Schiit Audio Tyr Monoblock Power Amplifiers

Amplifiers are, however, closer to being transducers than the other electronic components in the chain, because amplifiers are not pure voltage devices. Up to the amplifier, one is not really passing on power. There has to be a little power involved, but something like a preamp or a DAC is essentially a voltage generator, not a power generator. In technical terms, this means that the next stage has a very high input impedance, so that even fairly large voltages do not call for much current in order to maintain themselves. When a preamp puts out say one volt, it is putting out almost no power, because the 1 volt is terminated by the input impedance of the amplifier, which is typically tens of thousands of ohms. The stages before the amplifier are like information transmitters—voltage information, but next to no power. The power amplifier has a different job. It amplifies voltage all right, typically by about a factor of 20 or so. But it has to maintain its output voltage signal into a rather low-impedance load, a speaker—so there is a lot of current and quite a lot of power required.

You can think of this by analogy with pressure. When you press your fingertip against a rigid object, say a brick wall, you can exert a good deal of pressure without the wall moving. If there were a handle in the brick wall, you could pull back and forth, exerting pressure alternately in both directions but with nothing much moving and without you really doing any work. But if you pull back and forth on a wheelbarrow of bricks, the bricks move back and forth. And it takes a lot of power to maintain the motion. This is the difference between applying voltage to a really high impedance with essentially no current flowing (pulling back and forth on the handle attached to the brick wall, which won’t move), and applying voltage to a low impedance where impedance is low and current flows and power is transferred (the handles on a wheelbarrow). In short, this is why power amplifiers are tough to make: They are going to do work in the physics sense of the word.

Electric power is tricky. You cannot see it! So how much power are we talking about? The Tyr (one Tyr monoblock) puts out potentially 350 watts into 4 ohms. How much power is that intuitively, anyway? To put this in an everyday perspective, it is about the same as the power of a typical electric motor for an electric scooter of the kind that teenagers are whizzing around your neighborhood on. We are speaking now of power in the macroscopic appliance sense, not the sense of, say, computer electronics.

Now there is a curious thing here: A rather small fraction of this power generated by the amplifier, any amplifier, actually emerges from the speaker as sound power. Typical dynamic- driver speakers are only about 3% efficient or thereabouts. This is one reason why some people are interested in horn speakers: They convert more of their electric power input into sound, so they do not need as much driving power as do ordinary speakers.

So, how much power does one use in practice? I suppose everyone knows that I am not a fan of extremely high levels, but even so, some years ago when I had an amplifier around with peak-level meters, I discovered that on orchestral music at a moderately realistic level with speakers of mid-80s sensitivity, I was hitting close to 200 watts on big bass transients. No wonder MBL makes an amplifier with more than 800 watts of power available to drive their 81dB-sensitive Radialstrahlers! You do not need a lot of power all the time, but you need more than you perhaps realize some of the time, for short intervals. Even with speakers with reasonable sensitivity levels, power counts.

At the same time, an amplifier needs to have very low distortion to sound pure at moderate or low signal levels, the levels that are occurring most of the time in music. The proverbial “inner detail” typically involves output signals at the 10-milliwatt level or even lower. So, one is looking in effect for the strength of an NFL lineman with the delicacy of movement of a neurosurgeon.

Enter the Tyrs

The Tyrs have got this covered! Not to put to fine a point on the general conclusion, these amplifiers sound super, so good that one is hard pressed to find anything wrong, at all. Or at least, I was. Whatever was missing from the musical experience came from the recordings or the speakers or the room around. I could not point a finger at anything that the amplifier should have been doing and wasn’t. Expressed the other way around, the Tyrs were letting through all the good things on recordings and getting the speakers to show their best.

In this gestalt sense, I was perfectly happy. But, of course, it is a convention of audio reviewing that one is supposed to analyze the unified experience into parts and evaluate things as separate items and even to speculate (which is usually all it is!) on the reasons for these parted-out behaviors. So, here goes.

For a start, there was a sense of ease that only high-power amplifiers give in my experience. The Tyrs never seemed stressed by anything I threw at them. Of course, every amplifier has limits, and in a large-scale live environment, enormous amounts of power can be relevant. (I understand the Wall of Sound system of the Grateful Dead in the mid-1970s used over 25,000 watts of amplification power.) But back in the world of domestic audio, I think the Tyrs, with their 200 watts into 8 ohms and 350 into 4, used with speakers of plausible sensitivity, are going to do the job for you with ease.(Exceptions might be speakers with capacitance loading of an extreme sort at the top end—electrostatics or speakers with extremely low but non-capacitive impedance or those with super-low sensitivity might be a problem.) For those extreme cases, one might want to try the appropriate Sanders amplifier, for considerably more money. But the Tyrs surely did the job for me. No stress and strain at all with the various dynamic-driver speakers I tried.

Does this large power window matter? In one theory, if an amplifier does not reach clipping level, it does not “know,” so to speak, that it is near clipping level. But as noted earlier, even moderate levels can make momentary large power demands. And in addition, psychologically, it gives listening ease to know that when the music starts to get loud, there are not going to be any amplifier issues. One does not want to sit there waiting for, listening for, hints of clipping behavior. I am no Death Metal headbanger, but I liked the power.

Moreover, and perhaps even more crucially than the ease of large power capacity, the sound of the Tyr is intrinsically very pure and, one is tempted to say, sweet at all levels, including low ones. Calling the Tyrs sweet-sounding is misleading if it suggests they are doing something in the way of altering the signal coming in. They are not. But sweet music coming in will come out sweet at the output. And if like me you have looked for recordings that ought to sound musically attractive in the top end, then the Tyr will let them do exactly that when you listen to them. Designer Jason Stoddard is, I gathered in our conversation, of the impression that the choke power supply (an unusual and “retro” feature) may have something to do with this. Or it might also have something to do with the particular output network used (to be discussed later). Or maybe something else. But in any case, there it is in the listening experience.

The Tyrs offer striking power and definition in the bass and middle bass. With the bass-capable Cabasse Murano Alto speakers, the Reference Recordings Rutter Requiem’s organ part was extraordinarily convincing. And the low plucked notes on Eight String Cello Religion positively exploded out of the speakers if the volume was up. Orchestral music had scale and scope: The opening of the ProArte Mahler Second (Dallas Symphony, Mata conducting) was properly hair-raising. The whomps in the opening of the ProArte Dallas (Mata again) recording of the Rachmaninoff Symphonic Dances really whomped.

And yet when the music turns delicate, the Tyrs go right along and present delicate fine detail in a most graceful form. Using the Jern speakers, with their remarkably silent (cast-iron) enclosures, the third movement of Kind of Blue, say, sounded the very definition of sonic refinement. And back to the Mahler Second, the Laendler (second) movement’s delicate string sound was exquisite in its realism.

And the Tyrs, with their low distortion and absolutely minimal deviation from accurate response, do an unexcelled job with human voice. Jane Monheit’s “I’ll be Around” (from Monheit at Home) was just as perfect as whatever speaker you are using will allow. In fact, the Tyrs make the best of everything in the sense of letting the whole sonic picture through intact. The recorded signal arrives at your speakers exactly as it should, and if your speakers are ready for it, there is nothing that will go wrong. In Douglas Self’s word, the Tyrs are really “blameless,” which means, in practice, that they sound fabulous—with material that deserves it and speakers that can do it. Little things about amplification that could potentially take away from the beauty of the picture are just not there, and indeed the Tyrs may reveal to you that there were such things, such micro-defects, in your previous listening that you did not know were there, until, with the Tyrs, they weren’t there anymore.

Silence and Resolution

Back in the “objective” world, the Tyrs are really quiet. They have an A-weighed signal-to-noise rating of >120dB down from full output. Some arithmetic is useful here: Full output into 8 ohms is 200 watts, 23dB above the one-watt level. So, the noise level is more than 120-23 = 97dB below the one-watt signal level. If you speakers have, say, 87dB sensitivity, then the noise has absolute level lower than –10dBa. This will not be heard in a room environment. In fact, this is below the threshold of hearing at any frequency even in an entirely quiet environment. (With extremely high-sensitivity speakers, there might be noise that would be marginally audible in an anechoic, totally silent environment, but in a real-world room, it would still not be heard, being masked by room noise.)

And this is true in practice Without an input, the Tyr produces no audible noise at all. When one adds in the fact that even a quiet listening room has a background noise level of 20dBa (that is a really quiet room, actually), the noise is even further below the audibility level. I read a review recently where the reviewer said that his listening room had a noise level of 30dBa (which is still very quiet), so he was going to hear any noise that there was to hear. In practical terms, that has some element of truth. But clearly the noise of low-noise amplifiers like the Tyrs is gone, gone, gone compared to that. In fact, getting quieter than the Tyrs is just “specsmanship.” Incidentally, the A weighting is a system that is not set up to paper over anything but rather simply reflects the fact (which is true for everyone) that noise at the frequency extremes (especially in the bottom) is less audible than in the middle frequencies. This is quite an extreme effect: the threshold of hearing at 20Hz is over 70dBa (cf., wikipedia.org/wiki/Equal-loudness_contour).

This low noise has something to do with perceived resolution. In speakers, extremely low-level signals can be “eaten.” Because of stick/slip behavior, a dynamic driver can fail to respond to a very low-level signal. I wrote about this in detail in my review of the DALI Grand speakers many years ago, which have linear response to –10dBa or lower. These speakers have real resolution in a sense that few do, not the kind of perceived but unreal resolution that speakers with lots of little resonances in the mid and higher frequencies can generate, e.g., in some speakers with certain kinds of resonances or with exaggerated top ends, which remain all too common. (This seems to be a tendency which never dies.)

But properly biased electronics are not like this. Even an older classic like the Hafler DH200 shows the story: Look in wikipedia.org/wiki/Equal-loudness_contour at the behavior at the Hafler DH200 at a 500-nanowatt level. The distortion is still 100dB down from the truly low-level signal. What happens with amplifiers that work like this (that is, correctly) is that as the input signal drops, the associated output eventually just sinks into the noise floor. So, low-noise, properly behaving amplifiers resolve low-level signals with lower noise—down to the level where the noise is inaudible.

This points up the difference between analog and undithered digital. In undithered digital, the decreasing signal becomes not just lost in noise but actively distorted. This is why people tended to hate undithered digital—it turned to garbage at low levels. But a Class A (or strongly biased Class AB) amplifier does not do this at all. Distortion does not rise as level goes down. And a fading sound smoothly fades out into the noise floor (which is very low in the Tyrs).

This short excursion into engineering and information theory is borne out in practice. The Tyr presents fine detail perfectly. Things like brushed cymbals sound exactly right. And low-level audience noises in live recordings (e.g., the coughs in Track 5 of Favorite Encores of Ofra Harnoy) are resolved as well as your speakers will allow. Same thing with the low-level bleed-through of helper notes on the Jacintha “Moon River” recording (those that PS likes to talk about). I venture to say that anything you have ever heard on a recording with any amplifier, you will hear with the Tyrs. One has to be cautious here: Every time you listen to a recording, no matter how familiar, you are likely to notice new things. The real test is whether those things, once heard with a new piece of equipment, are still there when you go back to other pieces of equipment.

Inductance, Upper-Frequency Behavior, and Output Networks

The Tyrs are wide-bandwidth amplifiers, with a –3dB down point in the top end of 500kHz (manufacturer’s specification). This is wide enough to minimize almost completely any effect of the top-end roll-off in audible terms without calling for large amounts of extremely fast negative feedback (the Tyrs actually have a relatively low level of negative feedback, between 30 and 40dB, according to designer Stoddard). 

A certain considerable amount of bandwidth is needed, but too much can potentially cause trouble. Hyper-fast amplifiers that amplify well up into the radio-frequency band can have stability issues from radio frequencies in the world around, for one thing. (I had such an amplifier at one point that played a local radio station with no radio being present.) The Tyrs have enough bandwidth to sound really right but were, in my experience, very stable and seemed entirely insensitive to radio-frequency signals.

The Tyrs have a small inductor for stability in their output network, but small it is, according to Stoddard again, under 0.5 microHenries. And this low inductance is actually maintainable in practice. Namely, because the Tyrs are monoblocks, one can use very short cables. This point seems to be overlooked all too often. The inductance of your speaker cables is in effect part of the output network of the amplifier. And cables have serious amounts of inductance, even the ones that are low inductance nominally—a typical figure is 0.15mH per foot. So, a six-foot cable would contribute 0.9mH—more than the output network of many an amplifier itself. All of a sudden, your 500kHz-bandwidth amplifier has been cut down to a considerably lower bandwidth associated with a larger inductor on the output. (The Quad 606, which was sometimes regarded as too band-limited for ideal sonics, had an inductive component in its output impedance of 1.5mH. So, 0.9 from cable + 0.5 from Tyr equals 1.4 microHenries, tied effectively with 1.5mH with no cables. Makes you think, right?)

Cable inductance comes in per-foot units, so and so many microHenries per foot. Really short cables have really low inductance. If one arranges super-low inductance per foot, this makes capacitance go up by theoretical necessity. And this type of load can present a problem for amplifiers. But capacitance comes by the foot, too, so a really short cable can have very low inductance and low enough capacitance that the amplifier is unbothered. Then you need long interconnects, but since interconnects live in the world of pure voltage devices, they present fewer problems. Short cables and long interconnects are the way to go.

This kind of riffing on engineering can be annoying and disturbing. But it does matter! Actually, there is some reason to believe that it is one of the main things that makes low-distortion amplifiers sound different. The noise and distortion are often under the thresholds of audibility, but the variations in top-end frequency response, while small compared to speaker variations, are, even so, often above the known thresholds. So, with a fixed speaker, these variations may well be and, to my mind, likely are main contributors to perceived sonic differences. Research done of this matter (e.g., by James Croft, personal communication) has suggested, for instance, that slight alterations in the response of an amplifier in the 4–8kHz region affect perception of transients, with a little extra 4kHz favoring pianos and 7kHz perceptually enhancing cymbals, at levels of response-shift that are very close to thresholds of perception and are almost surely below affecting overt tonal character.

It is possible to make an amplifier with no output inductor, although this is seldom attempted. The Carver A500x, designed by James Croft, along with Brian Aase and Vic Richardson, was one such, and it presented a flat treble response all the way out to 20kHz, flat to considerably better than the threshold of detection. It seems this was appreciated by few people at the time. But there were some. Ed Foster reviewed it enthusiastically for Audio magazine, and I used A series amplifiers for years as a reference for speaker-reviewing purposes, especially for top-end behavior.

The Tyrs with short cables are very close to this effect of true neutrality, so close that the issue is all but moot—something that is not the case with most amplifiers, I think. In a direct comparison, using actual speakers, not resistive test loads, there can seem to be a subtle, almost subliminal difference from what I perceive as the totally neutral response of the A500X, involving a microscopic dip in the presence region, but this is subliminal in musical terms—if it is even there at all! (This is down at the level where such things are very hard to verify definitively, down at the same level of slight changes in listener position.) I was reluctant even to mention this because audiophiles often find it impossible to believe that an effect could be real without being important. But things very close to threshold level like this may be observable under certain very controlled artificial circumstances (like not moving your head at all) but in musical terms and in common use completely insignificant.

But if you use long cables with substantial inductance—on anything—be aware that you are probably not hearing the wide-bandwidth amplifier (if such it is) but a bandwidth much reduced by the cables. (I wrote an article long ago in Issue 71 about this, with numbers on the effects of output impedance, but it seems to have passed the audio public by, even though these matters are well known in theory. The tendency to think of an amplifier as a thing unto itself with its bandwidth unchanged by external circumstances is apparently hard to get rid of.)

So, what does this—the low-inductance, high-bandwidth, plus short cables setup—sound like? It sounds very smooth and to my ears very truthful! The little presence range edge that sometimes appears is gone. The treble (assuming the speakers are good) just floats there as it should. All is well. Such effects are subtle, but they are apparently enough to change people’s reactions to the sound. The cable market is after all lively! And to my ears, yes, it is worth getting right. In a quick A/B with unfamiliar speakers and material, perhaps it would be hard to detect, but in long-term listening with familiar material and speakers, it is likely to count for more.

Amplifier History and Soundstage

The beginning of high-end audio was practically synonymous with an emphasis on the importance of power amplifiers. The introduction of transistor amplification had led to amplifiers which measured well according to the distortion-at-full-power measurements that had been more appropriate for tube amplifiers, and these often did not sound very good because of high crossover distortion at low levels.

Recognizing this by listening became the badge of honor of high-end reviewers, and quite rightly so. But new and less convincing aspects of the situation arose. In particular, the notion was voiced by Harry Pearson in particular, though it was seldom entirely explicitly formulated, that the amplifier was somehow supposed to create the “soundstage,” not just let whatever stereo information there was on the recording through to the speakers without alteration. This idea has somewhat faded out, but not entirely: One still sees online references to amplifiers “throwing a soundstage,” even though this idea is surely peculiar, at least as stated.

What does happen, at least in principle, is that stereo information at low levels—not direct stereo images as such but rather the micro-structure of recorded room-reflections, for example—could be lost in a high-distortion or high-noise amplifier. But this will not happen in an amplifier like the Tyr. And, indeed, recordings that deserve to have a lot of depth of image, for instance, most definitely do. The instruments at the back of the stage in The Swan of Tuonela of Sibelius on Reference Recordings Reveries CD are very far back, indeed. And width is equally well served. To my mind, this is practically an inevitability with an unstressed low-noise, low-distortion amplifier arrangement with total channel separation, but in any case, here it is. All the space that there is on the recording (and that the speaker/room arrangement allows) is there. (The Swan of T recording mentioned is really something. Don’t miss it.)

Be not Blinded by the Price

There is a certain difficulty in encountering the Tyrs as review items. There is what amounts to a convention in reviewing devices of medium price that one is supposed to conclude a review with a tribute to the quality offered but add what one could get more for more money. I am afraid this is a convention I am finding impossible to follow. There are theoretical ways in which the Tyrs could be better: Any specification can always be improved. But in musical listening terms, there are thresholds. And my experience is that in those musical terms there was really nothing I would change. It invariably happened that whatever musically meaningful restriction there was in the sound I was hearing came from the recording or the speaker or speaker/room combination. The Tyrs always let the musical experience through as well as the recording permitted, provided the recording itself and the room/speaker combination allowed it to be reproduced. Assuming, as will be the case in almost all situations, that the Tyrs had enough power, the faults lay elsewhere to my ears.

Now people differ, and longtime TAS readers will know that I am a person who blames speakers and recordings more than electronics quite generally. Perhaps a person who is obsessively interested in the behavior of power amplifiers might find some severe fault or another with the Tyrs. But I did not, and indeed even at the level of subliminal, personal, and perhaps not-transferrable-to-others impressions, I found the Tyrs to match the performance of high-powered Class A amplifiers, which in the past have been my subjective favorites—but without the practical inconveniences of high-power Class A amplifiers.

The best contemporary amplifiers tend to sound rather similar—they are all quite good and hence are quite similar, perfect amplifiers all sounding alike, or so I would suppose. And I would not want to claim that I might not have some similar feelings about some other amplifiers. But I am unaware of any other of this sort that has as much power as the Tyrs but does not cost a lot more money. In the context that I am somewhat of a speaker person, you might think, well, REG is just not all that observant of all the things that are bothering us about amplifiers. And I cannot prove that that is not so. Indeed, it is surely the case that far more speakers strike me as seriously defective and even somewhat goofy than do amplifiers. But I can only report my own experiences. In any case, there is a way for you to find out.

Get a hold of a pair of Tyrs and compare it with any high-priced amplifier that you can. But for this, I would suggest, in a modest proposal, as it were, that you arrange that you do not know which one you are listening to. There is an enormous amount of evidence that if you know which one you are hearing and know the prices, this will distort your impressions of quality. A striking example in my other field of work of this kind of thing is the study which showed that mathematicians, who to a person all firmly believe in gender equality in mathematics, even so in practice statistically rate mathematics papers lower if they know they are written by women than they do if they are asked to evaluate them without the gender of the author being revealed. We can be prejudiced even when we are sure we are not.

I am not speaking here of quick A/B comparisons of amplifiers, which can arguably be misleading and confusing, but rather of long-term listening. It is not hard to set up a situation of listening to different amplifiers at leisure without knowing which one you are hearing. You just need someone to help. Under these circumstances, I would honestly be surprised if the Tyrs did not prove the musical equal of anything at all to your ears. But you should please try for yourself before you decide whether this could not be true. I cannot help wondering, though, how many people will do this! (Cruising around on the internet surely suggests that there are a lot of people who want to have opinions on amplifiers without trying hard to substantiate these opinions in listening terms.)

Regardless of the outcome of that for you, I think everyone will agree without demur that at the very least the Tyrs are a remarkable item at their price. And by the way, I personally liked their engineering presence, so to speak. Their somewhat “retro” approach to power supply (via chokes), their weight and solidity, their battleship build-quality, which suggest longevity in use., and their careful monitoring of operating conditions internally, all added up to a very attractive picture, in addition to the superb sound quality itself.

Cheers all around for Jason Stoddard and the Schiit team.

P.S. Incidentally, Schiit products are sold by direct online order unless you happen to be able to visit the Schiit store in Southern California (address in the Specs & Pricing sidebar). But they have a 15-day return policy. I don’t think very many will come back.

Specs & Pricing

Type: monoblock amplifier
Topology: Nexus discrete differential current feedback, no coupling capacitors or DC servos, continuity transconductance output stage
Power output: 200W into 8 ohms; 350W into 4 ohms
Frequency response: 20Hz–20kHz –0.1dB, –3dB 500kHz
Total harmonic distortion: <0.01 % at 200 watts RMS 8 ohm
Intermodulation distortion: <0.01 % CCIR 200 watts into 8 ohms
Signal–to–noise ratio: >120dB referenced to full output
Damping factor: >200, 20Hz–20kHz
Gain: 22 (28dB)
Input impedance: 47k ohms single-ended, 47k ohms balanced
Size: 16″ x 3.875″ x 12″
Weight: 55 lbs.
Price: $1599 (per channel, two required)

SCHIIT AUDIO
The Schiitr (store)
22508 Market Street
Newhall, CA 91321
schiit.com