problem with 4780

[Faber]

Hi all!
I'm experiencing some problems with my new project using the lm4780ta

I'm using the datasheet schematic that you can see as Figure 7. Only difference is that there are no caps on the input.

the problem is the following:
On the channel A all it's very nice, almost no DC on the output and so on.
On the channel B I have almost -18V DC out.... I can't understand were this voltage come from.
the strange thing is that I already made 4 PCBs for this circuit, and all 4 pieces behave in the same way (channel B with DC).

Another strange, on the pin 27 (V+) there are -12V instead of +24

I've already detached all the NC pins from the PCB but the problem persists. I've checked with friends the pcb traces and all it's compliant with the schematic... any idea? Do I have 4 broken chips?

TIA, Fabrizio

[Faber]

here is the image of the pcb.
Any ideas?

[Faber]

here is the complete pcb

[platenspeler]

Fabrizio,

Difficult to analyze your amp based on the two small files you posted, also you did not post a schema of your power supply.
You tell that channel A seems right, but with -12 Volts on pin 27, I take it that the complete amp does not work ... right?

? No shorts between the housing and ground for the non-isolated TA package?
? Without chip in place, does the power supply put the right voltages on the board (and pins 5, 27)?
? You get -12V on pin 27? What do you get on pin 5 and the other V+ pins?
? Is R2 the feedback of channel B connected between pin 25 and 21? (Hard to tell on the picture)

You have more than just 1 change from figure 7 I think:

• R10 and R11 on the output,
• R6 and R7 directly connected to ground without DC decoupling caps,
• No input caps
• and I would expect the inputs to be connected to the intersection of R3/R8 and R5/R6.

Hard to say whether you have broken chips, as these chips seem to have a large tolerance for abuse/mistakes.

Maarten