2
\$\begingroup\$

I'm currently working on a simple audio amplifier circuit using the TDA7052 IC. So far I have been able to achieve amplification, with a little distortion.

My problem is that I'm observing a large DC offset at the audio output, and the IC is getting very hot.

I'm feeding in 12Vdc and the amp is drawing around 100mA, which makes for 1.2W of power. The gain is set to 20dB with pin #4 being driven at 1Vdc. Here is my circuit: TDA7052 based audio amp

And here are the input and output waveforms (Yellow waveform taken at pin 8, with reference to system ground):

Waveforms

So, the input is being driven with a signal generator with the parameters displayed. The output is indeed being amplified, but it has a DC average value of around 5V, how do I remove this DC offset? and is there a way to mitigate this slight distortion?

I've tried adding two 10uF caps in series with OUT+ and OUT-, this is what happened: (and I'm not surprised, the datasheet doesn't show them in the typical application circuit, nor do other implementations found on the internet). enter image description here

\$\endgroup\$
19
  • 1
    \$\begingroup\$ How did you measure the output? What did you do with (both) outputs (the circuit diagram doesn't show)? The speaker is supposed to be between OUT+ and OUT- and none of the outputs must be connected to GND. \$\endgroup\$
    – Curd
    Commented Aug 25, 2016 at 20:12
  • 2
    \$\begingroup\$ @brhans: no, generally they don't (i.e. only if the signal is 0). The specialty of the TDA7052 is that you don't have one output for a load based at GND but two complementary outputs with the load between the outputs. That way it is possible to do without the output capacitor (saves money and space). \$\endgroup\$
    – Curd
    Commented Aug 25, 2016 at 20:30
  • 2
    \$\begingroup\$ Your signal generator GND and scope GND may be shorting out one of the outputs if you are connecting the scope reference (GND) to one of the outputs. Adding a capacitor just makes it an AC short. \$\endgroup\$ Commented Aug 25, 2016 at 20:31
  • 1
    \$\begingroup\$ "The problem is that both these outputs have a 5V DC offset" Yes, they do because (a) Each output has to maximise the swing from a single supply voltage (so they need a DC offset otherwise you would only get half the cycle output and (b) Each output has the same DC offset so that when you connect your speaker between them the DC offsets cancel each other and leave no net DC voltage across the speaker. That's how a bridge output works. \$\endgroup\$ Commented Aug 26, 2016 at 12:40
  • 2
    \$\begingroup\$ @Curd the word 'offset' usually refers to the steady-state DC level - ie: the state when the signal is 0. \$\endgroup\$
    – brhans
    Commented Aug 26, 2016 at 14:02

2 Answers 2

0
\$\begingroup\$

What you have is a bridge output which is self biased to be at 1/2 the supply voltage. That is normal. The outputs will be at or close to 1/2 the power supply voltage with no signal.

If the IC is getting hot, well it needs to be mounted to a heat sink. Aluminum is fine. 1.2 watts is very hot to touch, so a heat sink is mandatory.

It sounds like you have a perfectly good amplifier that is just missing a heat sink. All of your parameters (100 mA for bias current for bridge output) seem normal and you stated it put out a clean sound, so I do not think anything is wrong-except for a missing heat sink.

\$\endgroup\$
0
\$\begingroup\$

It appears that the DC offset of the outputs are not "cancelling out" and that's why the chip is getting excessively hot.
To see the "correct" output signal, use the input as your trigger and one channel on the + out, and the other channel on the - out (across the speaker terminals).
If the chip still overheats, it could be faulty.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.