Increasing the power of the audio amplifier TPA3128D2

Thread Starter

electro_fan

Joined Feb 6, 2018
9
Hello,
I am currently working on the modernization of a certain construction.
It is a Class D audio amplifier system built into a certain device. The system power is ~15VDC. The schematics is based on the TI TPA3128D2 datasheet. I would like to increase its power a bit.
In the present solution, the analog input signal is about 300mV (pk-pk). Because I can not raise the supply voltage, I thought about increasing the input signal. Is it good idea? I thought about a simple solution consisting in inserting between the signal source and the TPA3128D2 amplifier "preamplifier" based on two LM358 operational amplifiers (for left and right channel). Is this a good idea and does it make any sense?

crutschow

Joined Mar 14, 2008
25,432
Depends upon how close to the maximum power output, the 300mV input gives you.
The TPA3128D2 is rated at 15W into an 8 ohm load with a 15V supply, giving a maximum output voltage of √(15*8) = 11Vrms or 30Vpp.
Measure the output voltage across the speaker and that will tell you whether additional amplification of the input voltage will give you more output sound without distortion.
But if you do add amplification, don't use the LM358, as they are noisy and can have crossover distortion.
Use an audio op amp, such as the NE5532.

Jony130

Joined Feb 17, 2009
5,183
But how can increase in the input signal incresst he amplifier power? Only increase in the supply voltage can for real increase the output power.

KeithWalker

Joined Jul 10, 2017
1,222
But how can increase in the input signal incresst he amplifier power? Only increase in the supply voltage can for real increase the output power.
The TPA3128D2 needs about 1Vp/p on the input to get the maximum power out of 30Vp/p output, so it will need a pre-amp stage with a gain of x3 to achieve this with an input of 300mVp/p.

crutschow

Joined Mar 14, 2008
25,432
But how can increase in the input signal incresst he amplifier power? Only increase in the supply voltage can for real increase the output power.
That's true if the 300mV gives the maximum output power, but it may not.

Thread Starter

electro_fan

Joined Feb 6, 2018
9
Thanks guys for all replies.
I would also like to say that there is no symmetrical voltage in the system, and the measurements of the input signal are made in reference to GND and the output to the + and - the speaker.

Today I measured the output voltage by connecting an 8ohm resistor at the output. I used a sine as a test signal. Here's what I observed:
Input signal sine 1kHz, has about 300mV, the output signal is about 12V. This gain is obtained with a gain factor of 36dB (the largest possible). It seems to me that increasing the input signal from 300mV to about 400mV should give me a TPA3128D2 output signal with an amplitude around 15V.
Will a regular non-converting operational amplifier handle the matter?

crutschow

Joined Mar 14, 2008
25,432
Input signal sine 1kHz, has about 300mV, the output signal is about 12V.
Is that 300mVpp or 300mVrms.
Will a regular non-converting operational amplifier handle the matter?
If you mean non-inverting, yes.
But you could also use an inverting amp, as the phase of the input signal doesn't matter (unless you have stereo inputs).

Thread Starter

electro_fan

Joined Feb 6, 2018
9
Thanks for help. Problem description:
The output signal from the VS1003 IC is about 250mVpp. After amplification by TPA3128D2, the output signal (1kHz sine) on the 8Ohm speakers has about 12Vpp.

I made a small experiment and as an input signal I connected the 1kHz 350mVpp signal generator. Then the TPA3128D2 amplifier strengthens it to 15Vpp and is clearly louder than when the input signal has about 250mVpp.
I would like to achieve the same effect using the signal from VS1003. Is the diagram below right?

I tried to connect it to the universal board, but it did not work.

Attachments

• 84.2 KB Views: 12

KeithWalker

Joined Jul 10, 2017
1,222
Thanks for help. Problem description:
The output signal from the VS1003 IC is about 250mVpp. After amplification by TPA3128D2, the output signal (1kHz sine) on the 8Ohm speakers has about 12Vpp.

I made a small experiment and as an input signal I connected the 1kHz 350mVpp signal generator. Then the TPA3128D2 amplifier strengthens it to 15Vpp and is clearly louder than when the input signal has about 250mVpp.
I would like to achieve the same effect using the signal from VS1003. Is the diagram below right?

I tried to connect it to the universal board, but it did not work.
That looks OK. Double check your connections and measure the output from the IC. You have a bit too much gain ( x 2 ) and that will cause the amplifier to clip the peaks. You will have to decrease the value of R1 and R3 to 39 K to give you a gain of 1.4.

Thread Starter

electro_fan

Joined Feb 6, 2018
9
Are capacitors on input and output mandatory? What about their values? I am testing it with LM358.

KeithWalker

Joined Jul 10, 2017
1,222
Are capacitors on input and output mandatory? What about their values? I am testing it with LM358.
You have a good point. You do need them but it drew my attention to the fact that you have no reference level for the inputs on the IC. Try this:

sghioto

Joined Dec 31, 2017
1,463
Try this:
That circuit has negative gain.
Using a LM358 the circuit below should work. Gives a gain of 1.39
SG

Last edited:

crutschow

Joined Mar 14, 2008
25,432
I tried to connect it to the universal board, but it did not work.
The circuit didn't work because the (+) inputs need a resistive path to ground for the op amp input bias currents, which are blocked by the input capacitors.

Audioguru

Joined Dec 20, 2007
11,249
I guess you do not care about distortion:
1) The datasheet of the TPA3128D2 shows a graph of output power vs supply voltage. With a 15V supply it is clipping pretty badly into 8 ohms producing about 12.5W. If you turn up the volume so it clips more then the distortion harmonics are louder, not the original signal.
2) The lousy old LM358 produces crossover distortion because its output is biased for very low supply current.

sghioto

Joined Dec 31, 2017
1,463
The lousy old LM358 produces crossover distortion because its output is biased for very low supply current.
The LM358 isn't that bad. It's design states for Active Filters, General Signal Conditioning and Amplification
I didn't see any crossover distortion in the circuit I posted and with a gain of less then two, what noise?
The TS stated in post #8 that he only needed 15Vpp output to achieve the required volume level.
That's only about 3.5 watts rms
SG

Audioguru

Joined Dec 20, 2007
11,249
The TPA3128D2 stereo amplifier has bridged outputs so each output produces almost 30Vp-p when the supply is 15VDC.
The LM324 is a quad opamp and the LM358 is a dual using the same opamps. Here is how bad is the crossover distortion:

Attachments

• 52.3 KB Views: 6

sghioto

Joined Dec 31, 2017
1,463
Similar threads