The Weak Link

The major problem with hollow state is that the vacuum tube is a high voltage, low current -- and therefore Hi-Z -- device. Electromechanical speakers are just the opposite: requiring high currents at relatively low voltages. The usual solution to this problem is the impedance matching transformer. Unfortunately, no transformer that has a ferromagnetic core is a linear device. Air core transformers that can operate at audio frequencies are not exactly efficient, and what you gain in core linearity you promptly lose in the enormous stray capacitance that the huge number of turns such a thing would require.

So the ideal situation would be to get rid of that transformer completely. There are a couple of factors that go against this: the Hi-Z nature of the device itself, and the fact that there is no such thing as a "P-Channel" VT. Given that, you are limited to some sort of "totem pole" output stage. Here, the problem is the lack of symmetry: one half will be acting as a grounded cathode amp with its characteristic voltage gain, while the other half will be, more or less, working as a cathode follower with its less than unity voltage gain. There have been various attempts to deal with this problem such as asymmetrical drive -- the SEPP output, or the Futterman variations that include feedback to either make the cathode follower half behave more like the grounded cathode half, or the inverted Futterman that applies NFB to the grounded cathode half to make it look more like a cathode follower.

There is another topology that at least takes care of the balance problem: the Wiggins Circlotron design, this being a bridged output stage. Even though originated for use with an OPT for the purpose of eliminating Class AB switching transients from the primary, it also makes for a very good OTL design as well.

The other problem remains: the low current nature of the VT itself. You really don't have much choice here, so far as possible types are concerned.

There have been VTs that were intended to give some serious output currents, these being the horizontal deflection finals. Even though the screen voltage specs limit the screen voltage to rather low levels, these can still pull some serious current at lowish V_PK's. Still, you are going to have to parallel up individual units in order to get the necessary output currents. The 6BQ6GTB can easily pull some 350mA of peak current when the plate and screen both operate at 150V_DC, if you go a bit into Class AB₂. For 30W of output into 8R, it will still require 16 6BQ6GTBs (8 per phase). That's a bit too much.

Other possible types would include the 6C33C (Series pass regulator used on the MiG fighter). There have been a lot of problems concerning this type, as its reliability is questionable, even with the special "burn-in" trick. Also, the pins are way too thin for the heater current requirement. MiG technicians replace not just the 6C33Cs, but also the socket as part of routine maintenance. I would avoid it.

The most likely type is the 6AS7 (and its relatives: 6080, industrial type, or the 6082, 26.5V heater version) -- another type originated as a series pass regulator tube. As such, it has the high current capability with reasonable V_PK's, and it is also a dual triode, so that you require half the number of bottles. Though its plate characteristics aren't spectacular, they aren't half bad either. It's not likely you'd want to use this in a conventional output circuit as there are better power triodes for that. Still, the 6AS7 can pull significant currents without taking the control grid positive. Eight 6AS7s (16 sections, 8 per phase) can get you 30W of output without serious spec busting. The type is also well known for having a high g_m and a very low r_p. The 6AS7 has been used to drive VT Class B audio output stages (AM plate modulators) for a very long time now.

Does getting rid of the OPT really represent an improvement in overall linearity? What you might gain in one area, you might lose back in another: operating the VTs into a very steep loadline (eight sections in parallel to drive 8 ohms is just 64 ohms per section). You might be just as well off in substituting MOSFETs for the hollow state output. There have been some complaints that paralleling finals leads to sonic degradation, since the individual VTs won't all be operating at precisely the same currents with the same exact characteristics. Whether or not this makes a difference is a whole 'nother story. At least the very low (for hollow state) impedance of the 6AS7 does help to mitigate this effect. This is another one of those areas where you just might have to try it to see and make up your own mind if it's worth it.

As for trying this, it is definitely something to take a shot at. It is especially helpful in that the 6AS7 is a current production VT, and at least in this case, the new production is superior to NOS offerings: better matching between sections, and a more robust design that will stand up to the slight spec-busting that can reduce the number of triode sections per phase. The 0.125A plate/cathode current rating is for DC regulator use. That peak current can be increased to 0.375A in audio final use. Music and voice programs won't be spending very much time at the maximum peak current. (It would be a different story for DC use or continuous max P_RMS applications. (If you were doing that, then consider the 0.125A spec to be a DC/RMS value.)

The only other requirements are DC offset detection/correction since you definitely want to keep DC out of speaker voice coils, fault detection in case one 6AS7 (or one section of a 6AS7) develops some fault that causes it to pull more current, and HV delay. Being that the intended purpose was series pass duty, the heater-to-cathode insulation is a good deal thicker to stand up to higher than normal voltages. This also slows down greatly the cathode warm-up, and the specs call for preheating to make sure the cathodes are at operating temp before hitting 'em with the HV.

The specs also strongly discourage the use of fixed bias. The problem of using the much more desirable fixed bias can be solved by mixed fixed/cathode bias, and over current detection. A 6AS7, OTL, Circlotron design is definitely "in the works".

The Numbers Game

In most of the electronics field, the characterization of
performance is pretty straight forward. If a video amplifier has
inadequate bandwidth that causes the loss of fine detail, and excessive phase nonlinearity that smears images across the screen, or puts out all the wrong colours, no one who isn't blind as a bat will disagree that it is a very poor design. If a servo doesn't track accurately, no one will argue about the finer points of that kind of distortion.

The audio field is something else entirely. Eyes don't lie, but ears deceive. Characterizing the performance of audio electronics is complicated by the very fickleness of the end user. What sounds just great to one will be judged inferior by another. Different people hear different things, and there may very well be those (the "golden ears") who hear things that the rest of us can not hear at all. Throw in musical preferences, and it gets even more complex. It's highly doubtful that the head-banger and the Classical aficionado will agree that the same amp is superior to other offerings.

The marketing department needs a selling point, and that has long been THD. The Williamson of 1948 made THD fashionable, and started the fallacy that gNFB was the cure for all ills. Between 1948 and 1956, the Williamson design was all but universal. The commercial "Big Box" manufacturers went for more power and ever greater feedback factors to drive those THD numbers ever lower. By 1960, Big Box products had excessive bandwidth (necessary for stability) and enormous feedback factors. UL EL34s in Class AB had become quite fashionable. The demand for more feedback led designers to abandon the lower distortion, lower gain types (6C5, 6J5, 6SN7, 6CG7) for high-u triodes and small signal pentodes. These high gain devices are less linear, adding more open loop distortion that was then corrected by high gNFB. Solid state was a natural progression since transistors have almost unlimited gain available, getting rid of the OPT removed an obstacle to driving up gNFB, and, of course, the gross nonlinearity of transistors demanded even more correction. This naturally led to a certain sloppiness in the open loop design. Why worry about poor open loop performance, add enough gNFB and you can sweep your mistakes under the carpet where no one will ever find them. Unfortunately, you also sweep away much of the musical detail.

The end result: sound-alike amps, with that lifeless "Big Box" sound.

The supreme irony of this is that there is almost no correlation between the THD figure and listener satisfaction. Unless something measures really badly can it be said that it will sound really bad. If THD had everything to do with it, we would have attained audio perfection a long time ago. Solid state amps have THD figures of 0.001% or even better, yet when was the last time you ever heard a manufacturer of VT equipment brag that their amps had that great transistor sound?

There are further complications that you just don't see in other areas of electronic design. An RF amp works on a very narrow band of frequencies into a well characterized load that's almost pure resistance: either a transmission line or antenna. If there are any reactive components, you can tune them out. Whatever device distortion is produced is of no consequence since one or more tuned LC circuits and/or bandpass filters will reject any harmonic distortion. This does not apply to audio. Even though we draw nice, straight loadlines representing a resistive load impedance, speakers are anything but. A spec that says a speaker is an "8 ohm speaker" refers to the DC resistance of the voice coil winding. In operation, the device is an AC synchronous motor. The only difference is that the voice coil moves back and forth instead of
going around. Any time you have coils moving in a magnetic field, you have a voltage induced in the coil. That induced "back EMF" adds to the incident voltage to produce an impedance that varies in both magnitude and phase. In practice, your loadline opens up into an ellipse (for one frequency only). This is bad news for your output devices, which like to see a constant load. Then there's the question of OPTs. A transformer designed to operate at a frequency of 20Hz is going to be very different from one that's designed to operate at 20KHz. And yet, OPTs are expected to do both, everything in between, all at once. By all rights, it shouldn't work at all, let alone as well as it does.

Distortion is not created equal, nor is distortion necessarily undesirable. If that were the case, then no one would ever use a tone control or graphic equalizer. Since these suppress or enhance various frequency components, what comes out is no longer the same as what went in. In other words, the output is distorted. The undesirable, but inevitable, distortions are not created equal either. You can see this with an oscilloscope. If you add 10% of second harmonic to a pure sine wave, the difference is hardly noticeable at all. Even if you know what to look for, you'll probably miss it. It's not until you get to about 20% h₂ that the disturbance becomes obvious (flattening of the peaks on one side only, all even harmonics produce asymmetric waveforms). Just 10% of the third harmonic will make for clearly distorted waveforms (nearly triangular waves). Smaller amounts of higher order harmonics make for even more severely distorted sine waves. This means that you can tolerate more lower order harmonics than high order. If the THD is several percent, but it is all h₂, you probably won't notice it at all. It is this type of distortion that low-u triodes tend to produce. On the other hand, even very small amounts of high order harmonics can be all but unlistenable. The cross over distortion produced by underbiased transistor amps -- or Class B VT amps -- is a good example of this, and it does sound quite nasty, even in small quantities. It is for this reason that Class B is relegated to RF amplification, or audio amplification where power and efficiency are more critical than fidelity.

As for what the various distortions sound like, well, that depends. Unless the distortion is extremely bad, it isn't really all that obvious. Some harmonic distortion doesn't sound any different than background static, or hissy "esses", or background crackles. Often, there really isn't any one thing one can identify, other than listening becomes fatiguing after awhile. The least harmful is h₂.
Since the second harmonic is musically correlated, being one octave higher, it tends to make for a "richer" sound. Excessive amounts tend towards a "darker" sound, as has been the complaint against the very nonlinear 12AV7. From a design point of view, this would not look like too much of a problem, since even order harmonics are nulled with balanced circuitry. As for h₃, this, too, is musically correlated, being nearly an octave-and-a-half. The effect of h₃ is to lend a sense of "detail". More h₃ lends "edge" or "brightness" to the sound. For some types of music, this can be desirable. As for higher order harmonics, these aren't musically correlated, and sound very dissonant and unpleasant. The higher order the harmonic, the less of it you can tolerate.

THD numbers really tell you very little. The spec sheet for the 6L6, for example, gives a THD figure of just 1.8%, before any NFB is ever applied. That is really quite outstanding. The lowest THD figure for the 6V6 -- another audio type -- is 3.5%. However, operating open loop without any NFB, 6L6s definitely sound worse, with lots of "pentode nastiness" that quickly wears the listener out. With some program material, the effect is as annoying as fingernails on a blackboard. O. Schade, the developer of the 6L6, recognized this problem back in 1936. O. Schade recommended that local NFB be applied to correct for the problem of higher order harmonics. Even though the 6V6 measures worse by a significant margin, it produces mainly h₃. This doesn't mean that you won't need NFB, however, running open loop, produces mainly an overly "aggressive", "edgy", or "bright" sound. The type doesn't need the added assistance of local NFB in order to sound right. Fifty or so years ago, Norman Crowhurst proposed a weighting system of distortion measurement that would account for this: undervaluing the less destructive low order harmonics, and emphasizing the higher order harmonics. Simple THD measurements don't do this. Of course, the industry wanted nothing to do with it.

There is another type of distortion that's much worse than harmonic distortion, and that's IMD. IMD arises from the same source as harmonic distortion: active device nonlinearity. IMD occurs whenever a lower frequency mixes and modulates a higher frequency. It's really the same as AM with a very small modulation index. All IMD is quite bad since its frequency components are rarely, if ever, musically correlated. This is the reason to stay away from nonlinear devices and nonlinear operating points. IMD is also the reason why poorly filtered DC sounds worse than AC for filament heating. With AC, you have just the one frequency to deal with. DC, unless well filtered, adds quite a few harmonics of the line frequency. Even if the level doesn't seem all that high, there is sufficient to mix with signal frequencies via intermodulation to degrade sonic performance.

The simplistic answer is to add as much NFB as you can manage while maintaining a reasonable level of stability. Unfortunately, it's not that easy. Over fifty years ago, Norman Crowhurst looked into the problem, and discovered that the main effect of NFB is the reduction of h₂ and h₃. It doesn't do much, if anything, for removing low levels of higher order harmonics. NFB can make the situation even worse by filling the noise floor with lots of low level, high order harmonics. This can, and does, ruin any pretense to sound stage. In much the same way as NFB, the unavoidable quantizing noise that accompanies analog to digital conversion also adds lots of uncorrelated harmonic noise to the noise floor. It has been proposed that this quantizing noise is the reason that some find the CD sound to be inferior to vinyl and tape. This is also a problem with certain audio file compressions, such as mp3. This is a "lossy" compression, in that some of the bits are thrown away. The first consideration in which bits those should be were the ones encoding near noise floor audio. These mp3 audio files just don't sound quite right because of that missing detail (And to think some people actually expect you to pay for mp3's!) What's going on at the nearly subaudible level seems to be a good deal more important than once suspected.

The main lesson is to design for the best open loop performance, and add just as much NFB as needed to improve the overall sonics. Of course, it is a trade-off: sound stage for better bass and less of the excessive harshness that often comes with open loop operation.

Common Amplifier Myths

 

You Should Always Use (Some Brand) of Tubes

There is seldom any reason to pay premium prices for "special" VTs. I'm certain that Western Electric NOS 300Bs are quite decent tubes. I'm also certain that they are not worth the $High Four Figure prices these command. You can see certain types going for some ridiculous prices these days. That's nothing but pure audiophoolery. Put an over priced tube in a mediocre amp, and you still have a mediocre amp. "NOS" is also not necessarily the magic word either. Back in "the day" good, bad, and indifferent tubes rolled off the assembly lines by the tens of thousands every day. Some war surplus is positively hideous, which is why it's surplus -- military rejects. Truth is: some NOS tubes are horrible, and there are current production tubes that are every bit as good as those made in "the good ol' days".

This is not to say that the seller is necessarily trying to put one over (although that does happen all too often). The advertised tubes may work especially well with whatever equipment he's using. It doesn't mean equivalent performance in random gear with random speakers.

"Tube rolling" does have a basis in fact. Unlike transistors, VTs are low gain devices. Being low gain devices, circuit performance is more dependent on the characteristics of the active device. Transistors have so much gain that circuit performance is totally dependent on the circuit itself. Unless you need some special property (such as an unusually low noise figure, or that can operate at unusually high voltages) it doesn't matter what transistors you use so long as the transistor in question can process the signal frequencies. Not so with tubes, especially in designs that don't incorporate any NFB. You may get a different sound from different tubes of different construction, name brands, model years. Whether it's worth it or not is something else.

It is not necessary to pay audiphool silly prices for good sound.

Point-to-point is Good, Circuit Boards are Bad

This is definitely true for RF circuits. For P-2-P wiring, the insulating medium is air. The resulting stray capacitance is much lower, and performance definitely enhanced. This is a good reason to opt for "dead bug" or P-2-P construction of RF circuits. At audio frequencies, there isn't enough of a difference in stray capacitance to make a difference. P-2-P has the advantage of allowing the free use of "parts on hand". Circuit boards have the advantage of easily replicated construction, and much better quality control for mass production. To be sure, a badly laid out circuit board will ruin the sonic performance of an otherwise excellent design. So too will a poorly laid out P-2-P job.

If You Use Circuit Boards, These Should be Teflon

Again, a good idea for UHF circuits, but the savings in stray capacitance will make no difference at audio frequencies.

You Should Always Use Teflon Insulated Silver Wire

Another excellent idea for UHF construction. The tiny savings in resistance and stray capacitance will make no difference at audio frequencies. The added expense isn't worth it for frequencies as low as the audio band.

You Should Always Use Carbon Composition Resistors

Another idea held over from RF practice. C-comp resistors have the advantage of having a very low inductance. C-comps hold their DC resistance well into the VHF band. Otherwise, C-comps are very noisy. Save them for your RF projects. Metal film resistors are a good deal quieter, and better suited for sensitive applications, such as low level stages. Those few nanohenries that are a big deal at 400MHz are of no consequence at 400Hz.

Negative Feedback is Bad

This bit of nonsense does have a basis in fact. In all too many cases, NFB is used to cover up a poor open loop design. Pour on enough NFB and you can sweep your mistakes under the carpet, along with a good deal of the vitality of the music. This, however, is not an indictment of NFB. Used properly -- to improve an already good open loop design -- NFB can make a good amp even better. This is the key: the open loop design must be a good one. If you find that you need excessive amounts of NFB, then it's better to go back to the drawing board and correct your open loop design. All too frequently, this is not the case.

Furthermore, there is no escaping NFB. Every device, vacuum tubes or transistors of every sort, have inherent NFB. This is the cathode/source/emitter resistance: r~= ¹/g_m. This resistance will cause degeneration just as surely as if it were an unbypassed resistor soldered into the circuit. Just because you don't see it doesn't mean it's not there. Furthermore, triodes have an additional feedback mechanism. Plate current is dependent on the V_PK. When a signal tries to pull the control grid positive, plate current rises, but the V_PK drops and tries to pull the plate current down. Since the plate and grid are pulling in opposite directions, this is negative feedback by definition. Yet triodes are preferred to every active device that doesn't have this property (pentodes and transistors).

If anyone claims to have a "no feedback" design, they're either ignorant or trying to put one over -- or what's worse: both.

Solid State is the Spawn of the Devil

Another partial truth here. The first solid state amps were truly awful: noisy, unreliable, and with hideous sonics. There has never been a technology with no down side, and there never will be. There are some things that SS does very well, and some things it does very poorly. The key is balancing the strengths with the weaknesses. The only reason to avoid all solid state is if you are replicating 1930s-era equipment for the nostalgia appeal. Otherwise, do you also say "No" to modern line, interstage, and output transformers and chokes that are made from exotic materials that didn't exist in the 1930s? All these devices can perform much better these days, and make those old circuits work much better now than they ever could then.

With zero impedance and perfect regulation, plate-circuit distortion does not exceed 2.0%... The driver stage should be capable of supplying the grids of the Class AB₂ stage with the specified peak grid voltage at low distortion. The effective resistance per grid should not exceed 500 ohms...

-- 1624 Spec Sheet

Back in "the day", this was asking the nearly impossible. Getting Z_o below 500 ohms is very difficult even for a cathode follower, unless it uses some of the big power triodes. These days, any power MOSFET source follower can come very close to the ideal of zero drive impedance, certainly below 10 ohms. Use solid state where it gets the job done right. The fact that it was a considerable challenge is testified to by the lack of Hi-Fi designs left over from "the day" that operate Class AB₂. Most of the designs that did are for AM plate modulators and PA systems where the premiums were lotsawatts and efficiency, not sonic performance. These days, there is no reason not to take advantage of going Class AB₂.

Another bit of mythology is the supposed noise that solid state diodes cause. Except for audiophile circles, I have never heard such a claim. Furthermore, I haven't seen it on the o'scope screen and I have never heard it from any receiver operated right next to any audio amp that used a solid state power supply. Gas and mercury vapor diodes, however, are notorious for producing all sorts of hash and incidental RF. Otherwise, solid state diodes offer much lower forward drops, don't get anywhere near as hot, have significantly higher I_surge ratings, and require neither heater power nor another hole in the chassis. They better lend themselves to voltage multiplier circuits, and bridged circuits. If you are that worried about high frequency noise from solid state diodes, the "fix" is a two or three pole LPF made from air coils and mica capacitors. Won't do any harm, but it sure doesn't look necessary either.

If there is any down side to solid state, it's that the high voltage comes up within a couple of seconds or so -- long before cathodes have a chance to warm up. This can overvolt direct coupled stages, possibly exceeding V_HK ratings. Since silicon diodes can source much higher currents, there is the temptation to use them with huge reservoir capacitors. The increased current demand can stress vintage power transformers that weren't designed to handle currents in excess of those expected from vacuum diodes.

A solid state power supply may or may not lead to a different sound. If there is a difference, it's due to the better voltage regulation of the solid state power supply. For some applications, good regulation isn't desirable. This would apply to over driven guitar amps where the voltage drop at max current demand lends to compression, and "sustain" as the voltage comes back up. The same effect can color the sound of a Hi-Fi system. If you're used to that sound, or actually prefer it, then, of course, solid state will be considered inferior.

It is questionable whether it makes a difference in a PP amp since the AC sums to zero at the primary center tap, making it a virtual AC ground. It might make more of a difference in a SE design, since the power supply is in the signal path. Even more reason to use solid state, as the AC impedance is considerably lower.

Tube Watts are Bigger than Transistor Watts

A watt is a watt: one newton * meter per second. Transistors clip hard and fast. This is what makes them superior digital devices, but horrible analog devices. Add NFB, and clip behavior becomes even worse. Tubes clip much more gently, first simply rounding off the peaks, as opposed to just flattening them. This gradual rounding generates mainly low order harmonics that roll off quickly with increasing frequency. The sharp discontinuities of peak flattening generates high order harmonics that roll off gradually with increasing frequency. Since low order harmonics are much less detrimental, one doesn't hear the occasional clip on fast transients. A solid state amp needs to stay out of clipping. Therefore, a tube amp will sound louder than an equivalent solid state amp, since you don't hear the occasional clip, you can turn up the volume higher before obvious distortion is heard.

Tube Amps Have Poorer Bass

This may have been true at one point. One of the advantages to NFB is that it reduces the output impedance. A low output impedance is needed to damp the woofers, otherwise they will tend to produce their own resonance note, as opposed to the notes the musician actually played. Back in "the day", the main problem here was inadequate open loop bandwidth. You can have a billiard table flat frequency response from 30Hz to 20KHz, however, if it's NFB that's causing it, the bass will not be as good as you'd expect. If the open loop response prematurely rolls off the low end, the NFB decreases, thus driving up the closed loop gain. Less NFB is less effective NFB, and won't dampen the woofer(s) properly. This problem was due to interstage coupling capacitors that were too small. At one time, your choice for capacitors was limited to either PiO, wax paper, or ceramic. PiO and wax tend to be leaky (not the oil or wax, but of current) and deteriorated with age. Ceramic wasn't leaky, but also didn't come in the sizes and voltage ratings you'd like to see. Consequently, there wasn't enough open loop gain at the low end.

These days, with new polymers, it is possible to get capacitors in the sizes and voltage ratings you need for good open loop performance at the low end. Vacuum tube amps can have good bass, with adequate speaker damping. (About the only exceptions are sub-woofers with ceramic or carbon fiber cones and heavy voice coils. These sub-woofers are difficult for even a solid state amp to keep under control.)

"Magic" Resistors

Yet another out growth of a real phenomenon. During the early 1970s, it was discovered that capacitors really did have an effect on sonics. Up till this point, no one really paid any attention to passive components. They knew about such things as ESR, DF, Loss Tangents -- however it was only VHF designers who needed to worry about this. Capacitors introduce distortions, due to D/E field nonlinearities, piezoelectric effects, and even self-excited resonances. So far as dielectrics go, the worst offenders tend to be metal oxides (ceramics, electrolytics) and other high permittivity materials. The more benign being the polypropylenes and teflons. Unfortunately, these are also materials with low permittivities, and so do not lend themselves to packing a lot of capacitance in small packages.

While dielectrics have a proved effect on sonic performance, this has also occasioned a lot of audiophoolery. From legitimate claims about the effects of various dielectrics with data to back the claims up, to utter nonsense regarding the "sound" of resistors. There are lots of claims with zero data about the superior attributes (available at superior $Prices) of these things. Don't be conned into paying dollars for some part that would normally cost pennies.

If you see extravagant claims with nothing more than vague generalities backing them up, better get a tighter grip on your wallet.

Component "Burn-In"

You see these claims about a variety of components: they require X number of hours of "burn-in" before they start to sound good. This is complete nonsense. Capacitors polarize within seconds of the voltage application. I've seen claims of 100 -- 500 hours for Teflon capacitors, and with AuriCaps. The first time I replaced an electrolytic coupling capacitor with an AuriCap, I noticed the improvement at once. Of course, electrolytics make lousy signal capacitors, and AuriCaps are pretty good. There was no "burn-in" required at all.

Resistors sometimes burn out, but they never "burn-in". As for cables, well, you figure that one out. The only components that require a certain amount of "burn-in" are vacuum tubes. It may take minutes to hours for the bias to settle down when new tubes are put in service. This is completely normal, but does not apply to any sort of passive component, nor does it apply to solid state devices at all. Such claims are largely disengenuous, in that the time it takes for the new component to "burn-in" is equal to the time it takes for you to get used to the new (and most likely) inferior sound of whatever over priced, whiz-bang, gizmo you just likely spent way $Too Much to acquire.

Next time you see this particular claim, ask yourself one question: why didn't they burn the thing in at the factory before sending it out the door?

SRPP Demystified

SRPP

There is considerable misunderstanding and mystification regarding this particular circuit, including even what it's called. "SRPP" seems to mean: Shunt Regulated Push Pull, although you see different words assigned to the acronym.

As for what, exactly, this thingy is, it should be immediately recognizable to anyone with a solid state design background: it's an active pull-up/active pull-down circuit. (It greatly resembles the output stage of the TTL family.) The main difference is that it is quasi-complimentary by necessity since there is no such thing as a "P-Channel" VT. This means it is, indeed, push-pull by definition. The upper triode acts to source current to the load, and the lower triode sinks current from the load. As with any push-pull topology, it reduces distortion by nulling even order harmonics.

The big point of departure from solid state is that the SRPP is balanced for three load conditions only: a dead short (makes both triodes into grounded cathode stages -- not very useful) an open circuit (or at least a very high load impedance -- if there is just one path for the current, equal currents must flow through both triodes) and the one impedance for which it was designed. For any other load conditions, the SRPP goes out of balance, and distortion rises, more or less, rapidly. This is OK, considering the purpose for which the SRPP was originated: a line driver. As a line driver, it operates into the characteristic impedance of the T-line.

As an audio circuit, this leaves a lot to be desired. If it works into Class A*₁ grids, then it's OK since that's nearly an open circuit. The trouble starts when the driven control grids are driven positive, and draw current. Under grid current conditions, the resulting impedance is neither constant nor linear. This is not the type of load an SRPP wants to see. Including such a driver will lead to poor clipping behaviour. Unfortunately, it is grid current conditions where you'd like to include active pull-up to not only supply that grid current, but supply it from a Lo-Z source to minimize distortion.

You also see SRPPs used as audio finals. This, too, is not the place for this since any speaker represents anything but a constant load. Speaker impedance varies not only in magnitude, but also phase angle. This will play hell with an SRPP, as it will be operating off its optimum load impedance almost all of the time. That will generate lots of avoidable distortion which will require that much more NFB to correct. This is not what you want in good open loop design.

As for how the SRPP develops voltage gain, consider the R_P. The far end of that plate load resistor has a cathode follower sitting on top of it. This means that the voltage across that resistor is much less than it would be if connected directly to the DC rail. The effective AC resistance is much higher than its DC resistance. This gives the lower triode more voltage gain that it would otherwise have. The next question becomes: can we make that resistor even larger to increase voltage gain? If you break the DC coupling, the answer is "yes". This gives us the variation called a "Mu stage", so called since its voltage gain can approach the amplification factor (μ) of the lower triode.

Mu Stage

The AC coupling allows for a larger R_P since it is no longer doing double duty as a cathode bias resistor. The mu stage no longer has any pretension for being a balanced topology. It is designed for large voltage gains. Is this a useful topology? It was back in "the day", however, it is obsolete and should not be used. Today, we have solid state devices which can operate as excellent CCSs. You will do much better loading a triode with a constant current, as this gives a horizontal loadline that both maximizes output swing, voltage gain, and minimizes harmonic distortion. Back in "the day" we didn't have the ICs, BJTs or MOSFETs that could have made for decent CCSs. Before then, your only other recourse was to use a pentode as a CCS. That would give you something quite mu stage-ish anyway.

If you need lots of gain, then use a solid state CCS. These old fashioned circuits serve no purpose these days, other than nostalgia appeal, or audiophool trendiness on the part of those who like "exotic" circuits for the exoticness.

Cathode Follower

 

AC Coupled CF

In this schemo, R_K is the normal cathode bias resistor. R_L represents the tail load in parallel with the load impedance. R_G is the control grid DC return. The CF gives excellent high frequency performance since Miller Effect is absent, and the C_GK sees very little current since the grid and cathode are always at nearly the same potential. This makes the C_GK effectively smaller than its static value. The main component of input capacitance will be the reverse transfer capacitance: C_GP. With small signal triodes, it is easy to present a Hi-Z, Lo-C load to the driving stage. This isn't just helpful at RF.

This is another circuit which has lately come under unjustified criticism within certain audiophile circles. Much of this is unjustified on the basis that the CF is a negative feedback circuit. This view that all NFB is all bad does have a basis in fact. It has been all too common to use NFB to cover up for poor open loop designs. If your open loop design is poor, just pour on the NFB to force "the numbers" to look good. Sure, you can sweep your mistakes under the carpet that way, but you will also sweep away much of the vitality of the music. However, this is a misuse and abuse of NFB. The blame properly belongs to these lazy designers who can't be bothered to correct their open loop designs.

Another big part of the problem lies with the nature of the cathode follower itself. Yes, it can present a low impedance source, but only to a high impedance load. There is a big difference between source impedance and load impedance. For example, a 6C4 small signal triode could be used to implement a cathode follower. If you implemented it thusly:

V_PP= 330V_DC
V_PKQ= 140V_DC
I_PKQ= 4.0mA
R_tail= 47K

r_p= 10.5K
g_m= 1.4mA/V

You could easily calculate a Z_o= ~647Ω Given that Z_o, you might think you could drive a set of 600Ω headphones with this cathode follower. However, you would be quite wrong about that. Your undistorted power output will be just under 10mW as you'll only be able to swing just 3.4V_p into that load. So what happened? When you connected a 600Ω load across the 47K tail, you killed most of your gain by making a nearly vertical loadline. Less open loop gain means less effective NFB. If it can not drive a 600Ω load connected to the plate it can not drive that load any better if you connect it to the cathode.

So what is it good for? This cathode follower would be ideal for isolating a grounded cathode gain stage from, let's say, a tone stack. The GC amp will have a very large Z_o. You could incorporate that source impedance into the resistances of the tone stack, to be sure. However, tube characteristics vary considerably with the manufacturer, models within brands, with age. The output impedance probably won't stay put, and if it varies, the poles 'n' zeros of your tone stack will change with it. Isolating the tone stack with a cathode follower representing an insignificant portion of the tone stack resistances will prevent this. It will also allow for more resaonable values of resistance and capacitance in the implementation.

Cathode followers can also be used as active pull-up circuits to drive the control grids of audio finals. Even if you stay with Class A*₁ operation, the input capacitances (C_GK + C_Miller + C_stray) are still going to require current to charge. If the current sourcing capability isn't there, then you will run into slew limiting at the higher audio frequencies. That sounds nasty. The CF can supply enough current to prevent this from happening, especially if you follow the "Rule of Five" from solid state practice: make the Q-Point current of your CF at least five times greater than your anticipated peak current.

Don't ever forget: the vacuum tube itself neither knows nor cares whether the load is connected between the plate and the positive rail, or if it's connected between the cathode and DC ground. It's always the same loadline, the same load resistance. You don't gain anything by trying to force the device into being something it will never be: a high current, low voltage device. A CF is not a magical power gain stage.

Attempting to use it otherwise will lead to degraded sonic performance. If you have a bad-sounding CF, blame the designer, not the topology.

As with any other audio subsystem, if the CF is designed properly, and used within its limitations, it is the most sonically transparent audio subsystem. If designed badly, it will sound bad. It's as simple as that.

Demystifying the Cascode

Basic Cascode

This is what the schemo of the cascode looks like. R_K serves to establish the Q-Point bias for the lower triode, and R_G is its DC grid return. R_P is the passive plate load. The voltage divider connected to the grid of the upper triode establishes its Q-Point bias.

So why would you want to do this? What you have here is a cascade of a grounded cathode stage driving a grounded grid stage. The GC topology has the advantage of a Hi-Z input. However, its high frequency performance is impacted by a high C_Miller that only grows worse with increasing voltage gain.

The GG topology avoids C_Miller for excellent high frequency performance, but it suffers from a Lo-Z input. It's not very often that a Lo-Z input is desirable. However, you can combine the two in a manner that work together. The Lo-Z of the GG stage loads down the plate of the GC stage, reducing its gain greatly. The lion's share of voltage gain comes from the GG second stage. Reducing the gain reduces C_Miller to manageable levels while preserving the Hi-Z input. This gives the cascode a characteristic more like that of a small signal tetrode, with its reduced C_Miller, high voltage gain, less the screen grid "kinks", and partition noise. Unlike a tetrode, the V_GK of the upper triode remains negative, and so you also don't get the partition noise that tetrode screen grids produce. That's where the name comes from: a contraction of "cascade" and "tetrode".

It is for this reason that the cascode is frequently cited as a VHF small signal amplifier. As for what this means for audio amplification, the reduced C_Miller is helpful since a volume pot with a high resistance won't produce the roll-off that bothers designs that place the high gain triode stage up front. The cascode, having higher gain than a single triode gain stage also looked quite useful. As an LTP phase splitter, the cascode could give enough gain to eliminate a second gain stage.

As for sonic performance, there is very little information concerning this. Most of the information I could find related to the design of guitar amps where voltage gain was the emphasis, as distortion is much less of a consideration in such designs. Cascoded LTPs are used quite frequently in solid state designs. However, I could not come up with any examples of hollow state designs that used this topology. Was that just because it was "weird", or required another dual triode, or had this been tried and found to be sonically inferior?

This was another case of try it to find out. As for VT selections, the 6SN7 didn't provide enough gain, and the 12AT7 suffered a fast g_m roll-off with decreasing plate current. Digging into the RCA Receiving Tube Manual (RC-30), I came up with the 6BQ7A -- a dual triode designed specifically for cascoding and operation up to 300MHz. Its μ= 38 is considerably higher than that of the 6SN7, but is still reasonable for this particular design. The main problem with the 6BQ7 is that it isn't an audio tube, has no audio use specified in the spec sheet, and has a very peculiar plate characteristic (undocumented variable-μ feature?). Finding a good audio loadline is not so easy, and this type likes to see a V_PK that's higher than usual. Still, it looked doable, and since this is a balanced topology, it should greatly reduce harmonic distortion if most of that is h₂.

Cascoded LTP Design

There are a couple of bug-a-boos with this topology. The output voltage swing is rather small for the DC rail voltage, the PSRR is less than that of most triode-only gain stages, and the output impedance is very high. This last feature is desirable in an RF amp since that means less loading of LC tuners, and higher loaded Q's. For an audio amp that has to operate over a rather large range of frequencies, it's not such a good thing. A Hi-Z output will interact poorly with the input capacitance of a subsequent stage or other load. To prevent premature roll-off, it is necessary to operate into a Hi-Z, Lo-C load.

The design also uses an active (CCS) tail load. The CCS presents a very high impedance to the junction of the two cathodes. This greatly improves both the phase-to-phase AC balance, and also equalizes the harmonic distortion between phases. It is this distortion imbalance that gives a great many phase splitters inferior sonics. The CCS is a cascode of BJTs. This gives better performance than would the more traditional small signal pentode. The BJT, having much greater gain, gives a higher tail impedance, and a more nearly constant current, with a lower negative rail voltage. This is one area where solid state really is better.

O'scoping the output of the cascode directly resulted in poor square waves with severely tilted tops, indicating a loss of high frequencies. The -3db_v point was barely 20KHz. That's positively horrible! However these were artifacts of the C_i of the o'scope probe, cable, and vertical deflection amp. So this meant it required a "friendlier" load. For that, a cathode follower works nicely. A CF stage has the high input impedance, and the low input capacitance since there is no Miller Effect. Since the voltage at the cathode "follows" the grid, there is very little current through C_GK, so this capacitance all but disappears. When operated into a cathode follower, the true picture emerged: nice flat square waves to 10KHz, a -3db_v point of 117KHz, more than enough for an audio amp.

So how did it sound? In a word: excellent. There were no pentode-like artifacts at all, and the sonics were identical to what a good triode LTP would produce. It beats all the paraphase splitters in their various iterations. This definitely should see much wider use than it does.

Welcome to Dolphin Hollow State Labs

 

For audio amplification, the first active device remains the best active device: the triode vacuum tube. The triode is unique among the active devices: vacuum tube pentodes and transistors of all sorts. With these devices, the plate/collector/drain current is largely independent of the voltage across the device. This property makes them excellent approximations of an ideal current source.

With triodes, however, this is not the case. The plate current can vary with the plate-to-cathode voltage. This is the meaning of "amplification factor" --

μ= ΔV_pk / ΔV_gk (at Ip= constant)

For most triodes, the amplification factor will usually be between 10 to 100. For power triodes, this can be smaller (audio power finals, vertical deflection power amps, series pass regulators) or higher ("zero bias" RF finals). Amplification factor is largely meaningless for other active devices since it's so difficult to measure directly.

The triode, like every other active device, has an inherent degeneration when operated as a grounded cathode amplifier. This being the cathode resistance: r_k= ~1 / g_m. This resistance acts in precisely the same manner as if it were an unbypassed resistor soldered into the circuit. Just because you don't see it doesn't mean it's not there.

There is another feedback mechanism at work as well. When the signal pulls the V_gk less negative, the plate current increases, and with it, the voltage drop across the plate load. This results in a decreasing V_pk. As this voltage decreases, it tries to pull the plate current lower. Since V_gk and V_pk are pulling in the opposite direction, this is negative feedback by definition.

It is this additional source of NFB that serves to correct for harmonic distortion to a greater extent than you will see with other active devices. What harmonic distortion remains is mainly the second harmonic, h₂. Sonically, h₂ is the least detrimental. It is this h₂ that lends to the so-called "tube sound", described as "warm", "rich", "full", etc. More h₂ is described as "dark". No distortion would be the best, but the perfect amplifier, like the perfect lens for telescopes, has yet to be invented. The question becomes how to minimize those defects which you will never completely eliminate.

Of course, not all triodes are equally good sounding. As far as the effectiveness of plate current control by plate voltage, this is measured by the term: plate resistance: r_p. A good many high-μ triodes attain that large amplification factor by driving up the plate resistance. For the 12AX7, the spec sheet gives: r_p= 80K (nominal). This is comparable with the r_p's of small signal pentodes. Is it any wonder why the 'AX7 tends to sound like a small signal pentode?

Even if the r_p isn't excessive, some triodes just won't perform all that well for audio. This being caused by excessive variation in g_m with plate current variations. For all active devices, amplification tends to increase with increasing current. If the positive going half cycle of a sine wave receives more amplification than the negative half cycle, then the two half cycles hit different peak voltages. That's not a sine wave any more, and something has been added. Since this is asymmetrical distortion, it is even order, and mainly h₂. Types such as the 12AV7 tend to produce much more of this distortion than you'd like to see. (Although it just might be useful for audio effects.)

Nothing sounds better. It is for this reason that this "obsolete" device is still around over a century after its invention. This has been known at least since the early-1950s when the quest for a solid state device that could match the distortion performance of the VT triode began. So far, no one has found such a device.