headphone amp. You know exactly which of those units suits your system, and don’t want to waste money on the type of limited or compromised integral DAC, phono preamp, or headphone amp that’s typically included in some integrated amps. Rogers makes a very interesting phono preamp, the PA-2, which is definitely worthwhile considering to match to the Corona, but doesn’t offer a DAC or headphone amp. The Corona has the typical integrated amp advantages: It requires no preamp-to-amp interconnect, and it occupies just a single shelf, albeit a very sturdy, well-ventilated shelf.
The Corona is built on a wide, flat chassis with a silver, brushed-aluminum faceplate in the front, tubes in the central area, and in the rear a transformer that spans the full width of the chassis. In a departure from the norm, the chassis is painted red instead of the usual Boring Black. The red color is not a garish, bright hue, but sort of a candy apple red. I thought the combination of the brushed aluminum faceplate and transformer trim plate with the red paint on the chassis looked quite elegant and attractive. (I’d welcome more color departures from black chassis. Maybe not lime green, though.)
The faceplate has, from left to right, four toggle switches (power on/off, standby, mode, and input) along the lower part, with a blue information window on the upper part. The labels for the switches are engraved in the front panel, which looks nice but in my view is harder to read than painted labels. To the right of the switches and information window is the silver, serrated, single-knob volume control, CNC-machined from aluminum. The right half of the faceplate is reserved for two VU meters. I don’t know that the VU meters serve a functional purpose, but they surely look cool, with their blue illumination matching the status window on the right. There’s no mute control, balance control, mono switch, or phase-inversion switch, either, all of which might have been useful. However, there’s a fair chance none of those controls would ever be used, so leaving them off makes financial sense.
The tube choice is a bit unusual, comprising two output tubes per channel–normally KT88s or KT150s (KT120s will also work)—and one 6SJ7 and one 6SN7 low-level tube per channel. The 6SN7 is a very common dual triode, but the 6SJ7 is rather unusual in current amplifiers. It’s a pentode with a metal rather than a glass envelope. The 6SJ7 tube serves as the input stage and provides all the gain for the amplifier. I asked Rogers’ President and Chief Engineer Roger Gibboni why the 6SJ7 was chosen instead of a traditional triode and he said, “I use the 6SJ7 in the front end for very specific reasons. First, they are still in production. They are available from Sovtek and a Chinese manufacturer. In general, we prefer the Russian versions for reliability. The 6SJ7 was in such frequent use that there are many sources of NOS tubes still available and affordable, and we will use them as long as we can get them. You were supplied with the 5693—a special red version of the 6SJ7. These were designed for the military and are high-reliability versions. They have extra structures inside that allow them to be employed in environments of shock and vibration up to 10Gs. This extra rigidity also gives them much lower noise characteristics than standard 6SJ7s, and we sell them as an upgrade for $125 each.
“I use the 6SJ7, which is a sharp-cutoff pentode, in the front end rather than the traditional triodes because you can achieve much higher gain in a shorter signal path. We get 35dB of gain in one stage that with traditional triodes would require two or three stages of amplification. This yields much lower phase distortion because the gain is achieved in a shorter signal path.
“In almost all cases, the small signal tubes—6SJ7 and 6SN7—have almost infinite lifespans. They run very little current and are not under stress in the amplifier. The user should expect 5000 to 10,000 hours with these tubes.
“The KT88 or KT150 power output tubes on the other hand run significant current—approximately 125mA per tube. This operating condition achieves the dynamic range and transient response you hear in the Corona. These tubes typically last 2000–3000 hours, about two years of normal use. The beauty of the Corona is that the bias condition of these power tubes is managed by the on-board processors. When a tube exceeds its normal self-bias range due to aging or failure, the processors shut down the affected channel and indicate which tube requires replacement on the screen and app. Because the output tubes are auto-biased, the user does not have to replace the power tubes as sets. Only the failed tube requires replacement. We supply tested and burned-in KT88s for $135 each and KT150s for $175 each. While we have very few requests, the Corona will also run the KT120.
“The Transparent power cable you were supplied is a $75 upgrade to the normally supplied power cable.”
The 6SN7s drive the output tubes for a 100Wpc rating in Ultralinear or 80Wpc in triode mode. The output tubes sound different, and since I was provided with a set of each type of tube, I’ll explore the sonic differences. I’ve experimented with both types of tubes in my Audio Research VT80SE amplifier and can see why some people would prefer the KT88 sound. KT88s are popular tubes in production by several manufacturers, so you can pick the ones you like, but, again, Rogers rigorously tests the tubes it provides, so it’s not a bad idea to buy their replacements. The same goes for the 6SN7 tubes, which are in production from several manufacturers. The Corona shipped with Gold Lion tubes, except for the KT150s, which were made by Tung-Sol. Those are all high-quality, current-production Russian tubes.
The signal circuit may use tubes, but the heart of the amplifier, the power supply, is all solid-state. Built around a 30-pound toroidal transformer (that’s half the weight of the entire amplifier), the power supply is rated for 1800W—way overbuilt.! The Corona is burned in for 100 hours to prevent any infant-mortality problems, and get you well on your way to total burn-in.
Rogers emphasizes that the amplifier is constructed of aircraft-grade aluminum with military-spec wiring and electronic parts. I don’t know if such construction makes the amp sound any better, but it certainly does make it more rugged and stable. It’s got to be a factor in longevity—the warranty is transferrable limited lifetime, which is about as good as you’ll find, and far better than most competitors. It even applies to amplifiers that are re-sold by dealers. With the exception of replacement tubes, it sounds as if you’ll have a hard time spending any money on your Rogers amp, unless you do something goofy to it.
On the rear panel you’ll find four unbalanced inputs, an IEC jack for the power cord, and two Furutech solid-copper output jacks for the speakers. Input impedance is 100k ohms, which should work with any conceivable source. Unfortunately, there are no line-level outputs to drive powered subwoofers or other external amplifiers. There is also a connection for an included Bluetooth antenna, which is how the tablet or phone connects to the Corona. The volume graphic on the app drives a motorized control which operates the front-panel volume attenuator.
The remote-control app on my iPad was named Rogers High Fidelity Corona at the Apple App Store. All I had to do was download and install it, and it connected immediately via Bluetooth to the Corona.
The Corona is a very expensive integrated amp designed to provide extremely high-quality sound from a relatively compact unit. The price tag will place it beyond the reach of most readers, but as with any pricey item, the question of value comes into play. In other words, can you attain the same level of performance for the same price or less?
Setup and Use
Normally, when I get a component for review, I first read the owner’s manual, or at least the set-up guide. The Corona’s manual was provided in digital form as a Microsoft Word document on a USB drive. I quickly printed it out and checked out Section 2, Quick Start. It had pretty good step-by-step instructions for installing and setting up the amplifier—except, several items were omitted. For example, there was an antenna in the amplifier box, but no mention of its function (it’s a Bluetooth connection for the remote control). It appeared to screw onto a fitting on the rear panel of the amp, so that’s where I put it. If the manual had contained a photo or drawing of the rear panel, it might have shown me where the antenna should have gone. There was no specific reference to the remote-control app, either. After checking out a few guesses, I asked Rogers and learned that the app is named Rogers High Fidelity Corona. I probably would have guessed that eventually. But why not just put it in the manual? Once installed, the app connected effortlessly (as noted) to the amplifier via Bluetooth. I installed the app on both my iPad and iPhone, and discovered only one device at a time will connect to the Corona. The iPhone screen is necessarily smaller, but still easy to use. I found myself using it most of the time, since I carry my phone with me around the house.
The Corona isn’t a huge amplifier, 17″ x 14″ x 11½”, but its 60-pound weight needs a sturdy shelf or support with plenty of ventilation for the tubes. For me, that meant the top shelf of my equipment rack, and was I ever glad to have help lifting the Corona to eye-level.
I connected the Corona to my DAC via High Fidelity Cables CT-1 Ultimate unbalanced interconnects and to my speakers with Van den Hul Mountain speaker cables. Rogers furnished an accessory power cord manufactured by Transparent. It’s a molded cord, but needless to say, it uses top-quality parts in its construction, and is way better than most stock power cords. Of course, it’s not stock, selling for $75 for a 10-foot length. I normally use the stock power cord when reviewing a component, knowing that using an aftermarket cord could make noticeable differences, but since this aftermarket cord came from the Corona’s manufacturer, I used it. The power cord was plugged directly into the wall, not into the Shunyata Denali power distribution system where the rest of my system is plugged.
When I swapped out the tubes, removing the KT150 and inserting the KT88s, I waited the normal amount of time, 30 seconds, before switching from standby to play, and the new tubes worked without any fuss. The new tubes showed up in the information window and the app as being in good condition. I wished the amplifier had used the computer to track the time the tubes had been in service, so I could replace them when they were worn out, but maybe that feature will appear on a later model.
I found myself indulging in some Old Geezer nostalgia: I remember when tube gear was noisy. Not so much now, though; the only noise I heard from the Corona was a slight pop when the amp turned off and when switching between Ultralinear and triode mode. (Rogers says that this has been fixed in current production.)
Of course, as with any amplifier review, no subwoofer was used. Otherwise, I would be listening to the sound of the 1200-watt subwoofer amplifier instead of the Corona. The Corona took longer than normal to completely warm up.
A thin line is scribed on the volume control to show its setting, but from my listening position about ten feet away it was hard to see. I wished for an LED on the volume knob to show the volume setting better at a distance—or a numeric readout in the computer window. The remote-control app installed on my iPhone 7 worked just fine. There was a bit of latency as the amplifier’s volume setting tracked the input, but I got used to it. The range of the remote control was much further than standard remotes; I could operate the Corona several rooms away from the amplifier. That’s cool and useful.
The Corona impressed me as a highly linear, neutral amp with explosive dynamics at all levels. I noticed one unusual trait, however: The channel balance varied a bit with different recordings. And there was no balance control to adjust that balance. This was not a show-stopper problem, just a perceptible one. (Well, I guess it could be a show-stopper for some listeners, who are ultra-picky about channel-balance adjustment.)
I started my formal listening with the KT88 tubes inserted in the amplifier. Generally, KT88 tubes had good harmonic accuracy and lots of detail. I began the formal audition with old favorite Folia: Rodrigo Martinez 1490, a realization of a truly old piece of music played by Jordi Savall and his band of Renaissance instrumental specialists. In triode mode, this information-rich piece was reproduced with lots of detail, especially in the percussion where leading-edge transients were unusually audible, though the soundstage was skewed to the left. Bass had plenty of rhythmic energy and descended fairly deeply. Band leader Jordi Savall solos on the viola da gamba (an instrument similar to the modern cello), and his tone was just a smidgen less harmonically rich than with some amplifiers. In the Ultralinear setting, the Corona was more explosive, exhibiting taut bass impact, but with the same slight leftward skew in channel balance that I heard in triode. The viola da gamba tone was ever so slightly bleached.
Moving on to “Miserere” from The Tallis Scholars Allegri’s Miserere & Palestrina’s Missa Papae Marcelli, an a capella setting of Psalm 51 performed by a small choral group with a smaller solo group some distance behind the main group. In the triode setting, lots of nuance and shading was delivered with ravishing finesse from the main group; however, the distant solo group was immersed in more reverberant echo than usual. Nonetheless, their lyrics still sounded perfectly enunciated and defined in space with not a hint of bloat or splashing. The solo tenor’s voice had a dense filigreed texture, free from distortion. In the Ultralinear setting there was still some left skewing of the channel balance, though the distant solo group was immersed in less reverberation than in triode.
The song “Snilla Patea” with Bjørn Kåre Odde, the composer, expertly playing a fiddle, backed by the Schola Cantorum chorus under the leadership of Tone Bianca Sparre Dahl, is an MQA-encoded piece. My DAC won’t decode MQA, but the recording still sounds great. In triode setting, the piece seemed better centered than the other pieces. The composer’s solo fiddle was not as harmonically vivid as it is through some amps. The chorus also sounded slightly less detailed than I’ve heard. In the Ultralinear setting, the soundfield was well-centered. The chorus sounded quite lucid, and the fiddle produced a full harmonic tonal envelope. The chorus was remarkably agile.
Next, I replaced the KT88 tubes with KT150 tubes. My first impression was that the KT150s sounded wide open and relaxed. In the Triode setting, Folia: Rodrigo Martinez 1490 enjoyed a well-centered channel balance. The opening cascabels were more distinct—it was easy to distinguish each strike on the bells from the other strikes. The initial transients of percussion instruments were (appropriately) dissimilar. Bass was deep and impactful, and it was easier to tell the pitches of each note. Savall’s viola da gamba sounded harmonically complete and tuneful. In the Ultralinear setting, it was also easy to tell difference between cascabel notes. Likewise, percussion instruments were distinct and realistic. Castanets swarmed like bees. Bass was ever so slightly less impactful. I could even tell the difference between baroque guitar and harp as they repeated the same musical passage, where the two instruments often sound identical. I’ve seldom heard this much detail from an amplifier. Dynamics were also unusually forceful.
In “Miserere,” channel balance in the triode setting was skewed slightly to the right—barely perceptible. The solo group sounded buried in reverberant echo, but was still clear. The Ultralinear setting produce slightly less bright highs, but the solo group was less detailed.
In the triode setting, “Snilla Patea’s” soundstage was well-centered. The fiddle sounded a little bright. On the other hand, the chorus was articulate, airy, and dynamically agile. In the Ultralinear setting, the fiddle exhibited the same very slight brightness.
So which tubes are better, the KT88s or KT150s? I slightly preferred the latter, but several visitors preferred the KT88s. I guess that makes it a personal choice.
You probably wouldn’t use a $15,000 amplifier to drive a $1500 speaker, but I did. In order to see how the Corona drove a different speaker, I connected my KEF Q700 speakers. This three-way cost $1500/pair when last produced, but represents an alternate load for the amp. I’m glad I tried it, as the KEFs have never sounded so good. I was particularly amazed at their bass performance. In the Ultralinear setting with the KT150 tubes, on Folia: Rodrigo Martinez 1490, bass extended quite deeply, with good pitch definition and surprising impact. I haven’t heard such bass from the KEFs before. Surprisingly, their treble performance was equally good; the opening cascabel whacks were extended and harmonically accurate, with just the right amount of impact.
“Miserere” had an agreeable soundstage, possibly a bit left-skewed. There was a slight high-frequency distortion (it’s a characteristic of the recording). The sound of the distant solo group was a lifelike blend of echo and direct sound, which produced realistic separation of the groups.
The “Snilla Patea’” fiddle was spot on, with a full, rich harmonic spread sounding whole and accurate. Choral dynamics were agile, ready to change levels at the drop of a hat.
My Audio Research LS28 linestage and VT80SE are somewhat similar to the Corona in that the VT80SE has a similar automatic bias circuit that allows it to use different output tubes. I’ve tried KT88s in addition to the stock KT150s. It’s only rated at 75Wpc, however, with tube life anticipated at 3000 hours. The 6H30 tube is used as the driver tube. A meter in the rear of the amplifier tells you how much time is on the output tubes. The LS28 linestage uses only 6H30 tubes in a hybrid circuit. Unlike the Rogers amp, the LS28 has many controls, including a channel-balance adjustment, phase-reversal switch, and even a display of hours used on the tubes (a 4000 hour lifespan anticipated). Prices for the VT80SE and LS28 are $9500 and $8500, respectively. Of course, an interconnect is required to connect the two components, and that can be as expensive as you like. I use Van den Hul’s balanced The Mountain interconnects in my system, priced at $1454 for a meter length. So that makes $19,454 for the Audio Research/Van den Hul system, a bit more than the Corona. The Audio Research requires two shelves, and is larger than the Corona amp. So there’s a space cost as well as a monetary one.
The Audio Research LS28 has a normal metal remote which operates smoothly and quickly. All the controls on the linestage are on the remote. Tubes were KT150s, the stock tubes offered with the VT80SE. The amp has no remote.
On Folia: Rodrigo Martinez 1490, cascabels were located in the expected position in the soundstage, a bit to the right of where they appeared with the Corona. Bass was deep and pitch accurate, but less punchy than with the Corona. The viola da gamba tone appeared spot on. Dynamics were forceful, but the Corona’s were even more so.
On “Miserere,” the solo tenor had a bit of the left skewing the Corona had exhibited, but the soundstage was more evenly spread between the speakers. The distant solo group sounded more deeply immersed in reverberant mush, obscuring some of the words. The Corona caught the echo and words pretty close to ideally. Finally, there was some overlaid shatter distortion, which I hadn’t noticed before.
Finally, “Snilla Patea” sounded as realistic as I’ve heard it. The fiddle harmonics were perfect, the chorus sounded like very well-trained vocalists singing together, capable of startling dynamic shifts. Wow!
So does the Corona represent a good value for its high price? Unfortunately, I had no similarly-priced integrated amp for comparison, but the Corona was quite close in performance to the slightly more expensive Audio Research amp and preamp, which are highly regarded, so I think its value must be excellent. And it has some technical features, like the very useful remote control, that surpass the Audio Research duo.
Let’s review the positive and negative factors for the Rogers KWM-88 Corona integrated amplifier. Appearance—outstanding. The red chassis adds a touch of class. Technology—very advanced. Uses cell phone or tablet app for remote, and an automatic bias circuit to manage tube settings. Sound—very high quality. Best bass I’ve heard from a tube amplifier, fantastic dynamics, very good harmonic accuracy. Size and weight—heavy but manageable. (Sooner or later this will be important to you.) Ease of operation—extremely easy to operate. Controls are a piece of cake; the remote app running on a cell phone makes remote operation easier than ever and informative. Someday all remotes will work this way, but the Corona had it first. Build-quality—dreadnought construction. Warranty—best in the industry. In sum, the Corona is expensive but worth it.
Specs & Pricing
Power output: Ultralinear 100W RMS, triode 80W RMS
Tube complement: 2x 6SJ7, 2x 6SN7, 4x KT88 or KT150
Frequency response: ±0.1dB from 20Hz to 20kHz
Input/output impedances: Four unbalanced, 100k ohms/2 ohms to 32 ohms
Weight: 60 lbs.
Dimensions: 17″ x 14″ x 11.5″
ROGERS HIGH FIDELITY
10 Holden Street
North Adams, MA 01247
Speakers: Affirm Audio Lumination speakers
Amplifiers: Audio Research VT80SE stereo amplifier
Preamplifier: Audio Research LS28 linestage Digital sources”Dell Latitude E6330 laptop computer running 64-bit Windows 10 Professional and Roon music server software version 1.6; PS Audio DirectStream DAC
Interconnects: Van den Hul The Mountain balanced interconnects, High Fidelity Cables CT-1 Ultimate interconnects
Speaker cables: Van den Hul The Cloud speaker cables
Power cords: Purist Audio Design Venustas power cords, Blue Marble Audio Blue Lightning power cords, Clarity Cables Vortex power cords, Audience powerChord e, Au24 SE LP powerChord power cord
Digital cables: Paul Pang TZ YUN Red II USB Cable, Audience Au24 SE USB cable
Power conditioners: Shunyata Denali, Audience aR6-T