audio with realistic timbre
 
search

Preamplifier design using power triodes

Do preamplifiers add to sound?
The use of a preamplifier is now-a-days technically debatable, reading all kinds of forums proves this. From the old era where sources output voltages were relative low, sometimes had high output impedance and tone control was common, a preamplifier stage was useful. In professional studios lots of filtering and levelling of signals is needed which calls for a line level stage, also a sort of preamplifier. In modern (home) audio there is no technical need for a preamplifier, output impedance and -level of sources are suitable for long cabling and almost all (power)amplifiers have sufficient sensitivity to do without. Tone control is most of the time not present in modern HiFi-equipment and on top of this most modern sources feature (remote) attenuation as well.
So why all the fuzz about preamplifiers? Imho the answer is simple: Preamplifiers DO add to sound, and that is what we like. It allows us to match music reproduction to personal taste. There simply is no better reason than doing just that, music touches the senses and senses are personal. And, of course, practicing tone control without a treble or bass knob makes us feel very good..😇

Design is key.
Lots of choices to be made. And each choice has an influence on sound. It takes time, insight and funding to be able to predict the sound. Experience and patience are key here. The more parts are present in the preamplifier, the harder it becomes to be able to predict the sound on beforehand. This Research and Development is seldom done by DIY-ers. Most of the time a design is made or borrowed and subsequently build. And of course that build will be THE best sounding when completed! And then, intriguingly, only after a few hours, days or weeks, a tiny change in the design makes it even better sounding. And that is perfectly fine, since that change added to the sound and touched the senses which are personal. No need to discuss or argue that modification. If it pleases, let it be. Trust your senses. The most easy part to swap are the tubes (or valves). Just replace them by another brand or equivalent type, and a change will add to the sound. Again, personal taste will determine if it sounds better. Perfectly fine! No need to argue about personal taste.

Preamplifier triodes.
Many vendors categorize their tubestock to application: Rectifiers, powertubes, preamplifier tubes (and so on). Most of the times preamplifier tubes are the more modern, smaller 7- or 9-pin indirectly heated ones. The reason for this is that these can be fed AC (have good isolation from heater to cathode), are less microphonic and less susceptible to transformer stray fields. This is very logical as a preamplifier needs to be silent apart from the music signal. Interestingly most of these preamplifier tubes have relative high gain, more than on average needed for a preamplifier. But since these are quiet, this is most of the time no issue. Also these run on relative low current and have a relative high internal resistance. Their inner structure is small thus more rigid and since current is low, expansion of inner structure is marginal. Energy needed from the heater is marginal, so total power supply demands are not huge and as result the build can be compact. Indeed very suitable for preamplifiers..

Power triodes.
These have typically lower gain compared to preamplifier tubes and have less isolation between cathode and heater, or heater and cathode are the same when directly heated. Also they run at a higher current and have lower internal resistance. That lower resistance comes in handy as most of the times an output transformer is used to drive the speakers, and the lower the tubes' internal resistance, the less challenging becomes transformerdesign. As result these are designed to amplify larger signals coming from its driver and that is why lower gain is no problem. Also the power triodes' bigger internal structure is susceptible to vibrations and transformer stray fields, resulting in being microphonic and causing hum when not used in the last audio stage. Also these are running on higher current thus the inner structure has to dissipate more heat which causes the metal parts to expand. All of this is normally no issue when used in the last audio stage having low gain and their amplified signal being stepped down using the output transformer.

Older triodes, pre-WWII.
Many of these have characteristics that are alike power triodes. Ergo, most of these were designed as power triodes, be it very modest in power handling. They are visually appealing to many, having globe shaped glass envelopes, experimental inner structures and are bigger in size compared to preamplifier tubes. As indirectly heated cathodes were yet to be invented, these are directly heated. And at that time sound reproduction beneath 70 Herz was a rarity, so hum was less an issue back then. Overall these older triodes expose the same challenges as powertubes when put into service in a preamplifier.

Ideal choice?
Preamplifier tubes are technically the ideal choice. No argue there. However some of the power triode properties are interesting as well for preamplifier service! So nowadays more and more preamplifiers are designed around power triodes and older triodes, as their sound (and looks!) pleases people on average more than preamplifier tubes. So seemingly technical and visual characteristics of these tubes add more to preamplifiers sound than preamplifier tubes do. This is so logical as, again, it is about senses, and these are personal thus add to perception of sound as well. And, aesthetics are important as well, these add also to the senses..
But how about the ideal choice? Looking at the construction of power triodes and older triodes, these must be prone to hum, be microphonic and be noisy. So they are! This is a challenge that faces lots of DIY-ers. Building a preamplifier using these triodes often result in lots of cries for help as it hums, is microphonic or is noisy. As much you like the sound, in the end these defects will override pleasure, unfortunately, and the preamplifier is set aside.

A solution.
Yes, A solution, not THE solution. There are many roads to pleasureville, this is just one of them.
For triodes there is a relationship between amplification factor, current, mutual conductance and plate resistance. Certain combinations are suitable for specific use. So when putting power triodes to service in a preamplifier, respect the conditions that are appropriate for that powertube. Do not use them alike conditions for preamplifier tubes.. What do I mean with that? well, do not run these below their advised operating points and do not amplify attenuated line level signals.
See below a list of solutions for the identified challenges.

  1. First of all, use a representative operational point for the triode, not to low in current nor supply voltage. Yes, that might mean building a hefty power supply!
  2. Hum from transformer stray field. Use a separate power supply chassis. As my friend Morgan Jones once stated: "Use distance, it is cheap and plenty available." Ideally only the power supplies' last capacitor should be incorporated in the preamplifiers chassis.
  3. Hum from cathode. Nowadays hum by heating a directly heated cathode can easily be overcome by using regulated DC. Alternatives like TentLabs low noise filament modules and Rod Colemans' heater modules solve all hum issues coming from the cathode. These work like a charm, with a pro for Colemans regulator, as this one allows you to bring in DC from the separate power supply where the Tentlabs is fed AC (but is very compact and efficient).
  4. Noise from expanding structure. Use a line-out transformer with a step down ratio of minimal 5:1, which often sells as a 10 to 15 KiloOhms to 600 Ohms transformers, which, by the way, is not always true (read Caveats, any?). Further on, see g as well.
  5. A to small input voltage. Use an input transformer to step up the incoming signal, step up what you step down using the line-out transformer. A good and affordable option is Lundahl LL1538XL.
  6. Microphonic. The combination of d and g will take care of this as well.
  7. A to sensitive power amplifier. Place the attenuator at the output of the preamplifier, This will supply only the to be amplified signal to the power amplifier, introducing an optimised signal to noise ratio.

When all these solutions are applied there will be almost NO noise present at the preamplifiers output when using power or older triodes. This depends a bit on type and age of the used tube, but at listening levels this type of design will be free of noises. Of course the quality of the transformers will influence the sound quality as well, but what the powertubes added to the preamplifiers sound will remain and keeps touching the senses. Music!

Concept design preamplifier with powertriode
Concept design preamplifier using power triodes. Beware that this design alters absolute phase by 180°.

Caveats, any?
Of course! There is no free lunch, everything has downsides as well.. When using a step-down line-out transformer, the input impedance of the power amplifier has to be considered as well. This is because of the input impedance will reflect its load to the line out transformer and may influence the actual load seen by the power triode. Now this does not have to be a direct problem as the optimal load will guarantee maximum power delivered by the powertube, something that is most certainly needed for driving speakers. However this power is not at all needed to drive a power amplifier, but a powertube will perform better when optimal loaded. This depends on the transformer design and many aspects define a good transformer. Old stock (vintage) (line) transformers are often specifically wound to 600 Ohms output, and may need correction by using zobel-networks etcetera. Using these, it might be worthwhile experimenting loading the secondary of the line-out transformer with a fixed resistor resembling the secondaries impedance, thus forcing the transformer to reflect the optimal impedance at the transformers primary for loading the power triode. Or, following the solutions, match the attenuators' impedance to load the secondary.. An interesting read (or rant) can be found by Dave Slage here. Following his statement most of the power amplifiers that have a 10K to 100K input impedance would not be appropriate to mate well to a line-out transformer with a step-down ratio. This is exactly why most of my power amplifiers have an input transformer, be it 1:1 or with modest step-up (*), thus being suitable for a type of preamplifier described in here. So this road to pleasureville might not be the road for all drivers..
Apart from this, my experience is that an autoformer (Slagle? Tribute?) will perform best as attenuator in this design.

(*) As Dave put it very clearly: " ..that is enough to give any engineer a heart attack.." but doing so has some advantages: Influence of interlinks are less prominent as it is now low impedance and they can be yanked out a running amplifier without loud noises being amplified (and it sounds way better..).