TDA2009 A amplifier help

[milix]

Hi all!
I really new to this, so I need some help from experienced members.
I tried to build a TDA2009A based amplifier but without success.
The amplifier's schematic is:


I follow this image to draw and produce the pcb (sorry I don't remember where I found it to give the credits - I remember initialy this image was blue):


I used a laser printer and iron to transfer the toner to the pcb board and after etching i got this board which looks fine to me:


I solder the parts, and connected it to a +12vdc power supply from a modified pc power supply (it actually gives +11.92 vdc). the input was taken from a mp4 player and i used a test speaker 3 watt/4ohm.


But it gave me no sound. I tried to inspect where the signal is lost and I found that it is gone when it reaches the IC. The output from IC at pins 8 and 10 is 0mV. What can I do now to see what went wrong? Maybe I had destroyed the chip when soldering?
Thanks for any help.

[pacificblue]

Originally posted by milix
Maybe I had destroyed the chip when soldering?

That could be. But if this is your first project, it is probably the soldering. Either some pins are shorted due to too much solder or the soldering joints do not make good contact, because they were not heated up sufficiently.

You can check them with a multimeter. Set it for resistance measurement and measure from the component pin without touching the solder to the trace. You should get near 0 Ohms. If not, resolder.

The 220 µF capacitors' voltage rating is a bit tight. They will work, but 16 V or even 25 V rating would be a better choice.

[AndrewT]

The schematic does not look right.
Is this from the datasheet?

[pacificblue]

Except for the typo in the drawing (13k instead of 1,3k) an exact copy. Even the PCB distribution is similar. The datasheet uses more rounded and wave-shaped traces.

[AndrewT]

where does the TDA get it's input offset current from?
Both inputs are capacitor coupled!!!
The feedback with that 220uF does not look right either.
The output Zobel with just 1r0 leading into the feedback line looks wrong.

[pacificblue]

Page 4

[gfiandy]

Hi,

Check the PSU is the right way round. If it is the wrong way round the chip is probably damaged.

Not sure how your speakers are wired up but at the moment you only have two wires going to the speaker output and no 0v wires. There should be 2 x 0V return wires from the speakers so the two speakers will require 4 wires.

Maybee you know all this stuff already I am just checking the obvious first.

Board appears to be wired up cirrectly to the schematic. Application note indicates the chip must have internal bias for the inputs as no external resistor is shown on any application diagrams, so that should be ok.

The layout is not all that good but even so I would expect it to produce some output. If the signal is not getting to the amp are you sure your cables are ok from the source. I would check that the signal gets to the end of the cable with it detached then attach it to the amp. If the signal goes away when you attach it to amp then you have a short circuit from the input to ground somewhere.

If signal is OK till the side of input caps connected to the amp then the amp is probably faulty. (powered up with rails wrong way round?) You could check this by pulling the cap leg out of the board and seeing if the signal apears once it is disconected from the amp. If it does then the amp is almost certainly the problem.

Regards,
Andrew

[milix]

Thanks you all for the replies! I had connected only one speaker for testing purposes. The input is ok, it goes through the 2.2mf capacitors (sound is different -without bass- after the capacitors). The 220mf capacitors i used in my pcb is rated at 35v. I had also checked the connections of the ic with a polymeter, i have no shorts and the tracks are connected to the appropriate pins, i tested this very carefully. My soldering iron works at 400 oC and the datasheet says that the right temperature is 250 oC. Can I use something like a slot (or make one of my own) to connect the ic without directly solder it? If so how can I ask for this in the electronic store? Thanks again

[pacificblue]

There are so called test- or burn-in-sockets for this type of IC, but they are hard to get by, because demand is too low for most shops to keep them in stock or buy them in minimum order quantities.

This should be the right model. However they do not seem to have a sales office in Greece.

In AN-1192 National uses the Yamaichi SMT-15420 for an IC with the same case, but according to Yamaichi's ordering code it would be 15-pin instead of 11-pin. And no sales office in Greece either.