Auto Bias – Keeping the Tube within spec for Classes A and AB
I used to be a radio officer in the merchant navy about twenty five years back. The MF/HF transmitters all used high powered thermionic tubes in the output circuits. In ten years I never had one catastrophic failure although I knew of guys purchasing replacement tubes over and above the legal requirement for spares use. If they only knew the price? Possibly 5 000 dollars (US) in today’s terms. Insane!
These tubes dissipated in the order of about 3kW when used in class AB. The two most commonly used transmitters were Marconi and JRC. Hagenuk was another manufacturer but they weren’t that common. Nonetheless, I never had to replace a tube once because of complete failure. There was an incident where the bias circuit failed, the anodes glowed hotter than a thermonuclear event but these tubes still worked 100% afterwards.
Tubes are bomb proof – they do fade away though
On the more modern equipment where the users were predominantly restricted licence operators there was a lot more fail safe circuitry – we were just told to get on with the job. If it failed we had to fix it. Now here is the thing – these transmitters were sometimes upwards of ten years old with still the same tubes – some blast air cooled (like the above) and some just huge bottles with a breath of air. In comes the modern fraternity of audio philes and everything must be the perfect match, perfect tubes, perfect bias and hugely expensive capacitors and resistors. I appreciate that. I also appreciate the fact that semiconductors also have to be a perfect match in the output stage. I am not forgetting the fact that many of the owners of these amplifiers (tube) are critical, vastly critical listeners but I fear for the manufacturing quality. Why the huge discrepancy between tubes and their competitive products. The prices are totally unrealistic even when looks at demands dropping (in fact they are now increasing). New 67cm CRTs used to cost about 100U$ in the eighties – this is the colour, in-line variety. The CRT has a lot more intricacies involved in the manufacturing process than a KT88 for instance. Sure, tooling is an issue, not many countries manufacture tubes these days.
Having not touched a tube amplifier for the last ten years I just cannot get my head around why we persist in blowing huge sums of money on a technology which is no longer superior to solid state for low power and within the rated dynamic range. In fact I’d say in the last twenty years semiconductor amplifiers manufactured by the high end boffins certainly sounds better than any tube amplifier I have come across for CD and DVD reproduction. Slow switch on times, hiss and heat don’t rock my boat. Certain amplifiers, such as the 2N3055 powered SKUs I believe sound uncannily like tube amplifiers. The NAD 3020 is a case in point. I had one and I really don’t think there was any comparison between a push-pull ultra-linear 15W (2 x EL84) and 2N3055 amplifier. The tube amp would run rings around it. (allow me to contradict myself here – the 2N3055 was and still is an incredible device and dollar for dollar, watt for watt is the better bet – but compared to fast switching devices of these days there is no comparison. We talk about sweet sound, depth, distortion – in a good solid state amplifier we get all these and more. Solid state amplifiers don’t lose emission and don’t blow when the load is disconnected (in properly designed circuits). They don’t like short circuits. They don’t like EMI. But they are instantaneous.
Tubes on the other hand now need to be biased properly for extreme performance, the tubes need to be perfectly matched. Using cathode bias can serve as auto bias. The philes tell us that cathode bias is no where as good as fixed bias. We also know that using a cathode resistor drops the power output. The cathode resistor is an energy hogger. We also know that valve amplifiers give the best high end frequency response (does it?) and adds colour. So if the end result is only to drive the audio spectrum at it’s higher limits we do not need high powered amplifiers so why worry about the cathode resistors. In fact all the amplifiers I came across in the late 60s and early 70s had cathode resistors. What about electronically controlled bias systems? Transmitters radiating thousands of watts peak envelope power have extremely advanced biasing circuits.
When replacing tubes it is almost always necessary to set the bias current to original spec. Ironically we do the same with semiconductor amplifiers yet they are more thermally unstable than tubes.
My belief is that the best amplifiers have the least components. Direct coupling. No output transformer. No phase shift. The pundits of semiconductor amplifiers positively raved about the dispensing of audio transformers – in fact they even developed single ended triode designs without output transformers to prove their point – Now we rave about the inclusion of audio transformers. I have not heard any tube amplifier produce the bass of a good class D amplifier. Possibly even a good class A or AB semiconductor amplifier. As one gets older, bass becomes more intolerable and we don’t hear the high frequencies. Possibly at 60 years of age we have cut-off of about 14kHz. Definitely if you were in a rock band.
I also note that many manufacturers are spending oodles of cash on the cosmetic appearance of their tube amplifiers. This will not enhance listening quality but will definitely increase the wow factor. Manufacturers of consumer products have taken the road to include a tube pre-amp into their circuits. This baffles many people, including me. The wow factor again. Transistors driving tubes, tubes driving transistors.
The 1000W amplifier with four 813s
There are two huge advantages of using tubes over their semiconductor sibs and that is, one needs less active devices to generate more power and the second, more importantly, when looking at extreme powers a tube has absolutely no match. The disadvantage of course is cost and weight. The link here will take you to the Champ 1000W RMS amplifier. Simple to build, lethal voltages, so be warned.
I definitely think tube amplifiers have something going for it in terms of amplified guitar use but from a high fidelity point of few, pre-distortion, I doubt it. Guitar amplifiers are more often than not designed to introduce distortion. We no longer stick to so called Hi-Fi parameters anymore which is a pity because we cannot compare apples to oranges. For many musicians, a tube amplifier for home monitor use is ideal but not for touring. The weight is certainly not for ideal for shipping and dollar for dollar the tube amplifier cannot compete with power to weight ratio advantages of semiconductors.
And now for something completely different…
Auto Bias – modern techniques
As mentioned previously, why battle changing bias currents to balance each tube individually when it can be had for next to nothing. In fact, also as mentioned prior, the cathode resistor/grid leak combination worked very well for thousands of years but the modern approach is extremely simple: Here we refer to tubecad.com, innovative ideas but yet you will see them in a lot of tube audio circuits.
Articles on tubes versus semiconductors:
- All About Ciruits
- The Audio Archive
- Rod Elliott’s Pages
- Making a transistor turn green – Really rock solid tube pages of John Chambers – Champ Electronics
- And finally, putting this article into perspective: http://www.amp-fix.com/Links.htm (another clever man, John Beer) and authorised dealer of what this article is all about: http://www.tubesync.co.uk/
In most high powered circuits, such as the Champ, the output tubes are driven by a fixed bias system derived through potentiometers and cathode followers. This is simple enough arrangement – high powered tube amplifiers rely on the users knowing what is good and what is bad. Losing bias to any of the grids is going to cause the tubes to overheat and fail (even burst). Buy the best quality components for these high energy amplifiers.
Auto Bias – summing it all up
There are many tube amplifiers today where the die-hards have stuck to using tubes throughout the design (e.g. including rectifier) and indeed if I was to build any of these monsters up to 100W I would no doubt follow suit but one thing I would change would be to add safety circuitry to the output stage using semiconductors in the big brute models. The 813 is a case in point. At sometimes up to 200U$ per piece we don’t take chances.
Many amplifiers these days are not taking into consideration auto bias is a sure way to go. Properly designed circuits do not affect the overall performance of the amplifier and yes, they could be used for setting the bias current automatically. Never say it cannot be done….
Next: Bucket Brigade Digital vs Tape Echo – coming up….
[P.S. I love valve amplifiers 🙂 so don’t get any funny ideas. Vinyl, tubes and great sensitive speakers. The glow of good sound – possibly a placebo for solid state? I doubt throwing in the odd semiconductor is going to damage the sound, auto-bias is a sure way of making things simpler].