# Bjt amplififer

Status
Not open for further replies.

#### merk24

##### New Member
Good day! I have a problem on our project (bjt amplifer) especially with the Watts. We already get the Voltage output gain but we cannot get the 3 Watts that we needed. I already tried changing the value of the resistor along the output node but the voltage output changes . May I ask what should I do to make it 3 Watts and what components should I change. Thank you

Specification of our project :
Vo GAIN of 8
3W
Vcc -> 12v
max input 1Vrms
Without clipping

#### Attachments

• 1.9 KB Views: 103

#### ericgibbs

##### Well-Known Member
hi,
Consider you have a 12V supply and a 30R load, how is it possible to get 3 Watts in the load.???
E

#### merk24

##### New Member
I don't know if it is possible to get 3 watts. But my professor suggested that we should try to change our transistor to get 3 watts. I don't have any idea how to come up for the 3 watts output.

#### ericgibbs

##### Well-Known Member
Hi,
Consider reducing the 30R to say 5R or 3R , increase the power rating of the two output transistors and repeat your Sim.
Post what you get.
E

#### Les Jones

##### Well-Known Member
If you want 3 watts into a 30 ohm load then you need the square root of 30 x 3 which is 9.49 volts RMS. The peak to peak value of 9.49 volts RMS is 9.49 x Root 2 x 2 = 9.49 x 1.414 x2 = 26.84 volts peak to peak. (I am assuming that you have a single transistor amplifier as a can't read a .asc file.) So to get this output you would need a power supply of at least 29 volts to ensure the transistor does not saturate. I can only see two ways to achieve this from a 12 volt supply. 1 Use an output transformer to step up the output voltage. 2 Build into your circuit a 12 to 30 volt step up converter to power the transistor.

Les.

Last edited:

#### MikeMl

##### Well-Known Member
I don't know if it is possible to get 3 watts. But my professor suggested that we should try to change our transistor to get 3 watts. I don't have any idea how to come up for the 3 watts output.
Guys, this is homework. Don't just hand him the answer...

merk24, Forget LTSpice for a moment. Ask yourself the following:

1. Can you make a amplifier that has a 8Vpeak-to-peak output swing (gain of 8) when powered from 12Vdc?

2. For what waveform shape is the power calculation usually done?

3. Knowing the peak-to-peak voltage across a resistive load, how do you calculate what the load resistance would have to be in order to produce a certain power?

#### Les Jones

##### Well-Known Member
Could it be that in the first post that "Vcc -> 12v" is intended to mean grater than 12 volts ? If so there woulfd not be a problem.

Les.

#### ronsimpson

##### Well-Known Member
I added "boot strap" and got the output to 11.5V on the pull up side and 1.5V on the pull down side. So 10Vp-p.
Yes Les, a 15 or 20V supply works well.
------
I would not build this type of amplifier in production. It is too dependent on the current gain of Q1. If you change the 2N3904 to a part make one second later in time, the amp will bias differently. The temperature will change the bias point too much.

#### Nigel Goodwin

##### Super Moderator
I would not build this type of amplifier in production.
As he's not posted the schematic in a sensible format I've no idea what it might be like?, but 4W RMS in to 4 ohms from 12V is easily (and commonly) doable, as done in all single ended car radio amplifiers.

#### ronsimpson

##### Well-Known Member
As he's not posted the schematic in a sensible format
He did, so most of us can run the simulation. Better than posting in PDF and we can not run the sim.

#### MikeMl

##### Well-Known Member
I say again:
Guys, this is homework. Don't just hand him the answer...

#### audioguru

##### Well-Known Member
The simple amplifier is class-AB. If the output power is 3W then the total heating of the output transistors is about 2.5W then each output transistor heats with 1.25W. But their datasheets lists a maximum allowed heating of only 0.625W so they might melt. Power transistors with metal tabs for a heatsink should be used instead.

#### JoeJester

##### Active Member
Posting the schematic is nice for those who do not use LTSpice. The same applies if someone posted a circuit in any of the other simulation software formats.

Posting it in the format easily read by LTSpice users is only easier for them.

I am shocked to see the results of my inquiry "spice simulation rankings" at http://www.eetimes.com/document.asp?doc_id=1326778

The second ranking page I viewed did have LTSpice, but not as the leader. https://mixedsignal.wordpress.com/2011/02/04/top-ten-circuit-simulators/

I do like the answers on StackExchange, especially when the one response compared a "book" oscillator circuit with a "real world" example.

#### EvilGenius

##### Member
I believe a resistor from collector of 2N4403 to ground will solve the problem. The output is sunk to ground during amplification. Once you add this resistor, you can play with output resistor to get your desired voltage to achieve your wattage.

#### spec

##### Well-Known Member
I believe a resistor from collector of 2N4403 to ground will solve the problem. The output is sunk to ground during amplification. Once you add this resistor, you can play with output resistor to get your desired voltage to achieve your wattage.
Did you mean base of Q1 (2N3904) to ground?

spec

#### EvilGenius

##### Member
Did you mean base of Q1 (2N3904) to ground?

spec
No. I am looking at it from logic perspective and its conflicting. It looks like the circuit is inverted logic which I am ok with it. I.e: sine in cosine out (90 deg phase shifted). The problem I see is that logic 1 in turns 3904 on, turning 4403 on grounding the output while 4401 is still on, shorting Vcc to ground.

#### ronsimpson

##### Well-Known Member
turning 4403 on grounding the output while 4401 is still on, shorting Vcc to ground.
D1, D2 keep the two output transistors on just a little bit. Some current flows from +12V to ground but not full current. (not a short)

#### EvilGenius

##### Member
D1, D2 keep the two output transistors on just a little bit. Some current flows from +12V to ground but not full current. (not a short)
Interesting. I am guessing the double diodes are to get rid of zero crossing distortion?! Looks like a push pull phase split.

#### Tony Stewart

##### Well-Known Member
He did, so most of us can run the simulation. Better than posting in PDF and we can not run the sim.
View attachment 101083
This design is less than optimal in many respects besides thermal due lack of open loop gain and insufficient input bias when output swings low. This results in nonlinearity and unable to maximize swing.

#### EvilGenius

##### Member
I had a chance to study this a bit. To improve on the output wattage, you need beefier NPN and PNP like TIP family. If output voltage is 4v in each directions, to obtain 3w the current needs to be around 750ma which 2N4401 and 2N4403 cannot supply.
The example I looked at had resistors on base and emitter of 2N3904 (change that to 2N4401 for higher gain). Also the 100k resistor was used to connect Vcc to base of 2N3904 for proper bias.
As a scholar in this post mentioned earlier simulation indicated a non symmetric output waveform which was disportionate to input.

Status
Not open for further replies.