How can I make Audio Amplifier from 2SK3549-01 MOSFET

[Afry]

Hi all I have small project using Microcontroller to output PCM Audio Samples . Everything works fine and the sound is clear , but it's very low I barley hear it , I lithely need to put the speaker inside my ear to hear it output (the Microcontroller runs at 3.3V ).

I use simple circuit using 2SK3549-01 N-CHANNEL MOSFET and S9018 NPN transistor to amplify audio signal but I got no sound output .

I already tried using NPN transistor alone the sound level improved little bit but it's still low , BTW I don't care about noise level All I want first is the voice become loader so I can hear it from meter away .

 

[duke37]

What is your normal language? Barley only rustles when the sun is on it.
What is I1?
What are the specifications for the FET?
If I1 is low, then there will not be enough voltage to turn on yhe active devices.
As drawn, there will be 0.6V less then I1 on the fet gate which may not be enough.
Connect I1 direct to the fet gate without the transistor. A power fet normaly needs a few volts on the gate to turn it on.
Pushing V1 up to 5V or so would help.

 

[Alec_t]

That particular mosfet is not a logic-level type and may need at least 5V gate drive to work satisfactorily.

 

[Afry]

What is your normal language? Barley only rustles when the sun is on it.
What is I1?
What are the specifications for the FET?
.

Hi thanks for reply obviously English is not my first language so sorry if any grammars mistakes I have on OP .
I1 is signal source from my Microcontroller it max voltage output is 3V here is datasheet for the MOSFET https://www.promelec.ru/datasheet/c/f/2SK3549.pdf

If I1 is low, then there will not be enough voltage to turn on yhe active devices.
As drawn, there will be 0.6V less then I1 on the fet gate which may not be enough..

I already connect the transistor gate to my signal source and it work but the sound were too low , as I said my signal output is between 3v - 0v which is PWM signal representing of PCM wave sampling from sound clip .

Connect I1 direct to the fet gate without the transistor. A power fet normaly needs a few volts on the gate to turn it on. Pushing V1 up to 5V or so would help.

I did that too and I heard no sound honest I dont know how connect this MOSFET to the circuit since it's N-Channel based how could I cut the circuit from it without NPN transistor
BTW L1 is 5-8 Ohm Speaker .

That particular mosfet is not a logic-level type and may need at least 5V gate drive to work satisfactorily.

it datasheet said N-CHANNEL SILICON POWER MOSFET , anyway could you recommend me better MOSFET or even a transistor that could output bit lout sound from 5V power source even if it suitable for earphone .

 

[Alec_t]

could you recommend me better MOSFET or even a transistor that could output bit lout sound from 5V power source even if it suitable for earphone .

Rather than a single transistor (mosfet or bjt), have you considered an LM386-based amplifier (such as one of these) available cheaply ready built?

 

[Afry]

Rather than a single transistor (mosfet or bjt), have you considered an LM386-based amplifier (such as one of these) available cheaply ready built?

We don't have such component esay avaliable on my country but I can order anyone from online ,,,
I forgot to mention I want the Amplifier works with very few components , since I have two complex Amplifiers on my hand CD2025HCP and CD7316CP but I couldn't make them to work (I guess all their pins need to connected to the capacitors I count 8 )

That's why I ask for simple circuit even with single MOSFET as i'm not transferring music with the device I design only human voice .
here is circuit on the OP I made it diagram with LTspice IV
http://sendanywhe.re/O7JTCNKP

 

[Nanren888]

Have a look at the configuration
https://forum.arduino.cc/index.php?topic=451897.0

I've not looked at the value and not for your components, but the configuration might be of interest.
Might want to add a resistor in the base of the transistor, to limit base current, depending on your transistor ans processor..
.
You have the transistor connected in a configuration called emitter follower, that provides current gain but not voltage gain. For small voltage input it provides voltage loss because of the diode drop base to emitter, as mentioned by duke37.
The link I have given uses a transistor in common emitter configuration, so the voltage swing at the collector will be larger than the base voltage.
The supply voltage for the transtistor and the fet could be greater than the processor voltage. I have not looked at your fet data, but as mentioned above, you need to get enough voltage on the gate to switch it.

 

[duke37]

You could try placing a high value resistance from the 3.3V supply to the fet base. Then drive the base through a capacitor so that the signal is centred about 3.3V. Make sure that excessive current is not passed or something my fail, the speaker is what controls the maximum current. A resistor in the fet source would also do.
The threshhold voltage could be as high as 5V so even this may not work and you may have to go to a logic level fet.

 

[Afry]

Alright guys firs thank you all for your help the info you give me sure will help me a lot
Anyway I discover the reason why the sound were too weak , my uC output voltage were too low ( 0.4V max ) because mistakes I made on PWM wave signal generator , once I fix the code (max output voltage went up to 3V ) so I hookup the signal output to Transitory base an connect the other end of the speaker (minus) to the collector and plus spear to 5V voltage source and the sound become way louder .

 

[Alec_t]

BC856B is a pnp transistor, so won't work well as shown.
A base resistor to limit current should be added.

 

[Nanren888]

Great, so you went for a common emitter configuration as suggested.
Found you could skip the FET you asked to use. ok.
.
That ground-like symbol at the bottom, on the emitter of Q1, is to the supply positive, right?
And the other one at the bottom of V1 is ground, zero volts, right?
Alec_t has also suggested a base current limit resistor.

 

[Afry]

BC856B is a pnp transistor, so won't work well as shown

Great, so you went for a common emitter configuration as suggested.
Found you could skip the FET you asked to use. ok.
That ground-like symbol at the bottom, on the emitter of Q1, is to the supply positive, right?
And the other one at the bottom of V1 is ground, zero volts, right?
Alec_t has also suggested a base current limit resistor.

Honesty I don't know what common emitter configuration means , all what got me to electronics is Youtube video I watch people experiencing some circuits that do stuff like audio amplifiers etc ,
The MOSFET still confuses me until now , but honesty this experience made me able to read datasheet and able to understand it better.
Anyway here is complete picture of the board I already but resistor on the base (but I accidentally remove it before take the picture )

 

[davenn]

Anyway here is complete picture of the board I already but resistor on the base

that black "link" near the top centre isn't a resistor

 

[Afry]

that black "link" near the top centre isn't a resistor

that a line I draw with Pinta I used my phone to take the picture and it's camera not that good It give me 1 acceptable picture out of three .

 

[Nanren888]

Thanks for the image.

>>Honesty I don't know what common emitter configuration means ,
Not to worry. Google it if you want. You'll get here.
https://en.wikipedia.org/wiki/Common_collector = emitter follower
https://en.wikipedia.org/wiki/Common_emitter
Recognising these two will speed you along. But entirely optional for your currrent aims.


Image shows quite well positive and ground connected. Have you gone back to the NPN transistors?
If you have the emitters to ground, the collectors to the load then the load to positive, then seems an NPN is where you should be.

This page;
https://itp.nyu.edu/physcomp/labs/m...o-control-high-current-loads-with-an-arduino/
about figures 17, 19 shows connecting a transistor to boost the current available form an arduino output pin.Might be helpful. Includes that base resistor that was mentioned.

Looks as if you have some transistors in parallel, all connected to the same things. Can be done if you don't have a bigger transistor.
Transistors don't always share well.when connected like this, as they are not actually identical, so sometimes you want to connect them all to the load but give each one a separate base resistor, and best also a small emitter resistor.
Here's a pagethat talks about it, but the images tell most of the story.
https://electronics.stackexchange.com/questions/77045/transistors-in-parallel
Note each transistor has it's own base resistor.
This all being a bit of a pain, people generally look for a bigger transistor if they have the option.
.
If you want to use fets, and they can at times be easier, google about them. On the data sheet you'll find somewhere a minimum voltage you;ll need to turn them on.

 

[Audioguru]

Why do you show three transistors in parallel?
Why didn't you write the part number of the transistors?

 

[Afry]

Alright guys I upload video demonstrate the working circuit (short 8 second clip )

BTW I'm still not satisfied with sound level (the speaker are 8 Ohm I hookup from old LCD-TV )

so sometimes you want to connect them all to the load but give each one a separate base resistor, and best also a small emitter resistor.

I'll think adding that to the feature design (e.g when I find way to put them on PCB ) ,

Why do you show three transistors in parallel?

to be able to handle the current (less heat )

Why didn't you write the part number of the transistors?

I forgot . here is the transistors I use S9014 and S9018 (obviously I can't edit OP now ) .

 

[Audioguru]

Audio is AC and a speaker works with the AC so its cone can swing back and forth. But your "amplifier" produces variable DC which might cause the speaker cone to be crammed against one side and it can swing only to the middle.

Usually your simple circuit would act like a rectifier producing severe distortion so you are lucky that the biasing causes the cone to be with an average voltage of about 2.5V most of the time. Then a simple calculation shows the output peak voltage swing is about 2V and the output power is about 0.25W which is very low sound level. A cheap clock radio produces 0.5W.

A real amplifier circuit uses two output transistors, one transistor pulls the speaker cone in one direction and the other transistor pushes the speaker cone in the other direction so that the speaker cone can swing back and forth. Power equals voltage times current so if the power supply is 30V then the speaker cone can swing about 12V peak which produces 18W into an 8 ohms speaker.

 

[Nanren888]

I guess the speaker cone's repsonse depends on drive, PWM frequency and momentum? Can't say I've done those sums.
If the speaker looks inductive, does the cone react more to voltage or current?
Perhaps a comment about how to go from PWM to something speaker-friendly?

If he has limited voltage range, you could talk about building a bridge output stage.

 

[Audioguru]

An audio amplifier produces a voltage output with an extremely low impedance (0.04 ohms or less) and a speaker is designed for it.
Then the cone 's momentum is cancelled and the cone does exactly what the signal "says".
A speaker has a low frequency resonance that would sound like a bongo drum if the amplifier has a high output impedance current signal.

The OP said the signal is PCM (audio samples), not PWM (Pulse Width Modulation). The audio system looks like a Bluetooth receiver that would produce ordinary audio.

I agree that many modern amplifiers use a bridged output stage for power multiplication.