[Absolute Zero] Design Amp with minimum component (understanding too)

[guitar89]

I'm going to build a symmetrical amplifier with minimum component (seems stupid) and I'm lack of knowledge (scratching from zero now), thus I name it 'Absolute Zero'.
(Just started to learn simulation too)

The aim of this amp is to let me understand (and some beginner too) by building amplifier beginning with only transistor (BJT now, maybe others later), then increase components that only 'necessary' to make the amplifier work sufficiently.

and the second stage is to make this amplifier sound nice and clean (improve the amplifier in short)
then maybe third stage is to make high power/performance, but we just make the first stage first.

So here I attached the schematic I have drawn, with symmetrical output transistors only, each individual responsible for halfwave(AC positive and negative portion) of the audio input.
Welcome to add any component to make this amp work stable and safe FIRST, and state (briefly explain would be good) the reason/function of adding it. (each single component explanation would be best)

[benb]

I'll go first, for the s***s and grins. The one component I would add is a single-cell battery, in place of the connection from the input to the top transistor's base, with the positive connection to the base. An alkaline cell with 1.5V would be good for silicon transistors (which I presume these are), but it's possible for that to be a little too much bias voltage, causing excess idle current. A NICD or NIMH at 1.2 volts might work better as far as lower idle current, the tradeoff of course being greater crossover distortion, but still a substantial improvement over the original circuit.

I'm reminded of an Albert Einstein quote that something "should be as simple as possible, but no simpler."

[guitar89]

Originally Posted by benb

I'll go first, for the s***s and grins. The one component I would add is a single-cell battery, in place of the connection from the input to the top transistor's base, with the positive connection to the base. An alkaline cell with 1.5V would be good for silicon transistors (which I presume these are), but it's possible for that to be a little too much bias voltage, causing excess idle current. A NICD or NIMH at 1.2 volts might work better as far as lower idle current, the tradeoff of course being greater crossover distortion, but still a substantial improvement over the original circuit.

I'm reminded of an Albert Einstein quote that something "should be as simple as possible, but no simpler."

I'm agree that quote, but my motive is not to make it THAT simple, but instead add on each component in order to make it work, the motive is to make me understand. (other tutorial or books I have seen always state the whole circuit as a start, already more than 5 component or so, thus couldn't understand what is the use of each component)
Hm... I also think of biasing for quiescent current, but how to achieve that ? (just putting resistor is fine ? as attached. seems easy to change Iq when temperature rise)

[DF96]

The two bases are still shorted, so no bias current just wasting supply current. You must put something between the bases. Bare minimum is a resistor, but this will give you poor temperature stability. Better would be a pair of diodes, as these would roughly match the base-emitter voltage of the BJTs.

[guitar89]

Originally Posted by DF96

The two bases are still shorted, so no bias current just wasting supply current. You must put something between the bases. Bare minimum is a resistor, but this will give you poor temperature stability. Better would be a pair of diodes, as these would roughly match the base-emitter voltage of the BJTs.

Do the diode also for thermal compensation (prevent thermal runaway) ?
Is it necessary to thermal couple the diode with output transistor ? like NJL3281 pair ?
Does the resistor I add there appropriate ?

I have attached the redrawn diagram, adding 2 diodes there. (should it be thermally coupled like NJL3281 pair ?)
The direction of diode can be change ? or this is wrong setup (I have the feel of that)
(A) configuration seem no different, also shorting.
(B) configuration seem no shorting, but no signal conduction. (maybe should reverse the direction.

Additional question : Does the 'usual' amplifier have bias current go through to input/VAS driver transistor ?

[dillmeister]

I am also very naive to the function of all the parts of an amplifier circuit. I have built a few kits, but to be quite honest I did not understand the function of each part. I think this is a great idea for a thread and I will be following it closely and I have my iron ready and my Radio Shack board ready to learn along with you.

[DF96]

B is correct. The signal will travel through the diodes, but you can also add capacitors in parallel with them. The diodes should be thermally coupled to the BJT.

The standard amplifier arrangement has output base bias current going through the VAS collector circuit. The output base voltage is set by a Vbe multiplier circuit, rather than a pair of diodes.

[guitar89]

Originally Posted by dillmeister

I am also very naive to the function of all the parts of an amplifier circuit. I have built a few kits, but to be quite honest I did not understand the function of each part. I think this is a great idea for a thread and I will be following it closely and I have my iron ready and my Radio Shack board ready to learn along with you.

Haha, thanks. I personally also build, repair, modding few amplifier, similar situation with you. I'm curious about what stage I would reach. (I'm learning simulation at the same time, multi-tasking i think)

If you saw anything that you doesn't understand,not sure, or not stated, please voice out (because maybe i would miss out something) since this is a total learning session

Originally Posted by DF96

B is correct. The signal will travel through the diodes, but you can also add capacitors in parallel with them. The diodes should be thermally coupled to the BJT.

The standard amplifier arrangement has output base bias current going through the VAS collector circuit. The output base voltage is set by a Vbe multiplier circuit, rather than a pair of diodes.

hm.... really ? it seems that positive halfwave 'should' go through the top transistor, but it seems the diode is in reverse bias, thus no positive signal allow to go there.
The opposite also same, the negative portion of signal can't go to the bottom transistor.

About the capacitor's function ? (block DC ?) why not we eliminate the diodes and put capacitor in between ? this will block biasing DC and allow signal to flow freely.

Based on your answer, i assume that Iq only apply to VAS and output transistor ?

I have attached another schematic, with also have (A) and (B).
(A) - Using Capacitor at value which allow AC signal to pass, but block bias DC.
(B) - Bias DC are block by diodes (reverse bias) and AC signal are allow to pass through diodes (forward bias)

[DF96]

Your new A would work, with the output devices now receiving current bias rather than voltage bias.

Your new B would only provide any output with a big input signal, and would suffer from serious crossover distortion. Essentially, it would only respond to inputs with an amplitude greater that about 1.2-1.4V and would delete the central part of the waveform near the zero crossing.

I don't want to rain on your parade, but your questions suggest to me that you still have a lot of reading to do on fundamentals before you can design even an 'Absolute Zero' design. Putting up bad designs then asking people to point out the flaws in them is an inefficient way of learning the basics for both pupil and teacher!

[SY]

I would absolutely recommend you spend some time studying Horowitz and Hill. Not only will you learn the basics that you need before trying to design something yourself, they have a cool pedagogical trick that's apropos here: at the end of every chapter, they show a set of "Bad Circuits" for the reader to diagnose. A circuit very much like yours appears at the end of one of the early chapters.