At your own risk.
Describing from end to beginning.
The speaker used is a vintage Philips 12in full range.
The Darlington transistor used is an NTE245.
Initially a battery of 6 volts was used.
Battery (-) was connected to emitter.
Battery (+) was connected to one terminal of the speaker.
The other terminal of the speaker was connected to the collector terminal of the Darlington. This makes the speaker the load resistor itself. So by proper bias the Darlington operates in class A and in a sense the transformerless speaker circuit is in itself in class A. (Class A operation means electrons flowing all the time, even when no signal is applied).
A variable resistor for biasing was connected between collector and base.
For a 6 volt battery a resistor of the order of 3.5KΩ moves the speaker cone as much as is needed shifting its equilibrium position but not excessively.
The Yamaha electric bass, (connecting directly to the pickup by bypassing tone control and volume control) was connected through a capacitor to the base of the Darlington.
There was no sound.
Then a portable Sony CD player was connected through the capacitor to the base. Sound was amazing, clear and dynamic as expected from just one transistor between CD player and speaker.
Then the Pleiades V5 was connected between electric bass pickup and power amp to bring the electric guitar to aux output level. Signal was taken out of the V5's second EF183 electron tube from its anode.
The sound was very nice and smooth.
By that time a 12V battery was being used. It gives much higher output. The biasing resistor has to be changed to about 17.5KΩ.
There is great risk of thermal runaway. The speaker cone keeps shifting as the temperature of the transistor is increased. The biasing resistor was manually increased. It is risky and tricky. Heatsink must be used and if thermal runaway persists a potential devider biasing method should be tried.
Everything was connected in the quickest way possible. Connections with crocodile clips with cables. This gives a nice inbox sound The speaker unbaffled. This gives a nice unboxy sound, and a very gentle 6dB per octave bass roll off. The Darlington transistor was without heat sink.
The smooth creamy sound of class A operation was a joy to hear.
The electric bass was being played along while listening to the radio.
There is some compression as expected as the V5 was designed for a tiny microphone signal. This compression was suitable for bass. More of the compression may be coming from the power Darlighgon being overloaded. In general the harder the bass was played the more compression was taking place. In class A amplifiers distortion tends to zero the smaller the signal is just like our ears.
The bass is connected across the secondary of the input transformer of the Pleiades V5. So there is a 5 Henry inductor across the bass guitar pickup making a Pleiades high pass filter. This compensates for excessive bass due to Fletcher Munson curves etc. And also makes every detail of the bass guitar strings shine.
Hiss has almost non existent but in fact there should not be any. Noise was non existent as everything is battery powered. From previous experiments hiss may be due to radio frequency pickup. When the V5 is connected to a balanced microphone hiss noise is almost no existent due to the extremely low anode and heater potentials supplied by a 3.7 li ion battery.
The overall result was very pleasing. Not very loud but of excellent quality.
Signal path:
Yamaha bass guitar - Pleiades V5 preamplifier- Pleiades power amp - Philips full range speaker
This little experiment was done inside a semi echo chamber. Even better bass definition is expected at a more normal room with Pleiades bass (panel membrane) absorbers.
Somwtimes the sound leaves something to be desired when listening to the power amp with a CD source. It sounds sometimes that the distortion does not tend to zero when the input signal tends to zero. This may be due to the voltage, current input nonlinearity of a bipolar transistor. This effect is absent on electron tubes and they are voltage controlled current output devices. More experiments are needed with transistors such as 2N3055. PNP germanium transistors (AD140 for example) give great results. How would power JFETs sound like on such a simple circuit?
On second thought there is no reason why such circuits should not sound the greatest. It can certainly be improved by inspiration from Neve, top of the world microphone preamplifiers using single ended 2N3055 class A operation.
Reference:
Audio Amplifiers - edited by J.R. Davies - Data Publications Ltd, London - chapter 7 - page 41 - 2 transistor miniature A.F. Amplifier
