TDA7379 simple amp noise

[smarg]

still i dont understand how am i suppose to fight the noise with voltmeter...any more detailed instructions would be welcomed...meanwhile i connected 2 9 volt batteries instead of wall brick for a test, and while the amp played with heavy distortion (probably the batteries cannot supply enough current), it was dead silent when idle and input cables connected.....so assuming the noise is some kind of power ripple , how do i eliminate it? adding more decoupling capacitors?

 

[Arty]

show exactly the psu, and how its wired up.
something is messed up there, and untill we see the exact wiring there is no way to tell what is the problem.

 

[Grasso789]

and i noticed now something interesting - as long as there is incoming signal (phone playing some music) - there is no noise in the output signal. i mean i can't hear it even when i low the volume almost to silent. however when i stop the playback - the noise returns after couple of seconds. maybe these hints can help....

This behaviour is due to the source shutting off its headphone amplifier when not in use. Shut off, source impedance becomes hi, so capacitive coupling to noise is not shortcut by lo source impedance. If we turn volume to full left, parking the potentiometer wiper on ground, noise should stop.

If impedance is really hi, then even a short cable might be too long, if not shielded. Also try moving things around.

 

[Arty]

su supposedly a resistor straight from input to ground heal it ?
something like.. 10K or so ?
(just asking. never had this kindof issue)

 

[smarg]

show exactly the psu, and how its wired up.
something is messed up there, and untill we see the exact wiring there is no way to tell what is the problem.

the psu is a generic wall brick providing 12v 2a output. i dont have its schematics. i googled its model and seems like it is a switching power supply, like the one below. switching is bad?

EU Plug Adapter AC 100-240V to DC 12V 2A Switching Switch Power Supply Converter | eBay

 

[Arty]

well, these chepo switching powersupplys can do interesting things so to say,
i would not be surprised if that was one source of your problems.

i would not trust these things ever.
maybe you could salvage a proper transformator and build a psu ?
that would be the route i would take.
as from battery power the thing works allmost fine, expect for the headphone cut out noise, witch should be solvable with a resistor paralell to the chipamp input.

 

[smarg]

hmm... i wanted to avoid dealing with mains voltages for now....i will see if i can find ready transformer based psu...maybe something that puts out AC so i only need to rectify it...thanks

 

[Arty]

something that puts out 10 to 12 VAC would be sufficient.
50-80 VA rated one will be more than good enough.
then supposedly you just have to rectify it with a diode bridge, and add some reservoir/buffer/tank capacitors. preferably rated to 25 volts.
something like 10000 uFD /25 V will do the trick.

 

[smarg]

Still...my original question is unanswered. When I touch the ground line, all noises disappear. This gives me some hope..how the circuit "sees" me? What happens here? What can be connected to the ground (instead of my finger..) to have the same effect?

Sent from my MI 4W using Tapatalk

 

[Arty]

well, you could try a pair of diodes in anti-paralell bypassed by a 10 ohm is least 5 watt rated resistor, separate power ground and signa ground and connect them with this contraption as a ground lift network.
it *should* work.
you would allso need a 15 Kohm is resistor from input to ground to set dc parameters for the amplifier, that may allso help.

 

[smarg]

Can you please explain what each such method does or point me to relevant material? I am trying to learn. ..thanks

Sent from my MI 4W using Tapatalk

 

[Arty]

if i find enough time in the weekend i make you a drawing.
rebuild your amp according to that and supposedly it should work as intended.

 

[smarg]

Arty, thanks for your help...drawing is certainly nice, however my goal is not to copy someone else schematics and just build the thing (this is too of course..), my main goal is to learn in the process and to understand why the things behave the way they do, according to what laws, etc...so if you could explain for example why i need resistor to ground from input, what this does ,etc...this would be great! thanks again

 

[smarg]

So, it was power supply to blame..I switched to transformer based power supply and the noise is gone. ..very faint noise still can be heard very closely to speaker ( and it still goes away when I touch the heatsink), but practically I can say there is no noise anymore.

Sent from my MI 4W using Tapatalk

 

[Grasso789]

Still...my original question is unanswered. When I touch the ground line, all noises disappear. This gives me some hope..how the circuit "sees" me? What happens here? What can be connected to the ground (instead of my finger..) to have the same effect?

A metal statue of thy size and proportions. Thy body acts as a capacitive (or electromagnetic anyway) shield. That switching power supply obviously uses a transformer with a strong stray field.

Cheers

 

[Arty]

i have nothing against switching mode powersupplys, that are well built.
but few dollar worth of chineese "stuff" is something i would allways avoid as you can see from this example..

 

[Grasso789]

Most "stuff" comes from East. West must know, because he is the designer for, secret owner of and shipper from these eastern factories.

 

[sgrossklass]

SMPS' tend to have some pretty interesting ideas about where their output ground should be. This problem really sounds like a common-mode noise issue. It would be interesting to know whether putting the circuit inside a metal case connected to ground eliminates the noise - if so, it's capacitive coupling of some part to the environment. Connecting it to a source that is earthed should have a similar effect.