I am getting noisy audio signal in my headphone. I don't know where the problem lies. To explain the situation, the stages are MCU -> DAC(PCM1789) -> LPF/diff. to single circuit(TPA6138) -> audio amplifier(TPA6138). The LPF and amplifier are both made up of TPA6138. I don't think there is problem upto DAC output.

I followed the datasheet of TPA6138 both for LPF/differential to single circuit and for audio amplifier. Attached herewith is circuit schematic. Attached is also output signal at the headphone. The signal is above 0V, is this correct? shouldn't there be negative going signal?

I have a 100kohm and 0.5uF between LPF and amplifier. The 100kohm is for reducing loudness and the 0.5uF is for DC blocking capacitor for 20Hz to 20kHz audio signal.

Trying for days now, would really appreciate if somebody can look into this.

Thanks

 

I am getting noisy audio signal in my headphone. I don't know where the problem lies. To explain the situation, the stages are MCU -> DAC(PCM1789) -> LPF/diff. to single circuit(TPA6138) -> audio amplifier(TPA6138). The LPF and amplifier are both made up of TPA6138. I don't think there is problem upto DAC output.

I followed the datasheet of TPA6138 both for LPF/differential to single circuit and for audio amplifier. Attached herewith is circuit schematic. Attached is also output signal at the headphone. The signal is above 0V, is this correct? shouldn't there be negative going signal?

I have a 100kohm and 0.5uF between LPF and amplifier. The 100kohm is for reducing loudness and the 0.5uF is for DC blocking capacitor for 20Hz to 20kHz audio signal.

Trying for days now, would really appreciate if somebody can look into this.

Thanks

Let me say up front that I had no prior knowledge of this chip or others like it, so feel free to take my thoughts with a grain of salt.

Having said that, I've looked through the datasheet and it looks to me like the chip can be set up as LPF and amplifier at the same time. I don't think you need two stages. I'd try removing the second stage, up to and including the 100k resistors, and see what kind of output you get from U2.

As for the negative signal, I'd agree that according to the datasheet, you should have an amplified audio signal centered around ground. If your output is biased around some positive voltage, that would seem to indicate a problem.

 

i am not sure what the datasheet means, is it both LPF and audio amplifier or are they separate application. But i did try both of them before. It did work assuming it is both LPF and amplifier.

 

i am not sure what the datasheet means, is it both LPF and audio amplifier or are they separate application. But i did try both of them before. It did work assuming it is both LPF and amplifier.

I'm not sure either, and the wording in the datasheet is pretty vague. The circuit they show as a second order LPF looks exactly like the input stage on their "typical application" circuit example. That's what made me think it was potentially operating as both at once.

You say it did work, assuming it's both? If so, why did you add the second stage? Were you having noise, distortion, DC bias, or any other problems when you ran the single stage version? If not, I'd go back to that.

 

at first i was confident that LPF and amplifier are two different application of the TPA6138. So I build both of them and tested. Later I just connected directly to the amplifier circuit with a capacitors at the input of the amp.

It works when both LPF and amp are used. It also works when only amp circuit is used.

I am getting clearer sound only when the volume in the PC is made low(case when only amp circuit is used with 2.2uF at the inputs, but other values also don't seems to make any difference). When I increase volume in PC there is high noise, high volume off course and audio is buried in it(noise).

Decreasing the PC volume the noise is lowered and I can hear and understand the audio(also music). Although the audio is clearer it is somewhat changed- like a man voice sounds like women sound but there is no background noise.

I have put various capacitors ranges at the input of amp(+LPF assuming this) circuit.

Attached is another schematic I found and I presently done this with the inputs and outputs(res and cap).

 

Again, let me start by saying I'm in over my head here, as I'm still learning most of this stuff too...

I've looked over the PCM1789 datasheet a little bit and a few things jump out at me:
1) Recommended output voltage range is 8V p-p, which appears to be higher than the input range of the TPA6138. Perhaps that's part of the issue with not being able to turn it up all the way? You could add some attenuation between the TPA and the PCM in order to utilize full ADC resolution while staying within PCM's expected input range.
2) Recommended load capacitance on analog output is maximum of 50pF. Maybe I'm misunderstanding this spec (I wouldn't be surprised) but if that applies to C8, 9, 12, and 13 like I think it does, then you're off by many orders of magnitude!

Hopefully someone smarter than me (@Alec_t or @AnalogKid among others) can help make sense of this?

 

i am not sure what the datasheet means, is it both LPF and audio amplifier or are they separate application.

The chip is two uncommitted opamps designed for headphone driving, plus a charge pump power supply to eliminate the need for an external negative voltage source or two output coupling capacitors. The chip is tweaked for total circuit cost where a few pennies adds up to real money over hundreds of thousands or millions of units.

The applications sections shows how to incorporate an LPF into the headphone amp function because that is such a common requirement. There is nothing special about the amps regarding the LPF function; their filter circuit will work with many other opamps.

ak

 

i put attenuation circuit between dac and supposedly LPF/amp such that input into LPF/amp is within 0 to 3.3V. This did not help but certainly attenuation had to be incorporated.

any other suggestions?

 

i put attenuation circuit between dac and supposedly LPF/amp such that input into LPF/amp is within 0 to 3.3V. This did not help but certainly attenuation had to be incorporated.

any other suggestions?

Did you look at the load capacitance specs on the DAC? I might be misunderstanding what those specs mean, but if I'm right, your capacitors between the DAC and the amp are way too big.

 

Did you look at the load capacitance specs on the DAC? I might be misunderstanding what those specs mean, but if I'm right, your capacitors between the DAC and the amp are way too big.

hi,

but the amplifier datasheet says to put blocking capacitor of 1uF. Latest result is that I have lowered the volume and adjusted the signal amplitude range between DAC and amplifier. The volume is ok now but there is one problem now- background noise present even in absence of any audio. would you check here too- https://forum.allaboutcircuits.com/threads/background-noise-in-circuit-in-absence-of-audio.139170

 

Did you look at the load capacitance specs on the DAC? I might be misunderstanding what those specs mean, but if I'm right, your capacitors between the DAC and the amp are way too big.

You are misunderstanding. The larger a series coupling capacitor is, the more it resembles a piece of wire. Whatever is on the other side of the capacitor that creates a path to GND is what determines the load current and phase shift. A max external capacitance (directly to GND) spec usually indicates that a capacitive load greater than that can cause the output amplifier to break into oscillation. The capacitor forms a phase shift network with the amplifier's output impedance.

ak

 

You are misunderstanding. The larger a series coupling capacitor is, the more it resembles a piece of wire. Whatever is on the other side of the capacitor that creates a path to GND is what determines the load current and phase shift. A max external capacitance (directly to GND) spec usually indicates that a capacitive load greater than that can cause the output amplifier to break into oscillation. The capacitor forms a phase shift network with the amplifier's output impedance.

ak

Excellent! Thanks for the explanation.

I realize now that I often get turned around thinking about capacitance questions because there can be unintended capacitance between so many things, my head starts spinning.

Sorry for distracting this thread from whatever the real issues are, and thanks for clearing things up!