Categories
hardware internet tech

That Elusive Valve Amp Sound, For Not A Lot! (There Has To Be A Catch)

It was with considerable interest last month that I set out to track down where in the world there are still factories making tubes. My research found them in Slovakia, Russia, and China, and it’s fairly certain I didn’t find all the manufacturers by any means. There appeared to be a whole class of mundane tubes still in production that weren’t to be found on their glossy websites. A glance at any outlet through which Chinese modules can be bought will find this type of tube in small audio amplifier projects, and some of them can be astoundingly cheap. When faced with cheap electronics of course I’m tempted to buy some, so I parted with about £10 ($12.50) and bought myself a kit for a two-tube device described as a stereo preamplifier and headphone amplifier.

An Unusual Tube Choice For Audio

What I received for my tenner was a press-seal bag with a PCB and a pile of components, and not much else. No instructions, which would have been worrisome were the board not clearly marked with the value of each component. The circuit was on the vendor’s website and is so commonly used for these sort of kits that it can be found all over the web — a very conventional twin common-cathode amplifier using a pair of 6J1 miniature pentodes, and powered through a +25 V and -25 V supply derived from a 12 VAC input via a voltage multiplier and regulator circuit. It has a volume potentiometer, two sets of phono sockets for input and output, and the slightly naff addition of a blue LED beneath each tube socket to impart a blue glow. I think I’ll pass on that component.

The 6J1 seems to be ubiquitous throughout the Chinese kits, which is surprising when you understand that it’s not an audio tube at all. Instead it’s a small-signal VHF amplifier, a rough equivalent of the European EF95, and would be much more at home in an FM radio receiver or turret TV tuner from the 1950s. I can only assume that somewhere in China there’s a tube factory tooled up for radio tube production that is targeting this market, because another tube you will see in audio power amplifier kits is the FU32 or QQV03-20 in European parlance, a large power beam tetrode that might have been found in a 1950s military radio transmitter. Still just as if you were to use an RF transistor in an audio circuit it would give good account of itself, so it is with an RF tube. There is no reason a 6J1 won’t do an acceptable job in a circuit such as this one.

There are no surprises among the parts.
There are no surprises among the parts.

Other than in its slightly unusual power supply, there’s nothing at all remarkable about the preamplifier circuit. The 6J2 is wired as a triode, and because a common-cathode circuit is designed to drive a high impedance, it’s safe to assume this won’t be a very good headphone amplifier. It’s a simple preamp circuit that has graced the small-signal end of countless tube amplifiers, including my youthful folly.

Building the kit was as straightforward as any other through-hole design, and made for an enjoyable half-hour or so. There are no special things to note, I simply worked my way through it, first resistors, then diodes (but not those LEDs), transistors, capacitors and finally larger parts such as the potentiometer and sockets. A visual look over showed nothing of concern, so I plugged in the tubes and applied 12 VAC.

It’s amusing, this must be the first time I’ve ever used a new tube socket rather than one scavenged from old equipment, so I was unprepared for how stiff it was to plug in compared to one with years of heat cycles to soften its metal. On turning the kit on I was rewarded with a soft glow from the tube heaters, and measuring the voltages I found that it was generating about plus and minus 30 volts. A quick check applying some audio to it showed that it was indeed amplifying audio and it didn’t sound bad, but I needed a little more than that. It’s time to characterize the amplifier, is it any good?

Way Too Much Instrumentation For A Ten Quid Amplifier

An Audio Precision APx525 audio analyser.
An Audio Precision APx525 audio analyzer. Bradp723 (CC-BY-SA 3.0)

How do you characterize an amplifier, or any other piece of audio gear, for that matter? There are four metrics worth knowing: the gain, frequency response, phase response, and harmonic distortion. Frequency response relates to the range of frequencies it accepts, phase response relates to the phase shift between output and input at any given frequency, and harmonic distortion is usually expressed as a percentage of the output spectrum that is due to the non-linearities of the amplifier rather than having been present at the input. There are specialized instruments referred to as audio analyzers that will automate all these measurements by injecting a very high-purity sine wave into the device under test and measuring its output, but they are extremely costly and beyond the budget of a iandroid.eu scribe.

For this amplifier the frequency and phase responses are not likely to be concerning, so I would have to find a way to measure its distortion. Fortunately I was able to borrow a Keithley 2020 precision multimeter from a hackerspace friend, this very high quality instrument unexpectedly has a THD (Total Harmonic Distortion) function and its associated signal generator built-in. I suspect it may be aimed at the comms industry rather than the audio business, but it gives me what I need and I am very grateful to my friend. Set up alongside my trusty Rigol 1054z, I was ready to characterize the board.

Measuring THD with the Keithley 2020
Measuring THD with the Keithley 2020

I set up the generator at the ubiquitous audio testing frequency of 1 kHz, with a 100 mV pk-pk sine wave as input to simulate a lower-level audio source. The output at maximum volume was 4.85 V pk-pk, at which the THD was 1.31%. I calculate that as 37.31 dB. Adjusting the volume for unity gain, i.e. 100 mV pk-pk output, gave me a 0.03% THD reading. Giving it a  1V pk-pk input to simulate a line-level input gave me a significant level of visible clipping and an astronomical 32% THD at maximum gain.

What do these figures tell me? At low signal input levels it has the potential for a low THD, but even then it’s not into the three-digits-past-zero zone you’d expect to see in a high-end audio product. At higher levels it starts to degrade significantly, but given that it’s a preamplifier rather than a line-level amplifier that should be hardly surprising. It’s possible that some negative feedback would tame it, however I’d then be worried that its phase response might suffer.

The question is, does it matter that it’s not a super-high-end preamplifier? For ten quid spent, not really. If you keep the volume low enough and hook it up to your hi-fi it’ll sound decent enough, indeed that’s just what I did. And it sounds, well, like my hi-fi system. No special “warm valve sound”, but it did give me the momentary cachet of a pair of glass tubes on top of the stack.

Have I Bought A Masterpiece, Or A Toy?

There is one final thought with respect to this amplifier, and it returns to what I said earlier about its being described as a headphone amplifier by the vendor. A single-ended tube amplifier has a very high impedance output, which is to say it’s good at delivering voltage, but not current. Something with an impedance in the tens of kilo-ohms is fine for it to drive, but not a typical pair of headphones with an impedance in the tens of ohms. I didn’t even try, because I know that it wouldn’t do a very good job. All is not lost however, because it’s possible to make a simple headphone amp with this circuit if you’re prepared to add to it a little. Back in the day it might have had a step-down transformer on its output, but those are hard to find in 2020, so an alternative might be to use a MOSFET source follower as a buffer. It’s one of those projects that I might find myself returning to with this amplifier.

There’s a lot of talk among a certain section of the audio enthusiast community about something special surrounding a tube amplifier. Pseudo-technical explanations involving distortion in even harmonics are trotted out, and a lot of fancy tube hi-fi kit is fawned over. It’s true, that a good tube amplifier can be a very good amplifier indeed, but after having been right through this subject over decades I have my doubts over whether the mere presence of a tube confers anything extra-special.

I think it has its roots in the first generations of transistorized amplifiers in the 1960s and 1970s, when germanium transistors and single-ended power supplies requiring hefty electrolytic capacitors on the output delivered some models with not-very-good performance, but we’ve moved on since then. Superlative quality transistor amplifiers were a done deal decades ago, and while their tube siblings stand alongside them in quality, I think it’s dubious to claim too much else beyond saying that in audio as everywhere else: you get what you pay for. Buy one of these little kits and have fun playing with a tube circuit for pocket money prices, but don’t expect too much from it.