Portable bluetooth amplifier speaker! noise problem, tips needed

[kaning]

Hi all,

Im using a pam8403 2x3w module audio amplifier and BLK-MD-SPK-B bluetooth module receiver (datasheet here). for this thread sake i've made a simple schematic to show you the problem im having. last time i built this schematic (last summer) this noise was not ocurring.

I've been reading a lot and search about possible solutions, like proper grounding, wich i dont think its the case. anyway its my first audio project and im quite newbie in handling audio signals. i've tried several other things but with no success, here is a demonstration.

how it looks on the breadboard.



View attachment PKB_peripheral_circuit_design_2 (1).jpg

the only diference from theese pictures is that im using a 100uf capacitors on both battery and 1.8v output. also tried to to add extra cap more near the bluetooth pins but nothing changed.

important note: this problem is somehow connected to bluetooth module because without it you can only hear small "ch" noise, when bluetooth module connected to phone and amplifier you hear whats in the video. many thanks in advance.

 

[KlausST]

Hi,

In my eyes most likely it is a GND reference problem.
To your amplifier (is it a class D?) You have one power gnd connection. Here you may expect high current peaks.
This causes voltage drop. If this shift your analog GND, thn you hear this as noise.

The same most likely is with your bluetooth circuit.
Here you have a power GND line carrying pulsed currents.

You either need very solid, low impedance GND lines from each device to a common star point
On a breadboard this is almost impossible.

Even better is to use a differential analog signalling between bluetooth circuit and amplifier.
Simplified it is possible to use a signal_gnd as reference...

Klaus

 

[FvM]

Looks like RF picked up by the amplifier input (or possibly other amplifier terminals). Not sure if this will ever work satisfyingly on a breadboard. Series resistors (e.g. 1 - 2k) directly at the amplifier inputs may help, also small SMD capacitors (10 - 100 pF) shorting the inputs against the a nearby ground pin.

 

[Vbase]

Hi all,

Im using a pam8403 2x3w module audio amplifier and BLK-MD-SPK-B bluetooth module receiver (datasheet here). for this thread sake i've made a simple schematic to show you the problem im having. last time i built this schematic (last summer) this noise was not ocurring.

.

The datasheet indicates that the speakers output from the bluetooth id differential, in probably means that it isn't referred to ground. It is OK when connected to earphones. The datasheet also recommends to use 4 serial capacitors if connected to an amplifier.
Try first to add those 4 caps.
If you still have problems try to connect the power of bluetooth to one battery and the power of the amplifier to another battery, the connections between the 2 circuits will be only 3 wires that go to your amplifier inputs.
Try also to connect earphones directly to the bluetooth to check if you get good sound there.
https://obrazki.elektroda.pl/6591333400_1430654382.gif

 

[kaning]

Looks like RF picked up by the amplifier input (or possibly other amplifier terminals). Not sure if this will ever work satisfyingly on a breadboard. Series resistors (e.g. 1 - 2k) directly at the amplifier inputs may help, also small SMD capacitors (10 - 100 pF) shorting the inputs against the a nearby ground pin.

thanks for the reply, i've tried that and no effect.

The datasheet indicates that the speakers output from the bluetooth id differential, in probably means that it isn't referred to ground. It is OK when connected to earphones. The datasheet also recommends to use 4 serial capacitors if connected to an amplifier.
Try first to add those 4 caps.
If you still have problems try to connect the power of bluetooth to one battery and the power of the amplifier to another battery, the connections between the 2 circuits will be only 3 wires that go to your amplifier inputs.
Try also to connect earphones directly to the bluetooth to check if you get good sound there.
https://obrazki.elektroda.pl/6591333400_1430654382.gif

with two diferent power supplies (both lipo) the "data sound" stops completely and just a small "ch" sound remains. but, for some reason, when using two diferent power supplies the sound stutters. with the 4 caps it just makes it lower but its always there.

- - - Updated - - -

Hi,

In my eyes most likely it is a GND reference problem.
To your amplifier (is it a class D?) You have one power gnd connection. Here you may expect high current peaks.
This causes voltage drop. If this shift your analog GND, thn you hear this as noise.

The same most likely is with your bluetooth circuit.
Here you have a power GND line carrying pulsed currents.

You either need very solid, low impedance GND lines from each device to a common star point
On a breadboard this is almost impossible.

Even better is to use a differential analog signalling between bluetooth circuit and amplifier.
Simplified it is possible to use a signal_gnd as reference...

Klaus

i havent quoted you because the other quotes answer you. also, the previous time i build it was on breadboard and this was not happening. yes, class d amplifier. Also, taking in consideration the tests i've made, it seems you are right and its really a ground reference problem. any tip how i can kill it?

 

[KlausST]

Hi,

I just reviewed all.

Then i saw that you connect SPKN of bluetooth module to GND of amplifier module.
But SPKN is not necessarily GND, therefore here current could flow.

You could try to disconnect SPKN completely. Try this.
But to be on the safe side you need the scematic of the bluetooth module.

Klaus

 

[kaning]

thanks for the quick reply! i've tried it now and it does exactly the same thing. only with two separate batteries it stops :/

 

[Vbase]

thanks for the reply, i've tried that and no effect.


with two diferent power supplies (both lipo) the "data sound" stops completely and just a small "ch" sound remains. but, for some reason, when using two diferent power supplies the sound stutters. with the 4 caps it just makes it lower but its always there.

I assume you tried to connect earphones instead of the amplifier and found that it works well.
The range of bluetooth is 1 meter, I guess you tried to get very near to the transmitter.
If you get no sound with batteries and some sound with power supply then it can be low RF reception which improves with earthing. Otherwise you have invented a clever device that can tell between battery power to mains power.

 

[kaning]

I assume you tried to connect earphones instead of the amplifier and found that it works well.
The range of bluetooth is 1 meter, I guess you tried to get very near to the transmitter.
If you get no sound with batteries and some sound with power supply then it can be low RF reception which improves with earthing. Otherwise you have invented a clever device that can tell between battery power to mains power.

first thing, im always testing on lipo batteries. the only difference is using 1 for both modules or 2 (one for each). im not using anything lse but lipo battery.

yes i tried it. only with bluetooth module operating with earphones/speakers it works perfectly.
using only amplifier direct to phone through cable and using speakers i get only the small "ch" sound.
NOTE: the "ch" sound comes from powering the module. only amplifier module and battery connectd its enough for the "ch"sound.

 

[Vbase]

Knowing that the bluetooth works well is a good start.
If the amplifier works well on its own then we know that it's the interfacing that doesn't work.
"using only amplifier direct to phone through cable and using speakers i get only the small "ch" sound." In this test you did is the phone connected to the input of the amplifier? If not do you have any source of sound that you can connect to the input of the amplifier to test if you can hear that sound well in the speaker? It can be MP3 player or portable radio.
If you establish that the amplifier works well then we will have to move on to the interconnection.

There is the possibility that the output of the bluetooth is digital (PWM) and it interferes with the oscillator of the amplifier causing distortions and hiss.
Put in series to each input a resistor of 4.7K to 22K and from each input of the amplifier to ground cap of 0.01uF to 0.1uF. This will be a low pass filter. If it helps then it will be needed to refine the filter.
Other possibility is that the signal level is far too high. Replace the capacitors of the filter with resistors of 1K and test it.

 

[kaning]

Knowing that the bluetooth works well is a good start.
If the amplifier works well on its own then we know that it's the interfacing that doesn't work.
"using only amplifier direct to phone through cable and using speakers i get only the small "ch" sound." In this test you did is the phone connected to the input of the amplifier? If not do you have any source of sound that you can connect to the input of the amplifier to test if you can hear that sound well in the speaker? It can be MP3 player or portable radio.
If you establish that the amplifier works well then we will have to move on to the interconnection.

both work perfectly by their own, the issue is on the interconnection.

the "ch" sound i fixed by using 100 ohm resistor on L+ and R+ amplifier output. with 50 ohm i still ear it so between theese values i should find the best option.

There is the possibility that the output of the bluetooth is digital (PWM) and it interferes with the oscillator of the amplifier causing distortions and hiss.
Put in series to each input a resistor of 4.7K to 22K and from each input of the amplifier to ground cap of 0.01uF to 0.1uF. This will be a low pass filter. If it helps then it will be needed to refine the filter.
Other possibility is that the signal level is far too high. Replace the capacitors of the filter with resistors of 1K and test it.

i've tried both options and it doesnt work quite well. the output sound gets quite lower and its still possible to ear it.

I dont know if you read that when im using two separated power supplies (one lipo for amp and other lipo for bluetooth) the sound stops. With 2 batteries and 100 ohm resistor (like i mentioned earlier) its working perfect! but ofc i will not have two separated batteries on the final product... any tips?

thanks a lot for the help so far!

 

[KlausST]

Hi,

On an internet page according the bluetooth module i found:

Stereo audio output can directly drive 40mW @ 32Ωspeeker without the need for DC-blocking capacitor

So it seems there already is an amplifier installed.
Maybe is class D. Do you know?

Especially the " without the need for DC-blocking capacitor" tells me that there might ba a problem with SPKN signal, as said in post#6.

Klaus

 

[Vbase]

In my first post I asked you to try 2 separate batteries and your answer was that the sound stops.
Now you telling that with 2 separate batteries and 100 ohm it works perfectly.
If you would have given me the second answer the first time it could have saved chasing our tails.

The reason for the problem is that the output of the bluetooth is floating and not related to ground.
In this case capacitors on the output will never work.
You need 2 audio transformers, one for each channel, in between the output of the bluetooth to the input of the amplifier.
Another option is to have an isolation inverter that is powered by your single battery and produces 5V for the bluetooth. This is an easier solution.
Only with isolation of power or signal you will be able to make it work.

 

[kaning]

In my first post I asked you to try 2 separate batteries and your answer was that the sound stops.
Now you telling that with 2 separate batteries and 100 ohm it works perfectly.

im sorry for the missunderstanding, what i meant was the bad sound stops (data sound).

Another option is to have an isolation inverter that is powered by your single battery and produces 5V for the bluetooth. This is an easier solution.

i cant power bluetooth with 5v, maybe u mean amplifier?

i was reading a bit meanwhile 2 audio transformers you mean like ground loop isolators right?(between bluetooth module and amplifier)

- - - Updated - - -

Hi,

On an internet page according the bluetooth module i found:


So it seems there already is an amplifier installed.
Maybe is class D. Do you know?

Especially the " without the need for DC-blocking capacitor" tells me that there might ba a problem with SPKN signal, as said in post#6.

Klaus

I will check and let you know!

 

[Vbase]

It is better if you call "bad sound" noise.

Because the amp takes more power it is easier to get an isolation inverter to power the bluetooth. If the inverter output gives higher voltage than you need you add a voltage regulator.
If you find inverter you can ask for our opinion before you buy it.

Use the high resistance of the TR (audio transformer) as primary.
Connect the primary of one TR to SPKLP and SPKN, connect the primary of the other TR to SPKRP and SPKN.
Connect the secondary of one TR to L input of amp and GRD, connect the secondary of other TR to R input of amp and GRD.

It is better to isolate with inverter because transformers spoil the bandwidth.
If you go for transformer get the smallest and the cheapest.

 

[KlausST]

Hi,

instead of transformer you could also use a stereo OPAMP.

differential OPAMP circuit with three identical 10k resistors and two 20k resistors.
The schematic (of a standard four resistor circuit) you can find in the internet. Just replace the single_resistor_to_GND with one 20k to GND and another 20k to VCC.

(you may change values. But mind to use 3 equal plus 2 equal with exactely twice the value of the others. )

Klaus

 

[Vbase]

Kaning,
Try Klaus idea first, maybe I will learn something too.

 

[KlausST]

Hi,

If the schematic is unclear, please ask..

Just a few words on how it should work.
The difference amplifier with four equal resitors reads the difference of both inputs an outputs it with reference to GND.
Now the use of two resistors intead of a single resitor makes, that the output no longer references to GND but VCC/2.
While both inputs must be wired from bluetooth module (short and parallel if possible) it is important that the vcc and gnd as referenc should be wired to the amplifier module.
I'd use supply VCC and supply GND from a star point. If there is no star point i'd try the amplifier's supply.
You may also test bluetooth supply.

If now there is some Ground bounce the circuit should compensate this.
The quality of this solution depend on OPAMP quality, the noise sources and the noise frequencies...

Goi dluck

Klaus

 

[kaning]

Hi,

instead of transformer you could also use a stereo OPAMP.

differential OPAMP circuit with three identical 10k resistors and two 20k resistors.
The schematic (of a standard four resistor circuit) you can find in the internet. Just replace the single_resistor_to_GND with one 20k to GND and another 20k to VCC.

(you may change values. But mind to use 3 equal plus 2 equal with exactely twice the value of the others. )

Klaus

i will need 2 channel opamp right?

edit: if i understood right is to use an opamp to connect the output signlas of bluetooth to the amplifier inputs, yes?

 

[KlausST]

Hi,

i will need 2 channel opamp right?

Yes

Klaus