Amplifier noise/heat up the wazoo (TDA2009)

[Rattanee]

The amps feed a Zobel network which will provide a heavy load on the amp at high frequencies. Does the amp get hot with no external load?

You have to stabilise the amp (no white noise) with appropriate capacitors near the amp. If you have a scope, check for high frequency oscillation.

I'm sort of tangled with all the different things I've tried, but I believe that when there were no speakers connected, the amps staid quiet and cool.

I do a lot of digital troubleshooting but this is pretty much one of my first endeavours into audio related things. What should I be looking for with the scope, and on what pins?

 

[davenn]

I think its time to show us several close up SHARP and WELL LIT pics of Your amp construction top side and bottom side of the PCB ( I assume you built it on a PCB)
sharp and clear enough that we can read component values etc

We may see a fault in the wiring etc

Dave

 

[duke37]

You should look at the output of the amps for signs of instability.

If the output capacitors are connected the wrong way round, they will short the amplifiers when the speakers are connected. Check for DC across the speakers.

The volume controls are connected in an unconventional way. Is this recommended in the data sheet?

Is the power supply stable?

 

[Rattanee]

You should look at the output of the amps for signs of instability.

If the output capacitors are connected the wrong way round, they will short the amplifiers when the speakers are connected. Check for DC across the speakers.

The volume controls are connected in an unconventional way. Is this recommended in the data sheet?

Is the power supply stable?

I'll see what I can dig up. Unless the datasheets' schematics are incorrect, no cap should be installed the wrong way.

What's wrong with the volume controls? I realized later that they would work in reverse when connected up as such, but I don't mind that... it's supposed to be set and forget so I don't mind. Or is there something else wrong there?

I powered it from the battery of my cordless drill and it still happened.

 

[duke37]

The circuits that I have seen have a 100nF capacitor across the power supply and also a big fat electrolytic. You do not have the electrolytic.

The volume control normally has the wiper as the output. You have put yours in the other way round so that when the control is at minimum, there is a resistor in series with the amp input and the input is shorted.

Yes, the caps should be connected the right way, but are they? Have you measured the DC voltage across the speakers?

Incidentally, the circuit shown by Google for a TDA2009 power supply is a good way of blowing up a transformer or starting a fire!

 

[Rattanee]

The datasheet says a 100uf cap across the poersupply lines, that's there, though I will add a 4700uf later on as 100uf is clearly not enough and it allows for some noise feedback into the powersupply, which causes visible interference in my video signal. (Long story, this is to be part of a PC-arcade cabinet interface and the video amp clearly gets noise from this amp I'm sure)

I obviously messed up with the volume control. I severed the ground connections of the trimmers for the moment, but that didn't solve the heating issue. (The old simple stereo amp runs cool even though it has a heatsink only a quarter of this one size-wise) I need some added resistance though as the inputs are way too hot for clear audio.

I checked all the caps against schematics and datasheet, they all check out. I'll measure DC across the speakers in a short bit, just waiting for it to cool down a little. I'll also snap a few pics, but I'll only be able to upload them in the late evening.

Edit:
I think I have the problem narrowed down to one channel of the amp, which doesn't simply distort, but crackles as well. I also noted that when I meter the output of the good channel to ground with a continuity test, it beeps for a short while before shutting up (as the caps charge up), yet on the channel that crackles, I only hear a brief crack before it shuts up. The two sides are supposed to be identical so... I might have a bad cap in there... will see shortly. Weird as I only use brand new Jamicon caps.

 

[Rattanee]

So my speaker setup is a midrange and a tweeter for both sides, with a filter cap on each. (Sounds fine with the old amp so there should be no probs here.) I metered the outputs on the PCB, and both speaker outputs have DC across them, 5.3v on one, 5.2 on the other. What does that mean ?

Also, the bridged amp was apparently silent. I found a solder joint that went cold, patchign that up, now the bridged amp works, and the random bursts of static are gone, except for a brief burst at power-on.

So technically it's no longer turning into a white-noise-fest if I leave it on, but it still gets warmer then I think it's supposed to be, and there is some fairly brutal crackling at any listenable volume, with one side being much worse then the other as far as the crackles go.

 

[Rattanee]

This is my layout. All cap markings are polarity correct, I double checked that.

 

[Rattanee]

Checked the DC against the working amp, it gives almost exactly the same 5.2VDC on the speakers... I'm really lost at this point... going to replace the TDA with the one from the working one and see what happens

 

[duke37]

"I metered the outputs on the PCB, and both speaker outputs have DC across them, 5.3v on one, 5.2 on the other. What does that mean ?"

What do you mean by speaker output? The amp output or the speaker input after the capacitor?
The DC level of the amp output should be at about half the supply voltage so the output can be equally driven both up and down.
There should NO DC voltage across the speaker. This should be blocked by the output capacitor. If the capacitors are passing a significant current then the the amp will get hot.

 

[Rattanee]

I measured VDC between the two pins of J2 and the same for J3, results as above. They go directly to the speakers.

The voltage sort of checks out then as it's fed by 12VDC.

I have no clue what's wrong at this point. I replaced the TDA with the one from the working amp and got the same crackling mess.