Transistor Audio Amp Design

[Steve]

Hi All,

I've designed a basic class-A common-emitter audio amp using a single transistor. This is biased correctly, and develops the voltage over an emitter resistor - the output of which can be seen clearly on a scope.

Now, whilst it's acting as an amplifier - as soon as I connect a speaker across the output (DC blocked with a capacitor) and ground, evrything fails miserably. I'm not sure if this is mainly due to the impedence of the speaker changing the operation of the circuit - or if it's simply an impedence mis-match (I guess the speaker is essentially 'shorting' the output to ground via a very low resistance).

I've seen two designs which seem to resolve this problem:-

Firstly, the use of an audio transformer in place of the collector resistor, with secondary wirings connected to a speaker. I don't like this - as the transformer would have to be pretty large to handle any reasonable currents.

Secondly, the use of a common-collector stage, which has less than unity voltage gain, but of which can provide the current requirements for a low impedence speaker. Is this right?

Assuming the second option is the best ... how would this work in reality? For example - If the first stage of the amplifier produces a signal 20 volt peak-to-peak, this would by far exceed the 5v maximum base voltage of the second stage ... even using a PNP and NPN in push-pull configuration would only give a possible input of 10 volt peak-to-peak ...

Should I somehow forget about the common-collector current stage ... and somehow modify the first stage to delivery the necessary current? I can easily develop a large current with the common-emitter design, but this current flows through the collector load resistor, collector, and emitter. How do I use this current to drive a loudspeaker?

I hope someone can understand the above - and point out where I'm going wrong. I obviously lack understanding somewhere ... but I'm not sure where.

At the moment I'm only interested in standard BJT's, not FET's or similar. I want to understand basic transistors before moving on to others ...

For Ref - the rough circuit (view in fixed width font):-

+ ----------------------------------------- +
| |
RES RES
| |----CAP---O output
input 0---CAP--------TRAN
|
RES
|
GND----------------------------------------GND

Any help would be very much appreciated!

Thanks,


Steve
Bristol, UK

 

[Ian Bell]

Steve wrote:

Hi All,

I've designed a basic class-A common-emitter audio amp using a single
transistor. This is biased correctly, and develops the voltage over an
emitter resistor - the output of which can be seen clearly on a scope.

A common emitter amplifier normaly develops its output between its collector
and ground. However its output impedance is relativeley high (of the the
order of several thousand ohms) and therefore not suitable for directly
driving a loudspeaker. You can easily measure this yourself. First
measure the output voltage with no load. Then try different value load
resistors until the output is halved. The load resistor is now equal to
the output resistance. Note: the circuit you show at the bottom is this
post is *not* what I would call 'properly biased.

Now, whilst it's acting as an amplifier - as soon as I connect a speaker
across the output (DC blocked with a capacitor) and ground, evrything
fails miserably. I'm not sure if this is mainly due to the impedence of
the speaker changing the operation of the circuit - or if it's simply an
impedence mis-match (I guess the speaker is essentially 'shorting' the
output to ground via a very low resistance).

Precisely

I've seen two designs which seem to resolve this problem:-

Firstly, the use of an audio transformer in place of the collector
resistor, with secondary wirings connected to a speaker. I don't like
this - as the transformer would have to be pretty large to handle any
reasonable currents.

Correct

Secondly, the use of a common-collector stage, which has less than unity
voltage gain, but of which can provide the current requirements for a low
impedence speaker. Is this right?

A possibility.

Assuming the second option is the best ... how would this work in reality?
For example - If the first stage of the amplifier produces a signal 20
volt peak-to-peak, this would by far exceed the 5v maximum base voltage
of the second stage

Sorry, you lost me here. Do you mean the base/emitter voltage?

.... even using a PNP and NPN in push-pull

configuration would only give a possible input of 10 volt peak-to-peak ...

Should I somehow forget about the common-collector current stage ... and
somehow modify the first stage to delivery the necessary current? I can
easily develop a large current with the common-emitter design,

Are you sure? What about the dissapation of the transistor?

but this
current flows through the collector load resistor, collector, and emitter.
How do I use this current to drive a loudspeaker?

You couple via a capacitor as before. however you must remember that as
this is a class A circuit the efficiency will never be more than 25% so a
lot of power will need to be dissipated in the transistor and its emitter
resistor.

I hope someone can understand the above - and point out where I'm going
wrong. I obviously lack understanding somewhere ... but I'm not sure
where.

One problem with a simple common collector (or emitter follower as it more
commonly known) is that its output impedance is asymatric for large signals
because current is sourced by the transistor and sunk by the emitter
resistor. A better design uses a current source in place of the emitter
resistor.

At the moment I'm only interested in standard BJT's, not FET's or similar.
I want to understand basic transistors before moving on to others ...

For Ref - the rough circuit (view in fixed width font):-

+ ----------------------------------------- +
| |
RES RES
| |----CAP---O output
input 0---CAP--------TRAN
|
RES
|
GND----------------------------------------GND

Any help would be very much appreciated!

You would probably benefit from reading Douglas Self's excellent series of
articles on power amplifier design that he wrote for Wireless World. He
has a web site at:

http://www.dself.dsl.pipex.com/

HTH

Ian

 

[John Popelish]

Steve wrote:


Hi All,

I've designed a basic class-A common-emitter audio amp using a
single transistor. This is biased correctly, and develops the
voltage over an emitter resistor - the output of which can be seen
clearly on a scope.

Now, whilst it's acting as an amplifier - as soon as I connect a
speaker across the output (DC blocked with a capacitor) and ground,
evrything fails miserably. I'm not sure if this is mainly due to
the impedence of the speaker changing the operation of the circuit -
or if it's simply an impedence mis-match (I guess the speaker is
essentially 'shorting' the output to ground via a very low
resistance).

I've seen two designs which seem to resolve this problem:-

Firstly, the use of an audio transformer in place of the collector
resistor, with secondary wirings connected to a speaker. I don't
like this - as the transformer would have to be pretty large to
handle any reasonable currents.

Secondly, the use of a common-collector stage, which has less than
unity voltage gain, but of which can provide the current
requirements for a low impedence speaker. Is this right?

Assuming the second option is the best ... how would this work in
reality? For example - If the first stage of the amplifier
produces a signal 20 volt peak-to-peak, this would by far exceed
the 5v maximum base voltage of the second stage ... even using a PNP
and NPN in push-pull configuration would only give a possible input
of 10 volt peak-to-peak ...

Should I somehow forget about the common-collector current stage ...
and somehow modify the first stage to delivery the necessary
current? I can easily develop a large current with the
common-emitter design, but this current flows through the collector
load resistor, collector, and emitter. How do I use this current
to drive a loudspeaker?

I hope someone can understand the above - and point out where I'm
going wrong. I obviously lack understanding somewhere ... but I'm
not sure where.

At the moment I'm only interested in standard BJT's, not FET's or
similar. I want to understand basic transistors before moving on to
others ...

For Ref - the rough circuit (view in fixed width font):-

+ ----------------------------------------- +
| |
RES RES
| |----CAP---O output
input 0---CAP--------TRAN
|
RES
|
GND----------------------------------------GND

Any help would be very much appreciated!

Thanks,


Steve
Bristol, UK

A common solution to your problem is a complementary emitter follower
(common collector). The NPN has its collector connected to the
positive supply, and the PNP has its collector connected to the more
negative supply rail. Their emitters are connected together and to
the coupling capacitor to the speaker. Often there are a pair of low
value resistors (10 ohms or less) between the emitters, with the
coupling capacitor connected to the common point between the
resistors.

The trick is to apply the same AC signal to both bases, while biasing
them to slightly different voltages to make both transistors conduct
just a little when there is no signal. The simplest way is to connect
two diodes in series between the bases, with the PNP base connected to
the collector of the previous common emitter amplifier, and the
collector resistor for that stage connected to the NPN base, co the
collector current of the voltage gain stage passes through the two
diodes, producing enough voltage drop to put both of the common
collector transistors into slight conduction. You can increase the
voltage swing of the output by splitting the collector resistor in
half (two equal resistors in series, and connecting a coupling
capacitor from the output back to that node. No need to worry about
base to emitter voltage, because the emitter voltage follows behind
the base voltage, never getting much more than a diode drop away from
it.

--
John Popelish

 

[Kevin Aylward]

Ian Bell wrote:

Steve wrote:


Hi All,

I've designed a basic class-A common-emitter audio amp using a single
transistor. This is biased correctly, and develops the voltage
over an emitter resistor - the output of which can be seen clearly
on a scope.

A common emitter amplifier normaly develops its output between its
collector and ground. However its output impedance is relativeley
high (of the the order of several thousand ohms) and therefore not
suitable for directly driving a loudspeaker.

Although the output impedance of the common emitter amplifier is high,
this has nothing whatsoever to do with the ability to use such a
configuration in driving a low impedance speaker. What matters is the
ratio of maximum voltage and maximum current that the transistor is
capable of. There are many commercial designs that use a push pull
collector output. Feedback is usually used to reduce the output
impedance to a low value.


Kevin Aylward
salesEXTRACT@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.

http://www.anasoft.co.uk/replicators/index.html

 

[Steve]

Hi,

Just a quick note to thank all those that responded. I now have a few more
ideas to work with ... and am very greatful for all your suggestions


Regards,


Steve
Bristol, UK

 

[Bill Bowden]

"Steve" <steve-news@moonfoundation.co.uk> wrote in message news:<newscache$bz9bqh$69e$1@catflap.electro-clash.com>...

Assuming the second option is the best ... how would this work in
reality? For example - If the first stage of the amplifier produces a
signal 20 volt peak-to-peak, this would by far exceed the 5v maximum
base voltage of the second stage ... even using a PNP and NPN in
push-pull configuration would only give a possible input of 10 volt
peak-to-peak ...

A common collector configuration is sometimes called
a "emitter follower" where the emitter voltage follows
the base voltage. So, you don't have to worry about exceeding
the base to emitter voltage because they are almost the same.

There is an example push-pull emitter follower circuit
using 3 transistors here:

http://ourworld.compuserve.com/homepages/Bill_Bowden/page8.htm#amp.gif

It only delivers about 50 milliwatts from a 9 volt battery, but
you can run the supply at 12 volts for more power.

-Bill

 

[ddwyer]

In article <ad025737.0312231351.5ec3f590@posting.google.com>, Bill
Bowden <wrongaddress@att.net> writes

"Steve" <steve-news@moonfoundation.co.uk> wrote in message news:<newscache$bz9bq
h$69e$1@catflap.electro-clash.com>...

Assuming the second option is the best ... how would this work in
reality? For example - If the first stage of the amplifier produces a
signal 20 volt peak-to-peak, this would by far exceed the 5v maximum
base voltage of the second stage ... even using a PNP and NPN in
push-pull configuration would only give a possible input of 10 volt
peak-to-peak ...


A common collector configuration is sometimes called
a "emitter follower" where the emitter voltage follows
the base voltage. So, you don't have to worry about exceeding
the base to emitter voltage because they are almost the same.

There is an example push-pull emitter follower circuit
using 3 transistors here:

http://ourworld.compuserve.com/homepages/Bill_Bowden/page8.htm#amp.gif

It only delivers about 50 milliwatts from a 9 volt battery, but
you can run the supply at 12 volts for more power.

-Bill
surely a common collector

--
ddwyer

 

[Ian Bell]

ddwyer wrote:

In article <ad025737.0312231351.5ec3f590@posting.google.com>, Bill
Bowden <wrongaddress@att.net> writes
"Steve" <steve-news@moonfoundation.co.uk> wrote in message
news:<newscache$bz9bq h$69e$1@catflap.electro-clash.com>...

Assuming the second option is the best ... how would this work in
reality? For example - If the first stage of the amplifier produces a
signal 20 volt peak-to-peak, this would by far exceed the 5v maximum
base voltage of the second stage ... even using a PNP and NPN in
push-pull configuration would only give a possible input of 10 volt
peak-to-peak ...


A common collector configuration is sometimes called
a "emitter follower" where the emitter voltage follows
the base voltage. So, you don't have to worry about exceeding
the base to emitter voltage because they are almost the same.

There is an example push-pull emitter follower circuit
using 3 transistors here:

http://ourworld.compuserve.com/homepages/Bill_Bowden/page8.htm#amp.gif

It only delivers about 50 milliwatts from a 9 volt battery, but
you can run the supply at 12 volts for more power.

-Bill
surely a common collector

I am sure you are aware that common collector and emitter follower are the
same thing.

Ian

--
UKFS

 

[N. Thornton]

"Steve" <steve-news@moonfoundation.co.uk> wrote in message news:<newscache$bz9bqh$69e$1@catflap.electro-clash.com>...

Hi All,

I've designed a basic class-A common-emitter audio amp using a single
transistor. This is biased correctly, and develops the voltage over an
emitter resistor - the output of which can be seen clearly on a scope.

Now, whilst it's acting as an amplifier - as soon as I connect a speaker
across the output (DC blocked with a capacitor) and ground, evrything
fails miserably. I'm not sure if this is mainly due to the impedence
of the speaker changing the operation of the circuit - or if it's simply
an impedence mis-match (I guess the speaker is essentially 'shorting'
the output to ground via a very low resistance).

I've seen two designs which seem to resolve this problem:-

Firstly, the use of an audio transformer in place of the collector
resistor, with secondary wirings connected to a speaker. I don't like
this - as the transformer would have to be pretty large to handle any
reasonable currents.

Secondly, the use of a common-collector stage, which has less than unity
voltage gain, but of which can provide the current requirements for a
low impedence speaker. Is this right?

Assuming the second option is the best ... how would this work in
reality? For example - If the first stage of the amplifier produces a
signal 20 volt peak-to-peak, this would by far exceed the 5v maximum
base voltage of the second stage ... even using a PNP and NPN in
push-pull configuration would only give a possible input of 10 volt
peak-to-peak ...

Should I somehow forget about the common-collector current stage ... and
somehow modify the first stage to delivery the necessary current? I
can easily develop a large current with the common-emitter design, but
this current flows through the collector load resistor, collector, and
emitter. How do I use this current to drive a loudspeaker?

I hope someone can understand the above - and point out where I'm going
wrong. I obviously lack understanding somewhere ... but I'm not sure
where.

At the moment I'm only interested in standard BJT's, not FET's or
similar. I want to understand basic transistors before moving on to
others ...

For Ref - the rough circuit (view in fixed width font):-

+ ----------------------------------------- +
| |
RES RES
| |----CAP---O output
input 0---CAP--------TRAN
|
RES
|
GND----------------------------------------GND

Hi Steve. Lets get your circuit working. Firstly remove the collector
resistor and put the speaker where it was. You wont need the
capacitor. Before someone yells, yse I know it will put sc thru the
speaker, but for a basic starter it'll work. Later we can replace that
with L-C-LS.

Now, add a resistor from base to ground so the transistor's bias point
is controlled sensibly. Set it up so you have say half a volt sitting
on the emitter resistor. Thus the base needs to be at 1.1v.

That'll get you up and running. Some more info on what power out you
want, supply voltage, tr type etc would help.


Regards, NT