hi ff,
Is this your basic circuit?.
Post some 'wav' files of the 3 audio sources, we can then simulate the result.
[change the *.wav extension to *.txt in order allow an upload to AAC.
E

 

Ive rebuilt the circuit on strip board instead of breadboard, but the LM4871 is still very quiet yet the PAM8302A is louder but there is a hum (there is also hum on the LM4871). I tried powering from a power bank to rule out switching noise, but the hum is still there.

So I am now looking at alternative solutions.

Amplifier:
https://uk.rs-online.com/web/p/analogue-development-tools/1360830
- uses LM48100Q-Q amplifier (class D ?)
- mono output but has 2 inputs (left and right channels)
- both inputs can be mixed and each input has gain control using I2C, so no need for the summing op-amp

Speaker:
https://uk.rs-online.com/web/p/miniature-speakers/2596151
- Same Sky 1.5W Miniature Speaker
- 8 ohm

Will this give me better results ?

I am really struggling with the audio/amp part of this project and its a road block to completing the project, so if I cant get an amplifer and speaker working (which I thought would be straight forward) then Ill have to give up

 

(you likely are aware that speaker impedance does not unarily define the share of E converted into sound pressure)
have you tested different headphones sets (?)

 

the hum is still there.

Is there hum with no input connected?

 

Is there hum with no input connected?

Remember that hum is likely from the power line, so if you away from that, you shouldn't have any hum.

 

(you likely are aware that speaker impedance does not unarily define the share of E converted into sound pressure)
have you tested different headphones sets (?)

I’ve tried 3 different speakers…

https://thepihut.com/products/mono-enclosed-speaker-3w-4-ohm?variant=31955927249

https://thepihut.com/products/20-x-...-ph1-25-connector-8-2w?variant=43764408516803

https://www.torkworld.com/collections/frontpage/products/tork-xpro-element

 

Is there hum with no input connected?

No not with no input connected

 

I used a powerbank and the hum was still there...

https://www.anker.com/uk/products/a...97764&gad_source=1&gad_campaignid=22353039033

 

No not with no input connected

How about when you ground the input?

It would appear the hum is from your signal source.

 

just, a stupid thought -- do you have an "on board" output filter = the speaker membrane has a mass and mechanical inertia ? when you HF juggle it around ?? the dynamic range may have an impact . . . maybe . . .

 

How about when you ground the input?

It would appear the hum is from your signal source.

Very slight increase in noise if I ground the input (compared to a floating input)

 

I would think main's hum would not be a factor for the circuit where it will be used.

 

I have been wondering and now I am finally asking : WHY put the audio signals into the non-inverting input, along with the DC bias??? The other posted TYPICAL circuits have them separate. Bias fed into the non-inverting side, and the signals, along with the feedback, into the inverting input. There are reasons for that and they make a lot of sense.

 

Agree. A non-inverting summer stage has less isolation among the inputs. Also, the through gain decreases as the number of inputs increases unless you adjust the summing resistor values. There is a case where that is a good thing, but the vast majority of summing amplifier applications I've seen have an inverting summer, either preceded by inverting input stages or followed with an inverting output stage to preserve signal polarity.

ak

 

I have been wondering and now I am finally asking : WHY put the audio signals into the non-inverting input, along with the DC bias??? The other posted TYPICAL circuits have them separate. Bias fed into the non-inverting side, and the signals, along with the feedback, into the inverting input. There are reasons for that and they make a lot of sense.

If I put them in the inverting input, wouldnt I need a bipolar supply (negative and positive) ? As it will be powered from a 1S LiPo battery, I would need a DCDC to generate a bipolar supply.

 

? is your chip authentic ? is it operational (hasn't not overheated)

 

If I put them in the inverting input, wouldnt I need a bipolar supply (negative and positive) ? As it will be powered from a 1S LiPo battery, I would need a DCDC to generate a bipolar supply.

No.
The LM4871 is internally biased to operate from a single supply when the input(s) are connected to the inverting input through a capacitor.

 

If I put them in the inverting input, wouldnt I need a bipolar supply (negative and positive) ? As it will be powered from a 1S LiPo battery, I would need a DCDC to generate a bipolar supply.

NO, you do not need a bipolar supply, because your design already biases the output to 1/2 Vcc. That should have been obvious. AND, that bias is needed no matter which input is used, for single supply operation.

 

So like this then....

 

hi ff,
The TLV has no gain, just a buffer???
E