[SOLUTION] VS1053B vs PAM8403 (Class D amp)

Designing hardware that uses VLSI Solution's devices as slave codecs such as an external MP3 decoder chip for a host microcontroller.
Post Reply
madias
User
Posts: 5
Joined: Mon 2018-10-29 14:51

[SOLUTION] VS1053B vs PAM8403 (Class D amp)

Post by madias » Sat 2018-11-03 0:42

Hello, I know a similar topic exists here: viewtopic.php?f=10&t=2017&p=10414&hilit=pam8403#p10414
But this didn't helped me out:
Simplify circuit:
STM32F103xxx connected to a VS1053b dev board (this with SD card onto it) VS1053b phone out -> PAM8403. Driven with a Li Ion battery.
In short: Driving the VS1053b with a PAM8403 Class D amplifier gives a lot of noise (accessing SD-Card, SPI Port....), but NOT over the headphones.
I read many topics about that problem used recommended input and output filters , but without a real solution:
I do not have other noise problems anymore: Connect GBUF with Pin 8 (Internal analog reference) of the PAM8403 eliminates every other static noise, so in idle mode it's very "clean".
The only solution I found out is to drive the PAM8403 on a separate battery circuit - but this is no real solution (even living with that, but in charge mode everything is connected together again). The problem is that the noise (like SPI handling) is independent from the main volume, so it's always "there".
I know (by now...) using a Class D amp isn't the best solution dealing with a VS1053, but I ordered 15+ pieces for a player I have to make in a small series.
Last edited by madias on Sat 2018-11-03 19:19, edited 1 time in total.

madias
User
Posts: 5
Joined: Mon 2018-10-29 14:51

[SOLUTION] Re: VS1053B vs PAM8403 (Class D amp)

Post by madias » Sat 2018-11-03 19:19

Ok, I nearly got it with this circuit:
http://esp8266-server.de/Internetradio.html
--> "Schaltplan analoges Teil"
There is some - quiet - random noise left, but I'm sure this is caused by the bread board.

Hannu
Senior User
Posts: 43
Joined: Mon 2016-05-30 11:54

Re: [SOLUTION] VS1053B vs PAM8403 (Class D amp)

Post by Hannu » Mon 2018-11-05 7:26

I gave a quick look at the PAM8403 datasheet. My discovery was that in the block diagram input is fed to amplifier and the inverted input is biased to VDD/2. That would suggest making good analog design and having a good power supply.

First thing would have more bypass capacitors with various values so that ripple would be filtered out.

Another solution is to give the amplifier a own regulator.

Post Reply