Music Module output signal is DC offset

The output from the Music Module is a DC offset and I could not get any sound out of a DFRobot Audio board.

At first I thought the audio module had no output signal but when I checked with the scope I saw that without anything playing the output was a noise DC signal or approx 2V

with the sine test running I can see that the audio signal is super imposed on this.

The audio module was seeing the signal at the input to the LM386 but there was only a zero signal on the output. Checking the output from the LM386 it had a DC offset.

After I fitted a capacitor between the Music Module output and the audio input I was not getting sound from the system.

For anyone else experiencing this issue, was out for this if you audio amp has no capacitor on the input to remove the DC component.

PS… The output from this music module is very noisy when nothing is playing. You can see this on the scope. I now need to consider adding a relay to the output to switch it off when not playing any sounds otherwise it is going to drive people nuts! :slight_smile:

Ok, sounds like an advice I dont understand. I am struggling with noisy output from the module as well, and decided it was the power-supply.

Are you saying the noise can be removed by putting in a capacitor, then please let me know exactly where and how big… :slight_smile:

I decided this as well, but not sure it would work.

The capacitor won’t remove any noise. It only blocks the DC portion of the signal and allows the AC part to pass through to the amplifier.

Page 14 of the VS1053 datasheet makes for interesting reading.

GBUF on the VS1053 is connected to what would become the GND connection if you use external amplifier and not headphones as the design is suitable for. In my case the external amp is powered from the same supply as the one that feeds the Music Module.

I have tried disconnecting the GND connection on the headphone jack and leaving the power GND to act as the GND for the 2 connections.

After doing this I still get the noise though but it may be that my input is too high impedance for the output so I may try an artificial load on it and see what happens.

Based on this, if you use an external amplifier, you should not connect the GND pin on the 3.5mm jack and connect the outputs to a capacitive input stage on the external amp.

I used a 150uF 35V cap on the output. It was just something I had lying around but you do need a large value here.

I also found this and it has the recommended output for driving an external amp on page 7.

You’ll also see a mute signal connection. Can probably use a GPIO pin for this.

1 Like

Hmm, I have tried comparing the schematics from the music module, and the page 7 from the pdf you linked. It looks like GHI actually follows the recommended transient free line out?

Yes but it is designed for use with a headset as the common on the socket is connected to GBUF and this MUST NOT BE connected to GND. If you us and external amplifier powered from the same supply that the Music Module is running from, your GND will be connected to GBUF.

The alternative line output in the datasheet shows GBUF not connected to the GND of the amplifier input.

I am scrapping these DFRobot amps as the output is too low and there is a lot of distortion. I am waiting for a newer higher output amp to try that.

All you need is to make sure the COMMON on the 3.5mm jack is not connected to anything and you have a 10uF cap on the output of each channel and then it will all work fine.

Ok, I will try running both systems in the future from the same power-supply, are you saying that it would not be possible?

Is there an external amp that will not have a GND connection, I dont think so…?

Confused.

I am excited to see what the result is.

I see a dangling common in front of me, and only one wire going to the amp, it sounds strange…?

I can see I am a noob here in the circuit arena.

@ Dave McLaughlin - How about a transformer coupler. That will also isolate the grounds.

That’s the thing, it is not 1 wire but 2. The audio output plus GND if your supply is the same. The output is then referenced to GND but needs the caps to remove the DC offset.

If you look at the second circuit in the datasheet it has GBUF unconnected and the GND of the circuit is the GND for the output to the amplifier. The outputs are via the 10uF caps.

Ahh, the dangling end just got connected. Thanks

@ Dave McLaughlin - I am have tried a new PSU and some caps, but i still have too much noise in my speaker, connected to a small amplifier module.

Can you draw a quick schematics of your proposed connection, it sounded simple, but I am sure I misunderstood you. I am using this amplifier http://www.ebay.com/itm/Mono-10-Watts-Car-Radio-Audio-Amplifier-Module-Board-/400397255443

Thanks again!

I’ll knock up something later today

I too get a lot of noise when the system is idle. I have switched to another device for now so I can get speech and sounds. I just never got anything good from the music module.

The noise issue is something that is in the manual for the IC and they have a solution for this.

You don’t need any caps for your amplifier. It has a 10uF cap on the input, after the volume pot.

Just connect the left or right output to the input on this amplifier. Don’t connect the GND on the headphone socket as we discussed before. If your power supply is common to the music module and the amplifier with a common ground connection, you won’t need to connect the GND connection on the input to the amplifier. You will be OK if the leads are short. If you need to use shielded cable, just connect the shield to the GND connection on the amplifier and make sure it is not connected at the music module. :slight_smile:

By the way, I never tried this but have you tried setting the volume to ZERO when playing is stopped and then switch this back to the last setting when music starts playing again? Does that work?

@ Dave McLaughlin - Does this noise only occur when using an external amplifier, or do you also hear it when using headphones connected directly to the Music Module? In my many hours of testing and listening I never heard any noise between songs or when a song ended when using headphones.

@ jasdev. Interestingly I never tried until now and you are right, there is no noise.

Reconnecting to the external amplifier and I hear the hissing sound again.

I wonder if this has to do with the DC offset?

I found this schematic of the output circuit of another manufacturer’s VS1053b. I don’t own this module, so I don’t know if it works better or not. Is this similar to your ideas for decoupling the output?

Where does L and R on the top right go to?

Is there a link to the full schematic?

The phonejack is still got GBUF connected so you can’t connect this to an external amplifier it the power GND is common. The datasheet states that GBUF must not be connected to GND at any time.

It is the mikroElektronika ARM Cortex M3 board.
Here is a link to their user manual:
http://www.mikroe.com/downloads/get/1833/mikromedia_stm32_manual.pdf

I see that GBUF is connected to the common on the phone jack, but the line out connections use board ground as the common. The line out L and R go to pads on there I/O connectors.

Hmm, when you two guys are struggling with this as well, I am not feeling that I am completely far out with the problem.

I was hoping that the noise was due the bad timing of the Spider board, and this is progressing slowly with the RLP driver coming soon I hope.

I was hoping that it was possible to simply hook up an external amplifier and this seems to be more complex than first anticipated.

Are we on a bad route here, is the music module a bad choice for playing music from SDcard in speakers?

@ njbuch - The music module works, but it would have been better if it had line out connections that are designed to be connected to an external amplifier. I’m starting to look for alternatives now.

Does VLSI provide any reference documents in this regards? Maybe contact them directly? I do not think any of their chipsets support line out. They are made to go in MP3 players typically.