Understanding Valve Amplifiers - By Charl Du Plessis

Background

In 1906 Lee De Forest invented the first radio valve "the Audion" which had three elements: the cathode, a metal grid, and the anode. This is the triode type of valve. It was capable of some amplification. Edwin Armstrong (who understood these things better than De Forest) did more development and introduced regenerative feedback, which led to much more effective gain and paved the way for the radio industry.

When applying a voltage of 6.5v to the cathode as the starting point, from where the cloud of electrons flows past the metal grid, to the anode, shows the direction of flow. How you control the flow is essentially how this valve will amplify.

Valve amplification circuits work by varying the voltage on the grid, and the stream of electrons from the cathode to the anode increases. If we negatively increase the voltage on the grid, in comparison to the cathode, we will measure as a result, a higher voltage at the positively charged anode. So essentially during this process, the valve will convert smaller alternating voltages, into larger alternating voltages. Bias refers to the control of these voltages and current, which in turn affects the power output. Increasing the Bias runs the tube hotter and increases power, but will shorten the life of the valve. Where there are a number of power valves in an amplifier, it is highly desirable that they be biased very similarly to achieve a balanced output.

As power valves were developed, we came to understand that more than one grid was required to better control the flow of electrons from the cathode to the anode, as the high velocity of the electrons at the arrival point of the anode may knock secondary electrons lose from the anode metal plate, this will usually reduce the effect of amplification.

As the charge cloud of electrons moves past the negatively charged metal grid a second grid, constant positively charged, was introduced to help reduce this effect. This became the tetrode type of valve.

Later on in Europe a third grid, the "suppressor grid", was introduced to immediately bounce electrons back to the anode, should these by any means be knocked off course. Adding the three of the grids, plus the cathode, and the anode results in 5 components and naturally this type of valve became known as a pentode.

What these valve component quantities indicate, is that naturally from the triode to the pentode we have on the positive side an increase in power output, however this may result on the negative side to have an increase in distortion. This balance clearly show that we have to be very careful in the amplifier design process so as not to opt for as high power delivery as possible, as this will logically negatively affect the end result in the quality of the sound reproduced.

This logically indicates the lower volume level used from a valve amplifier, the cleaner and purer the sound will be, as a result of a lower distortion factor from the type of valve used. This is quite contrary to solid-state amplifiers that usually work in the opposite way, as in higher distortion at lower volume levels, and less distortion at increased volume levels. This may help to explain why most solid-state amplifiers tend to sound better at a sweet spot/range in the volume control turning ratio.

Amplifier Design

If we look at the circuit diagram of a valve amplifier we will see that the components are logically placed in a chain of events. Firstly there is the mains power transformer. This is located on the front side as far away as possible from the output transformers. In high quality amplifiers the mains transformer usually has a choke circuit located next to it, to regulate the flow of voltage.

From here the circuit continues on to the pre amp section. This pre amp circuit usually consist of the triode type of valves. As in this part of the circuit we do not need to amplify large voltages, and we need to keep the amplification as pure as possible.

From this point onwards we need to start making decisions as to what we want to do with the power stage of the amplification circuit. We know from experience by studying the design that we cannot expect a huge power output from triode valves, and we should not expect too much power from pentodes based on the distortion increase problems discussed before.

Hafler and Keroes brought a design to the market in the 1950's to alleviate the above pondering. They referred to their design as Ultra linear, whereby the use of a valve as a phase splitter to work with pentodes in the power amp stage, by alternating the current, so at any time, one valve will be conducting, and the other will not in opposite phase. Secondly both halves of the output transformer primary windings are tapped, and the taps are connected to the screen grids of the power valves.

In a good design we normally ensure that the power valves do not work very hard so as to give good sound and a reasonable valve life expectancy.

As indicated before these power valves are usually connected to an output transformer. The function of this output transformer is to convert the dangerously high voltages usually of about 400v on the primary side of the transformer, to safer low voltages and high current output on the secondary side, to drive the appropriate speaker load.

Therefore over and above the importance of very high quality components needed for the mains transformer, and the electronic components in between, we now come to "in my opinion" the most important part in the discussion of a valve amplifier? The component known as the output transformer. This device can make or break the quality of a valve amplifier, as its design is what affects the performance of our valve amp from a mediocre level to what we perceive as high-end sound quality. Let's have a look and see what this entails.

There are of course lost of interesting points to discuss but most designers agree on the following five quantities as of utmost importance to the end result in the performance of the output transformer.

If we look at the circuit diagram of a valve amplifier we will see that the components are logically placed in a chain of events. Firstly there is the mains power transformer. This is located on the front side as far away as possible from the output transformers. In high quality amplifiers the mains transformer usually has a choke circuit located next to it, to regulate the flow of voltage.

From here the circuit continues on to the pre amp section. This pre amp circuit usually consist of the triode type of valves. As in this part of the circuit we do not need to amplify large voltages, and we need to keep the amplification as pure as possible.

From this point onwards we need to start making decisions as to what we want to do with the power stage of the amplification circuit. We know from experience by studying the design that we cannot expect a huge power output from triode valves, and we should not expect too much power from pentodes based on the distortion increase problems discussed before.

Hafler and Keroes brought a design to the market in the 1950's to alleviate the above pondering. They referred to their design as Ultra linear, whereby the use of a valve as a phase splitter to work with pentodes in the power amp stage, by alternating the current, so at any time, one valve will be conducting, and the other will not in opposite phase. Secondly both halves of the output transformer primary windings are tapped, and the taps are connected to the screen grids of the power valves.

In a good design we normally ensure that the power valves do not work very hard so as to give good sound and a reasonable valve life expectancy.

As indicated before these power valves are usually connected to an output transformer. The function of this output transformer is to convert the dangerously high voltages usually of about 400v on the primary side of the transformer, to safer low voltages and high current output on the secondary side, to drive the appropriate speaker load.

Therefore over and above the importance of very high quality components needed for the mains transformer, and the electronic components in between, we now come to "in my opinion" the most important part in the discussion of a valve amplifier? The component known as the output transformer. This device can make or break the quality of a valve amplifier, as its design is what affects the performance of our valve amp from a mediocre level to what we perceive as high-end sound quality. Let's have a look and see what this entails.

There are of course lost of interesting points to discuss but most designers agree on the following five quantities as of utmost importance to the end result in the performance of the output transformer.

1. The transformer turns ratio,(a): This refers to the ratio, of the number of turns in the output transformer primary side, in comparison to the number of turns on the secondary side. You will find that the number of turns on the primary side is significantly higher in comparison to the secondary side; a ratio of between 20 and 30 to 1 is quite common. The function of the output transformer is therefore to act as an accurate step-down device. Secondly the transformer ratio needs to convert the speaker impedance to the primary winding impedance.The turn's ratio enables us to calculate the relationship between the currents, the impedances, and the voltages on the primary side, and secondary side of this transformer. The result is to optimally couple the power valves and the loudspeaker impedance.

2. The primary winding inductance,(Lp): This refers to the size of the core, the number of the primary turns, and the magnetisation of the core. From a sound perspective the primary inductance is important, as this quantity will indicate as to how the transformer will handle the low frequency part of the audio spectrum. A sufficiently large value (Lp) is necessary to allow low frequency voltages to be faithfully, and true fully converted from its primary to the secondary side.

3. The primary winding leakage inductance,(Lsp): With this quantity we need to ensure that ideally each turn on the primary see each turn of the secondary winding. This would mean that the magnetic field lines should enclose the turns on both sides. A high coupling factor (K) of 1 will result in a very good high quality reproduction of the high frequency range of the audio spectrum.

4. The primary winding internal capacitance,(Cip): This is the second factor that could influence the high frequency range of the output transformer.Because the turns of the windings of our output transformer are so close together, there will be degrees of capacitance between the adjacent windings. The higher the frequency, more of the current will flow through a capacitance, this results in that the impedance will decrease, and that less current will flow through the windings. This capacitance effect needs to be minimised so as not to affect the frequency range over which the output transformer can work properly. In high built quality transformers this effect can be minimised out of the audio frequency range to above 80khz so as not to negatively affect this part of the audio frequency range.

5. The winding resistances, (Rip) and (Ris):When we wind wire around a core we have to consider internal resistances.As this resistance in the core and in-between the primary and secondary windings, effect results in a power loss for the output transformer. This affects the efficiency of the transformer and therefore what we can expect from the design. This resistance effect of power loss can significantly be minimised by using better quality wire material and optimum wire thickness in high built quality transformers.

The component next in line is the speaker terminals. The shortest possible route from the secondary winding of the output transformer needs to be followed with appropriate secondary load of four and eight ohm taps, to suit the specific speaker load, so as to maximise the damping factor for a particular load. Secondly the quality of the binding posts is important so as to minimise information loss. We can improve the quality of the binding posts by adding as much pure copper content as possible while maintaining a rigid structure to ensure longevity in the field.

FAQ - Frequently Asked Questions

Here are some questions that people might ask in trying to understand a valve-based amplifier.

(1) What care do I need to take; how do I bias the amp? (2) Do I need to bring it in to the workshop for a technician to do?. (3) How expensive will that be? (4) Can I do it myself? (5) How long does the amplifier stay biased? (6) Do I need to get the bias done once a week, once a month, once a year? (7) How long do the valve components last? (8) How long has the valves been used? (9) Can I replace the valve components myself? (10) Do I need to use the same type of valves? (11)Can I use different valves brands to create my type of sound? (12) Are the valves easy to get hold of? (13) What influences the sound? (14) How does one know the difference between good and very good?

Answers: The Plessis Model A Valve Amplifier

We believe that the above questions have been addressed with the Model A amplifier. Most people should be able to set the bias themselves by following the simple setup procedure of pushing the dip switches, looking at the built in MA meter, and adjusting the bias via the toggles clearly indicated in front of each valve.

The amplifier has a built in four digit electronic timer that will display the time of the power amplifier valves used in hours. This display comes with a reset facility when the valves need to be changed and a display off selection if required. This will give people confidence in the sense that everything is easy to control and maintain.

We included a toggle switch located on the back of the amplifier to choose between using the amplifier in ultra linear (higher power) mode and triode mode for the enjoyment of different styles of music and speakers to suit the taste and needs of the audiophile/music lover.

The input terminals are made of good quality copper so as to maximise contact, reduce information loss and to ensure longevity.

The volume control is located on the pcb and not where the rotary control is. This of course reduces information loss.

We have included high copper content 4 and 8-ohm speaker terminals at the back of the amplifier to suit the specific speaker choice of the client.

The Model A uses high quality transformers and electronic components. Secondly the technologies available today in regards to making these transformer components were not available before 1994.

It is of course understandable that these quality components combined will have a significant costing. We minimise the costing to the end user by significantly shortening the supply chain; the chain goes manufacturer, retailer and end user. I know of no other product that can compare in regards to this short supply route.