VS1053: very bad sound with normal in-phase audio, why?

Designing hardware that use VLSI Solution's devices as the system controller for the entire design.
Post Reply
DrDoom
User
Posts: 9
Joined: Tue 2018-12-11 16:21

VS1053: very bad sound with normal in-phase audio, why?

Post by DrDoom »

I looked again at one of my implemented projects because the sound was still very bad.
i adjusted and tested a few settings in my vs1053 library. With the SCI_MODE -> "SM_DIFF" bit = 1 (left channel inverted), great sound is played on the left speaker. ka ching. So is there any problem with the phase (hardly understand anything about audio signals)?!

I'm not sure at which point I have to make adjustments. I am sure it is wrong wiring. Can you give me a tip?

Since I have relatively little space in my housing, I had to solder the line-out cable to the VS1053:
IMG_20201231_095440.jpg
IMG_20201231_095440.jpg (535.11 KiB) Viewed 1102 times
line-out cable (yellow = left/right, black = GND)
line-out cable (yellow = left/right, black = GND)
IMG_20201231_095112.jpg (323.09 KiB) Viewed 1102 times

The audio signals run through a 3.5mm jack into a mass separation filter and from there directly into the amplifier.
mass separation filter (red/white = left/right, silver = GND)
mass separation filter (red/white = left/right, silver = GND)
IMG_20201231_095225.jpg (267.73 KiB) Viewed 1102 times

As I understand it, everything should be ok, but only when the phase was changed....one of the speakers worked perfectly.
Without the SM_DIFF, the sound on "both speakers" is terrible.
User avatar
pasi
VLSI Staff
Posts: 1783
Joined: Thu 2010-07-15 16:04

Re: VS1053: very bad sound with normal in-phase audio, why?

Post by pasi »

Probably a ground issue or ground loop or you have connected GBUF to ground. GBUF is 1.2V, and should not be connected to ground.

Do you have any schematics of your connection? How are left, right, and gbuf connected?

Have you checked viewtopic.php?f=9&t=69 ?

You get good sound from left, but do you get sound from the right with SM_DIFF set to 1?
Visit https://www.facebook.com/VLSISolution VLSI Solution on Facebook
DrDoom
User
Posts: 9
Joined: Tue 2018-12-11 16:21

Re: VS1053: very bad sound with normal in-phase audio, why?

Post by DrDoom »

Ohhh k, sometimes you just need the pusher to test something that has already been successfully tested.

I wrote a test routine to switch to the left or right speaker during playback every 5 seconds. I noticed that only the volume changes (stereo or one speaker), but not the speaker itself.

I was sure, that I had written my library correctly. Then i discovered that the mass separation filter was the cause.
- Without the mass separation filter, I can set the volume for both speakers correctly but with digital noise.
- With the mass separation filter and "without" SM_DIFF, I get poor audio quality on both speakers.
- With the mass separation filter and SM_DIFF, I get super audio quality on one speaker (left) and poor or dull audio quality on the other (right).

Well, it's the way it is. i need the mass separation filter and will live with it in this project. the overall audio quality is perfectly adequate. in future projects (i love the VS1053) i will pay attention to another separate power supply for the components during the design phase and will do without a mass separation filter.

Thank you again and see you soon.
Daniel.
Post Reply