Articles for Audiophiles by Steve Deckert
AUDIO PAPER #020
ZEN TRIODE AMP POWER MYSTERY UNRAVELED
by Steve Deckert
Consumers of hifi gear seem to gain all of their knowledge about it from magazines and sales people. Manufacturer's literature certainly being at the heart of it. That means the knowledge is more about what's popular than how it actually works or if it actually works in some cases.
You would think that the equipement pictured here would be superior to magazines and published specs for gaining the real facts yet it to suffers from the same miss-understood interpretations. I am pretty sure any electronics engineer will be happy to tell you that using ohm's law, watts are watts be it solid-state using pulse-width modulated power supplies or tubes using linear power supplies. The reality is that watts are not really watts in audio amplification because there are so many variables that determine what each watt can accomplish. High efficiency solid state watts and tube watts are usually an apples and oranges comparison.
It pains me to see that every day someone makes a purchasing decision based on specifications. Besides RMS power, I'd have to say that THD is by far the worst overblown subject in audio. Many people who call me haven't had the opportunity to realize that there are two basic forms of distortion encompassed by the spec: THD, and that one can actually sound good. They don't realize that tubes are very different than solid state devices and that distortion and power output are tightly linked.
When people discuss watts, they equate it to how loud something will get without audible distortion. How loud an amplifier gets compared to another depends on its ability to drive a given speaker load. It is possible to have a 10 watt amplifier get a pair of speakers louder than a 100 watt amplifier based on this reality.
ZEN TRIODE POWER RATINGS
The Zen Triode for those who don't know, is not an ordinary amplifier. It's sole purpose in life is to see how pure we can reproduce sound. That is the reason for its size and power output. It's a classic example of the phrase "Less is more!"
Based on the observation that the less parts used in an audio circuit the less coloration the circuit has... And based on the observations that more powerful amplifiers require and use bigger parts and generally more complex circuits - it's easy to conclude that larger amplifiers while having higher power numbers will also have higher coloration's.
Figure 1 - If you made an imaginary line chart with Fidelity on one end and Power on the other, you could design an amp to fit anywhere in that line, but regardless of where you chose to put it, you would be balancing between the two. That's a classic example of the phrase "Good sound is a balance of compromises" Too far on the fidelity end and you don't have enough power to drive anything, hence it's useless. Too far on the power end and you no longer have a class A high fidelity amplifier.
I found during the development of the Zen Triode Amp that there is a magic spot on the line of power vs. fidelity where the balance of compromise is minimal. Guess where on the line the Zen Triode Amp sits?
ZEN TRIODE RMS POWER OUTPUT
When this amp was first released we listed the Zen Triode is around 5 watts per channel. Since then competitors have jumped at the opportunity to interpret this as RMS (Root Mean Square) power output. After all that has become the standard way to determine the power of most amplifiers albeit far less meaningful in the tube world. In fact around here (in the tube world) and in the spirit of good Zen thinking we found that RMS really stands for "Real Meaningless Spec". It would certainly go against the theme of our little amplifier to post a lot of real meaningless specs when we can instead post a lot of real meaningful reviews.
It is not my intention to apologize for misleading the public with a 5 watt rating, because it was never stated to be RMS power. RMS power is what you read using a 1000 cycle pure tone into a non inductive "dummy load" consisting of large silicon resistors. This makes a wonderful reference to the designer when comparing one amp to another, or modifications within a single amp. However it makes a rather pathetic reference of how music will sound through it or how loud it will get into a given loudspeaker load.
Obviously music is not a pure 1000 cycle tone, and speakers are not large non inductive dummy resistors. Our rating of 5 watts is a peak power rating that takes into account things like: Peaks in the music and the amplifier's ability to generate power peaks in excess of it's RMS abilities not to mention distortion signatures and clipping characteristics. It also takes into account the fact that speakers are ever changing complex loads that behave very differently than "dummy loads". We felt based on the less than clear common interpretation of RMS watts that a more realistic "music power" rating was in order.
In the picture above you can see we have the gear to get anal about specifications, but have always chosen not to complicate our sales approach or your mind with excessive specification information which we feel is commonly misinterpreted. Figure 2 shows the RMS output of a single channel of our Zen Triode Amp to be at 2.38 watts just before clipping. When the sign wave starts to flatten to a point where the listener can detect unpleasant distortion the unit reads around 3.40 watts RMS.
WHY DIFFERENT RESULTS FROM DIFFERENT PEOPLE?
If you have been enjoying measuring the RMS power of the Zen Triode amplifier, please be aware that output tubes (all of them) vary in emission. A matched pair of tubes simply means they match each other. It doesn't indicate if the tubes tested strong or weak. The difference between a weak and a strong tube in the case of the 6Bq5's and or SV83's can be enough to vary the reading by almost a watt. Additionally the voltage drop across the rectifier and the voltage at the wall outlet all make a substantial difference. Of course when you measure RMS power you have to pick a level of harmonic distortion that is acceptable for the design.
For the RMS tests pictured on this page, we used a matched pair of tubes (only one of which is measured at a time during the RMS testing) that tested at 7500 micromhos. This is done on our Hickok model 539a tube tester with the bias setting at 2 volts. We match all our tubes here in direct micromhos using this classic military tester. 7500 is about ave. for the SV83's that we've tested (around 750). Some are as low as 4840 and some as high as 9750.
So you can see just the tubes themselves can account for differences in RMS power measurement of a Zen Triode Amp.
SPEAKER LOADS ~ IMPEDANCE and SET's
There are some stereotypes of how an amplifier drives a loudspeaker that relate specifically to the impedance curve of the speaker. The stereotype is that most amplifiers, especially low power single ended tube amps, have difficulty driving low impedance loads. In fact this has gotten so overblown that speaker manufactures are now listing the minimum impedance right on the specs. Without fail everyone that calls me about possible speaker prospects is concerned that they're speaker selection may not have a friendly impedance curve. Well, friendly to who? The Zen Triode amplifier is not like other amplifiers, if it was it wouldn't carry the name Zen in the title. The Zen Triode is perfectly happy with low impedance loads and will drive nearly a dead short.
This is certainly against the grain, especially for single ended tube gear, but was by no means an accident. Since one of the biggest problems with low power single ended amplifiers is their ability to drive difficult or moderately difficult loads, it's no wonder that the SET market is so small and overlooked. It would seem to me that when you have only small amounts of power to work with, you certainly wouldn't want to handicap yourself further by having the amplifier's power curve suffer at the low end of a speakers impedance. Doing so would mean that at a speakers resonance, where the impedance is high, the amplifier would have no problems or even put out more power. But when it drops the amp begins to have trouble.
Ideally, the amp would put the same power at any given frequency and the speaker would have the same impedance at any given frequency but this is impossible. Industry has focused on this for many years creating high power amplifiers with lots of negative feedback and high damping factors to make them "NOT FEEL" the speaker and thereby not be so affected by its impedance. Speaker manufactures have also focused on this by creating complex crossovers with Zobel impedance matching networks that yield a flatter impedance over the frequency bandwidth of the speaker. The problem with this is POWER. To do it requires lots of power. We already know that high power amplifiers are not as pure on the fidelity line and have more colorations and less detail. Speakers with overly complex networks become grossly inefficient and thereby require lots of power just to sound good. If you throw only 1 watt at these units, some of it is spent as heat in the crossover before it even gets to the voice coil of the speakers.
In the world of Single Ended Triodes, it's all about the first watt, and the magic that lies there. This is why when used with conventional speakers most SET amps fall short of delivering a good performance. The pairing is not unlike trying to ask your car to start after pouring orange juice into the gas tank instead of gasoline. My solution to this problem was to do things backwards and design an amplifier that put out more power into low impedance than into a high impedance. This approach certainly won't solve all the obstacles of pairing single ended triode amps with speakers, but it will make it possible for most anyone to find out what single ended is all about.
After marketing the Zen Triode for 2 years, 90% of our customers were able to use a Zen amp with their existing speakers. Speakers that were designed for today's solid state power houses. And while I'm sure none of them get LOUD like a big amplifier, I do know that I was successful in letting these people not only hear what single ended triodes were like, but gave them one that sounded better than similar units many times the price. That by the way, is what gives me the most pleasure, and why I enjoy my work so much!
When I said I wanted to "Design an amp for the masses", this was what I meant.
HOW DOES IT WORK?
The primary thing that makes the SE84B unique is that it was not designed with a calculator to effect the max. power. In other words there was no greed involved when balancing power output against fidelity. A math guy will always match the output transformers primary impedance against the plate curves of the output tube for max power and lowest distortion.
In fact we had to make our own power transformers to come up with the right impedance for our needs. Our custom hand wound air-gapped grain oriented silicon steel output transformers have a primary impedance of 9800 ohms and the EL84 / SV83 output tubes would rather see 1/2 to 1/3 of that. This combined with a single 6 ohm tap gives ideal clipping characteristics on the average speaker load. Lets face it, 8 ohm speakers are often closer to 6 ohm and with a six ohm tap the Zen can drive either 8 or 4 ohm speakers. The clipping characteristics of this design are so graceful that you can't hear when the amp is soft clipping. It takes a hard clip before you can actually hear it.
It is this gracious clipping and lower distortion from the output transformer that makes it hard to give a power rating that means the same thing to everyone. If you were to measure it like a solid state amplifier at .001 % distortion, you would have below a watt. When the engineer measures the amp, it is his interpretation of the waveform on the scope and or the distortion threshold he sets that determines the RMS output. Consumers used to hearing that distortion is a bad thing from large solid state manufactures have been grossly mislead to assume that this would apply also to tube gear. I would guess that most people who listen to a good SET amp, Zen included, are pushing between 5 and 10% distortion and don't even know it. Reason: Even order harmonic distortion sounds good, is more natural. Like playing a two finger chord on the piano. Odd order harmonic distortion (the type solid state circuits generate when clipping) is the exact opposite. Nothing good or pleasing sounding about it. In fact it rather sounds like sandpaper going through a meat grinder. Or like playing the wrong notes on the piano. It creates ear fatigue by its very design because nothing God created on this earth related to music that I can find creates odd order harmonic distortion naturally.
In the power tests using the same output tube, the Zen Triode amp performed with higher RMS output into low impedance loads. For example, a 16 ohm load yielded just under or at 1 watt RMS before measurable (not audible) distortion. (The waveform shown in the picture) An 8 ohm load yielded around 1.5 to 1.9 watts RMS. A 4 ohm load gave us around 2.38 watts RMS and a 2 ohm load gave us around 3.21 watts RMS. In the same tests, the amplifier was bridged using our series bridging technique made possible by floating the outputs. In this configuration the now mono amplifier exactly doubled all the figures above.
The importance of audio specs are overblown. Specs for solid state and tube amps don't give the consumer even a hint of an idea what the unit will sound like or how loud it will actually get before the sound stops getting better as you turn it up and starts getting worse. RMS to the end user i.e.. the consumer of tube gear means "Real Meaningless Spec" and worse than that is THD, and worse than that is DAMPING factor. Please don't let the math guys trip you up with rigid specifications because the slightest tweak of the knob on our test gear (and theirs) can change the readings - making it largely open to interpretation.
Remember RMS does not account for Peak Power, Instantaneous Peak Power, Clipping characteristics and the distortion level where RMS is measured is open to the guy doing the measurements. The industry who has been in a who has the lowest distortion competition for the past 70 years has completely missed the boat on sound quality. If someone is telling you the Zen amp is only a watt, smile and say "well if that were actually true, then it makes it even more amazing that it sounds so much better than yours."