Lowest possible component count, simple transistor amplifier.

Thread Starter

john1211

Joined Sep 17, 2018
12
Hi Thanks for reading. Okay I want to build a simple audio amplifier with absolute minimum components.

So far with my amateur attempts i have built a single stage amplifier using a 2sc5200 transistor, with a 1uf capacitor and 1000 ohm resistor as an input stage, but it sounds like a glorified headphone amp, very little output about half a watt.

Do you have any minimal component designs which can put out about 5 watts up to 10 watts, or if this is not possible a two stage amplifier consisting of a driver transistor followed by an output transistor, again minimal components. I have 2sc5200 transistors so will appreciate any designs incorporating these.

Thanks
 

Thread Starter

john1211

Joined Sep 17, 2018
12
Thanks! Very true.
The link you supplied shows a lot of chip amps, i am more interested in transistor amps.

I want to learn and get my hands dirty and build my own, for educational purposes.

I did this before but used you tube videos as a source which did not work out well, this time i would like to ask experienced people in advance
 

Ian0

Joined Aug 7, 2020
9,667
I can do 8* transistors, 4 capacitors and 8 resistors. Would that do? I’ll draw it this evening.
*Do you count a BCV62 or a BCM857 as one or two? If you just count it as one, then it’s 6 transistors.
 

BobTPH

Joined Jun 5, 2013
8,812
Since you have not specified what the input i, I can do it with no components!

input spec: 10V RMS with < 1 Ohm output impedance.

So, what is your input spec? And output as well, i.e. what is the load impedance.

Bob
 

Papabravo

Joined Feb 24, 2006
21,159
There is no evidence to suggest that a low component count will produce a good amplifier. I can choose to have a less stable bias for the transistor and save a resistor or two. Is that a compromise worth making? I don't think so and IMHO opinion this is a silly quest.
 

zophas

Joined Jul 16, 2021
165
@john1211 You are on the right track. But single stage 2 transistor amps will never give you much of an output. But those single stages can form the building blocks for a better and more powerful design. Also there are IC's you could use to make an audio amp capable of more than 0.5 watt. You tube video is about the worse place to go to learn anything about electronics. And BobTPH is right, you need to consider more than just minimum component count. You get what you pay for.
 

MisterBill2

Joined Jan 23, 2018
18,167
So the goal is an audio amplifier to drive a speaker, powered from some unspecified voltage and with some unknown input signal voltage. But with the output power specified at 5 t ten watts, it gets a lot trickier, especially if the distortion level is supposed to be low, and the amplifier is not supposed to go into thermal runaway.
So here described is the amplifier circuit WITH the resistor to reduce temperature effects on the idling current.

Two transformers and two resistors and one NPN transistor with a fairly high Hfe:
The input has a transformer with the input signal feeding the lower turns count primary winding. The secondary winding connects to the base of the transistor and the other end to the supply negative.. From the base, also, a higher value resistor connects between the base and supply positive. From the emitter, a lower value resistor connects to the supply negative. The collector of the transistor connects to one end of the output transformer primary, the other end of the output transformer primary connects to the supply positive. The secondary of the output transformer connects to the speaker.
There you have a complete single stage amplifier able to deliver quite a few watts, depending on the transistor used and the supply voltage provided. The emitter resistor value is selected to keep the collector current fairly constant as the transistor heats and the resulting leakage current rises. The base pull-up resistor is selected to bring the collector current up to about half the desired peak value.
This circuit is not original with me, it was used in many of the mid-sixties car radios that had a one transistor output stage. I just revised it to use an NPN trnsistor instead of a PNP transistor.
 

AnalogKid

Joined Aug 1, 2013
10,986
You can do a one-transistor emitter-follower or source-follower with a power BJT or power MOSFET. This configuration has no voltage gain (gain = 1). For a 5 W output it will dissipate over 10 W in heat. If you want a gain + power stage, the next step is a one-transistor common emitter or common source stage. Again, lotsa heat plus not great fidelity.

ak
 

hrs

Joined Jun 13, 2014
394
There is no evidence to suggest that a low component count will produce a good amplifier. I can choose to have a less stable bias for the transistor and save a resistor or two. Is that a compromise worth making? I don't think so and IMHO opinion this is a silly quest.
OP said "I want to learn and get my hands dirty and build my own, for educational purposes".
There is no evidence to suggest that a high component count will produce a good learning experience.
 

Audioguru again

Joined Oct 21, 2019
6,673
A single transistor as an amplifier usually has the power supply, transistor and speaker all in series with the speaker as the transistor's collector load. Then all three produce a lot of continuous heat. The speaker has DC in it which is bad since it works properly with AC, not DC.

An amplifier usually has a push-pull pair of emitter-follower transistors with no voltage gain at its output and is driven with a transistor or two producing voltage gain. With enough voltage gain then the amplifier can have plenty of negative feedback to reduce distortion.

If the power supply has dual polarity then the speaker can connect directly to its output and not have DC flowing in it. If the power supply has a single positive voltage then the speaker should use a series capacitor to block the DC from the amplifier output.

The amplifier should be class-AB with a small idle current that can be set with two series diodes or adjusted with a trimpot. The small current eliminates crossover distortion.

Amplifier modules from China lie about their output power. One listed earlier using an 8002 IC says its output is 3W but that that is with extreme distortion into a speaker that is only 3 ohms. The amplifier becomes very hot producing 0.5W into 8 ohms but is cooler producing above 1W when the output is clipping into squarewaves.
It truthfully produces only 1.2W into an 8 ohm speaker with low distortion.
 

MisterBill2

Joined Jan 23, 2018
18,167
A single transistor as an amplifier usually has the power supply, transistor and speaker all in series with the speaker as the transistor's collector load. Then all three produce a lot of continuous heat. The speaker has DC in it which is bad since it works properly with AC, not DC.

An amplifier usually has a push-pull pair of emitter-follower transistors with no voltage gain at its output and is driven with a transistor or two producing voltage gain. With enough voltage gain then the amplifier can have plenty of negative feedback to reduce distortion.

If the power supply has dual polarity then the speaker can connect directly to its output and not have DC flowing in it. If the power supply has a single positive voltage then the speaker should use a series capacitor to block the DC from the amplifier output.

The amplifier should be class-AB with a small idle current that can be set with two series diodes or adjusted with a trimpot. The small current eliminates crossover distortion.

Amplifier modules from China lie about their output power. One listed earlier using an 8002 IC says its output is 3W but that that is with extreme distortion into a speaker that is only 3 ohms. The amplifier becomes very hot producing 0.5W into 8 ohms but is cooler producing above 1W when the output is clipping into squarewaves.
It truthfully produces only 1.2W into an 8 ohm speaker with low distortion.
For a large class of audio stuff the power specification was for operation at 10% harmonic distortion. The concept was that the equipment very seldom reached that point and so somehow it was acceptable. The number of lies associated with consumer electronics often exceeded the number of components in the system.
 

MisterBill2

Joined Jan 23, 2018
18,167
D.D. Wow, that is quite a textbook.. One thing to keep in mind is that at least for the lower powered amplifiers, the trnsistors are germanium PNP devices, which need biasing a bit different from current silicon transistors.
 

BobTPH

Joined Jun 5, 2013
8,812
Here you go. 7 components. 5W RMS into 8 Ohms with 0.5V p-p input.

Note that R1 and the MOSFET will each dissipate about 144 Watts when the amp is idle.

Should I count the two massive heat sinks as two more components.

This was done be a professional driver on a simulator only. Do not try to do this in the real world!

1631553989668.png
Bob
 

dendad

Joined Feb 20, 2016
4,451
This all brings back memories!
I has the Sinclair Z30 amps.. https://rk.nvg.ntnu.no/sinclair/audio/z30.htm and before that, the Mullard 10/10. These had the wonderful ability to protect the fuses as the transistors blew quicker that the fuses if the speaker leads even thought about shorting!
Simple amps are all well and good, but I think it is better to build an amp that works reliably and will give long service.
I still have a PA amp with an "Electronics Today international" ETI-480 100W module (1976 vintage). These, or similar kits are still available.
https://www.jaycar.com.au/sc-480-50w-amplifier-module-kit-version-1-to-218/p/KC5345
 

Ian0

Joined Aug 7, 2020
9,667
This is the lowest I could get the component count for a proper amplifier (21 components, counting the dual transistors as a single component). I forgot the gate resistors when I counted up this morning. Power is 10W into 8Ω. The output transistors are Exicon ECX10N20 and ECX10P20, the 2SK1058/2SJ162 in the simulation is the obsolete Hitachi[Renesas] equivalent. They are not the cheapest, but were chosen to reduce the component count as they need no temperature compensation. You can reduce the component count by another two because you can eliminate L1 if your speakers aren't weirdly capacitive at high frequency (no piezo tweeters or esoteric crossovers) and you can detele J1 and replace R5 by 6.8k if you are prepared to put up with a little hum on the output.

You can scale the design up to many hundreds of Watts just by increasing the power supply and adding more output transistors in parallel.

Spice calculates the distortion at -100dB or 0.001% - I think it might be a little optimistic. Efficiency is around 70% at full power, so heatsinking requirement is quite small.
Untitled 3.pngUntitled 2.pngUntitled 1.png
 

DickCappels

Joined Aug 21, 2008
10,153
If I may make a suggestion, you need to get your feet wet (or hands dirty, or something!). Do some small projects based on simple amplifier circuits you find on the web. Do not be very ambitious about performance and parts count -you will learn about that sort of thing with your gained experience.

In real life instant expertise is very rare (and not necessarily very good). Go with what you can get your hands on and the learning will come naturally, and we will be here to help you through the tight spots.
 

Papabravo

Joined Feb 24, 2006
21,159
OP said "I want to learn and get my hands dirty and build my own, for educational purposes".
There is no evidence to suggest that a high component count will produce a good learning experience.
I would suggest that you have a very low bar for what is "educational". Just because you can design a crappy circuit doesn't mean you have learned anything. I suppose anything that doesn't win you a Darwin Award could be considered educational.
 

Audioguru again

Joined Oct 21, 2019
6,673
I simulated the Simple Amplifier at Hackaday on the other thread that has its diodes upside down. It works poorly.
It acts like a high current class-A amplifier where Q2 is simply a low value resistor and Q3 is the output transistor.
It has no negative feedback.
 

Attachments

Top