300b lovers

Quick recap: actually, vacuum tubes are far from saturation when set to normal bias points. Look at a 300B, or any other power tube. Normal quiescent bias is set between 60 and 85 mA, if Class A operation is desired. If Class AB is desired, 35 to 40 mA is more typical. With 400 volts from cathode (or filament), that's a steady-state plate dissipation between 14 and 34 watts, well within the 40-watt rating.

But that's nowhere close to the peak current emission of the cathode. I've measured 250 mA from a generic 300B, and the exotic European 300B's can slam out nearly 500 mA (transient). The only time I've ever seen a 300B current-limit around 80 mA were some particularly weak Chinese tubes from the mid-Eighties ... they sounded and measured pretty bad, and were near-defective. Other vacuum tubes are similar; the recommended quiescent currents are set by plate dissipation limits, not cathode emission maximums (which are left unspecified). Transistors will melt the internal copper links, but damaging the cathode in a vacuum tube is really hard to do unless the tube is operated with no B+ present.

It's transistors that have Safe Operating Area (SOA) curves that are log-log in both current and voltage (with an additional time dimension), not tubes. The current saturation mechanisms are totally different and have nothing in common.

Unlike transistors, vacuum tubes have very large areas of peak current emission that are left untapped by most circuits. Of course, plate heating goes up when these areas are explored, but unlike transistors, tubes do not fail in milliseconds (this is shown in the SOA curves of transistors, and must be respected). It takes sustained abuse, over many seconds, before mechanical deformation dooms the plate.

I think Ralph will agree that Class AB operation is not "false". In Class AB, one device cuts off (goes to infinite impedance and conducts no current) while the opposing device goes to a large multiple of the quiescent current. In conventional Class AB transistor amps, the idling current is a tiny fraction of the peak current, and in Class AB tube amps, it's still a small fraction.

Let's look at what happens in pure differential circuit, either tube or transistor, with a current source setting the quiescent current. This circuit must always operate in Class A. Unless something fails, the current source will always deliver the programmed current ... that is a hard limit that cannot be exceeded under any condition.

The late Allen Wright actually built a PP 300B power amplifier that had a current source under the pair of VV52B's (massive Czech power tubes). He stayed at my house during one of the VSAC shows, and we compared his amp to my early version of the Karna (which has bypassed cathodes and can operate in Class A, Class A2, or Class AB, or even Class AB2, depending on current demand). Allen's output stage was true differential, and true Class A, with a powerful solid-state current source running around 160 mA (if memory serves ... this was in 2003 or so).

The two amps sounded completely different. That's when Allen, and I, realized that differential, and balanced, are not in fact the same. This is a common illusion, a hangover from the Fifties. The question is what happens when one device cuts off.

When this happens in a current-sourced differential circuit, the "ON" device can never pass more than the total current programmed in the current source (by definition). That's a hard limit. It is a brick wall. The circuit, as a whole, will always pass whatever the current source is programmed to do ... no more, no less, always the same. This is why this circuit is seen in the Mullard topology as a low-power, medium-voltage phase splitter. Allen, as a big fan of differential circuits in Tek scopes, took it all the way and used it in a power stage.

This is quite different than a Class AB, or conventional Class A, power stage. Whether cathode or fixed-bias, current flow through the output pair is dynamic. IF (a very big if here) the output tubes were distortionless, perfectly matched, AND never voltage-clipped or driven into Class AB, yes, it would behave the same as a current-sourced pure differential stage. Only then are they the same.

But we don't live in a world of Platonic ideals. Tubes are not actually the same as the tube models, they are not perfectly assembled in perfect factories by robots, loudspeakers have odd ideas when they want lots of current, and bass drivers in particular are notorious for nonlinearity and very long energy storage .., all of which affects output stages.

So a power amplifier must deal with speakers as they are, not as we want them to be. So peak current excursions can be accommodated when necessary, without the amplifier grossly departing from basic design assumptions. The loudspeaker conforms to Theile/Small equations most of the time, but both Neville Theile and Richard Small warn us that these are only small-signal approximations. They are not valid once the voice coils start to move significantly. Speakers are only linear on average, not all the time.

My goal with Class A output is to synthesize a fixed output impedance that remains constant with real-world loudspeakers, which I have been designing since 1975. I know how awful speakers are. Most power amps use 20 to 50 dB of feedback to synthesize a perfect voltage source, and they do a pretty decent job of it. With zero feedback, the best I can hope for is a fixed, moderate-value equivalent resistor, about 2 ohms or so, which a low-Q vented or closed box speaker can deal with. And an output stage that does not have a hard current limit, but soft-clips in both voltage and current, without requiring protection circuits.

The above, in a nutshell, is why tube amps behave very differently at clipping than SS amps. It is why, with the right speaker of course, that a 60 watt tube amp can sound like it has the drive of a 200 watt SS amp, and especially why these 300b monos with a mere 27 watts each can sound like a 200 watt SS amp. Those of you who heard them at the show when the system was cranked up could hear their drive capability on an open baffle speaker of approx. 88-89 dB efficiency with a stable 4 ohm that is very well behaved. You cannot clip the amps at any sane volume level, and they can deliver large amounts of instantaneous current, while maintaining their sound quality.

There are certainly speaker designs that require 200 watts of SS amp power to wake them up. These amps are not for those speakers. But any reasonably "tube amp friendly" speaker is no problem.   If you have a speaker that presents a difficult load for the amp, then certainly one of the class D amps may be a very good choice for you.  

@lynn_olson

You might want to read this article:

The Power Paradigm

Most zero feedback tube amplifiers are Power Paradigm devices.

Regarding some of your comments in your post above:

How an output section behaves was not the topic when you brought up this bit of conversation (balanced vs differential). I never said anything about an output section. FWIW its possible to build a differential circuit so a tube can saturate when the other half is in cutoff. Its all about operating points as you rightly pointed out.

FWIW the first differential amplifiers were single pentode circuits; the grid being one input and the cathode being the other. IOW all tubes behave differentially- they amplify what is different between the cathode and grid. On this account, you can see that setting the operating point is the crucial bit which may or may not allow the tube to swing from saturation to cutoff. Drive has a lot do do with it of course. My surmise is Allen simply didn’t set his operating point correctly in your anecdote.

We were building tube differential voltage amplifiers before Allen came on the scene- we were the first worldwide to offer them to the public in a audio product meant for home use. My recommendation is to spend more time working with them and see if you might arrive at a different conclusion.

The above, in a nutshell, is why tube amps behave very differently at clipping than SS amps. It is why, with the right speaker of course, that a 60 watt tube amp can sound like it has the drive of a 200 watt SS amp,

In case there’s any question about why this might seem so, its how the tube output section makes distortion at clipping. A zero feedback tube output section has a very gentle clipping character; at early onset you don’t hear the amp breaking up at all, but the distortion has skyrocketed and the higher ordered harmonics cause the amp to sound louder than it really is, despite no obvious breakup. Its an illusion.

It has led to the myth that tube power is more robust than transistor power. But in simple terms a Watt is a Watt; but how distortion interacts with our ears is a different thing altogether. A sound pressure meter will reveal the truth of the situation easily enough.

I’m always interested in boundary conditions ... what happens when the amp leaves its happy place and a surge of current or voltage is required. Does a circuit saturate and hit the wall? Does a transistor fail? Does it store charge and "stick" for a few milliseconds? How smooth is the transition in and out of the Bad Place?

I mention this because speakers are badly behaved much of the time. They store energy for tens to hundreds of milliseconds, then throw it back to the amplifier. The feedback network might, or might not, keep correcting this, but the error overshoots can be very large and can saturate an input section.

Many power amps do not accept boundary conditions gracefully. Not just the output section, but the driver as well. Driver transistors fail when SOA is exceeded by transient reactive loads (failure to accurately read a SOA graph almost bankrupted Audionics). In tube amps, drivers can’t push enough linear current into the Miller capacitance of the output tubes. The voltage-amp section of a transistor amp can’t charge the dominant-pole capacitor fast enough, resulting in slewing.

These are all boundary conditions, and they are audible not just when they reach 100% failure, but well before that, when nonlinearity just begins. The previous point about Class A operation in a differential stage still holds: what happens when more than 100% of the current programmed in a current source is exceeded?

This is a boundary condition problem. When current is exceeded, what next? What’s after that? Does anything fail? What does current clipping look like? Are there any energy storage mechanisms that result in "sticking", a well-known problem in solid-state power stages. If sticking happens, how long does it take before it gets unstuck? In milliseconds?

The approach in the Blackbird/Karna does not use current sources, nor differential stages. Each side is parallel, but in antiphase, and all phase splitting, and re-summing, is done by passive devices, which do not have slew limitations. The 1:1 interstage transformer makes sure that recovery from A2 grid-current events happen in microseconds, not milliseconds.

Overload happens in the tubes, mostly in the output section, and the overload condition is not affected by local or global feedback, so the overall boundary characteristic is that of a (very) fast-recovery limiter/compressor. There is no hard boundary between Class A, where it remains most of the time, and A2, AB, or AB2, depending on current or voltage demand.

During the development of the Karna amp, I was in a kind of perverse mood, so I was curious just how much abuse the circuit, and the tubes, could take. I set the oscillator level so the scope display was just below clipping, around 20 watts, and the 8-ohm test load was nice and warm. I increased the drive frequency beyond 20 kHz, and as transformer gain started to fall off beyond 50 kHz, I just increased the input level to keep the output at a steady, undistorted 20 watts. Because why not?

I finally lost my nerve at 500 kHz. The scope display was still an undistorted 20 watts, with no sign of triangle waves or flat-topping, but playing around with an AM-band transmitter (with 500 volts inside) was asking for trouble. I wasn’t trying to kill anything, but sooner or later some part was going to fail. (If any of you customers try this stunt, yes, we will void the warranty, so don’t do this. Ever.)

Not many transistor amps would survive full power at 500 kHz. Some would, some wouldn’t. It’s an absurd test, with no relation to audio use. But it’s interesting to know the development prototype survived it. No, I would never do this to the current production model, and don’t you guys try it, either.

Let me tell the story about how misreading a graph almost bankrupted a company. This is a true story.

Our chief engineer, a Tektronix veteran, designed a power amplifier called the Point Zero Three, or PZ3 (hey, I didn’t name it, OK?).

A 100-watt/channel transistor amp. It measured great, sounded so-so, but had a little fault ... it blew up without warning, and worse, took out all the power and driver transistors, scorching the circuit board as it went. Take out the circuit board, replace the power transistors (all of them), and put in a new one. Repeat as necessary.

There were days when more came back than were shipped out. Obviously, this couldn’t continue. Word was getting out, and a failure rate approaching 50% is unsustainable.

Our new engineer, Bob Sickler, looked more closely at the SOA curves for the driver transistors. Bob told me these curves are intentionally hard to read, and it usually escapes notice that both voltage and current axis are log scale. Not only that, a straight load-line is assumed by most engineers, but with real loads, that line opens into an ellipse. Once that ellipse touches the no-go line, and stays there for more than 10 milliseconds, boom!

The chief engineer, despite being an old Tek hand, had missed this tiny little detail. It turned out the driver transistors were undersized by a factor of three, so we had to parallel them in a 3-stack with individual emitter resistors to have a stable amplifier. The fix worked; but we had to recall every amplifier in the field and update the circuit board with the new driver stack. It wasn’t cheap, but it stopped them coming back, and they kept working.

This incident resulted in Bob Sickler, the new guy who saved the company, getting permission to design a new amp of his own, which became the Audionics CC-2, their most successful, reliable, and best-sounding product. They sold more than 1,000 with a failure rate of less than 0.3%, the lowest in the industry. The CC-2 was their most profitable product.

All this happened because one part, a driver transistor, had poorly understood behavior in a boundary condition. Ask for too much current for too long, go past the SOA boundary, and blooey! A scorched circuit board with mostly shorted transistors, all thanks to DC coupling propagating the single point of failure through the entire output section in less than a second. Engineers love DC coupling, but it can propagate failure very, very fast, in less time than it takes to jump across the room and turn it off.

That experience is why I am wary of dismissing boundary conditions. Poorly understood boundary conditions can destroy a product, destroy consumer trust, and take down a company. All from not reading a data sheet carefully enough.