LM4871 amplifier circuit not working, please help!

[catalin123m]

Hello,

I have made a board based on an atmega328P, that reads a wav file from the SD card, and then plays it trough an LM4871 amplifier.
I followed the instructions on the datasheet for the LM4871, but I had no sound into the speaker.
It would produce a pop when I was changing the shutdown pin from LOW to HIGH and from HIGH to LOW, or more popping when I touched the pins with my hands, but that was all. Everything else is working fine, because I connected an piezo speaker directly to the output of the Atmega328P (aka the input of the amplifier) and I could hear the song playing.

NOW, after I fiddled a bit with it, I`ve connected together, by mistake, the Shutdown pin and the Bypass pin of the LM4871, and I started to hear sound coming out of the speaker. At this point I could hear "EUREKA" in my mind ).
Few days after this I`ve noticed that the amplifier was not shuting down anymore and it is getting hot,when I connect a speaker to it (also there is no input, the output pin on the 328P is set to LOW)

I`ve never made a schematic for my project. I only made the PCB design, so I had to make one now to show it to you, and for simplicity, in this schematic I used a Attiny85, instead off the 328P.

CAN SOMEONE PLEASE, PLEASE HELP ME WITH THIS?

What is the problem??

 

[(*steve*)]

Reverse engineer your PCB design into a circuit diagram.

 

[catalin123m]

It's basically the same thing, except instead of attiny there is and atmega 328 with pin 9 connected to the amplifiers input, pin 8 connected to amplifiers shutdown pin, pins 10,11,12,13 connected to and sd card reader (mosi, miso, clk, cs), pin 2 to a coin acceptor, pin 3 to a servo, pins tx,rx and reset connected to a ft232rl chip and the rest of the pins connected to an i2c lcd, a keypad, a led, a button and a mosfet switch.

All works, except the amplifier, wich is connected to the rest of the components only by these 2 pins , input and shutdown. And also it shares the power rail with the microcontroller

 

[(*steve*)]

The problem is that you likely have a mistake on your board, in your soldering, or perhaps in your actual circuit diagram.

However we can see none of these.

If you post pictures of your board, the first thing we will have to do (unless there is some really obvious fault) is reverse engineer the circuit diagram from it, a task you should do yourself.

You MUST be able to confirm the PCB reflects the circuit you want. Then we can examine the actual circuit, look at the datasheet for the device (please post a link to it) and see what we can come up with.

 

[73's de Edd]

Sir catalin123m. . . . . . .


Replotted . . . . . . on the internals of the IC aspect .

Hmmmmmm . . .an activity stream . . . . . aural monitor . . . . .

Refer down to the RED boxed area on the data sheet and their treatment of pin 1 with its pullup resistor and then the required "LOW" action.

Then compare what you have with the series inserted selection of a 4.7 k resistor.

Wonder if there is a possibility that you have shot yourself in the foot with the selected ratio of those resistors in acquiring a solid and definitive LOW from atiny pin 5..


RePlot:





73's de Edd

 

[CDRIVE]

As per the LM4871 datasheet bringing the shutdown pin to Vdd puts it in Shutdown mode. The amplifier is active when the shutdown pin is grounded.

Are you sure you need R3 (4.7KΩ)? I don't see why you can't connect the shutdown pin directly to the uC.

Chris

 

[CDRIVE]

Ed, we doubled.

Chris

 

[catalin123m]

First , to answer to 73s Edd, there is no ratio formed by the resistors, because there is only one resistor. In the datasheet it`s sugested to use a Pull-Up resistor to pull the shutdown pin HIGH and a single throw switch to ground, to activate the shutdown function. That is one aproach. Instead of this, I connected the shutdown pin directly to the micro controller, so I can pull it HIGH or LOW from software ( to automatically turn the amplifier ON or OFF, when I neet it)

Second, in my original design the shutdown pin was directly connected to the microcontroller, without the 4.7K resistor, but I, decided to modify the board and insert this resistor, to protect the microcontroller, in case of a short or something like that. ( the microcontroller, only supports 40ma on the output) The shutdown pin is high impedance, so it does`t draw any current from the microcontroller. It only gets pulled to VCC or Ground. So the 4.7K resistor does`t affect the shutdown function.

Third, Just to be sure, I also tried bypassing the resistor with a jumper wire, and tried to connect it directly to ground. but the result was the same as with the 4.7K resistor.

I was hoping that somebody else used this amplifier and had success with it

PS: THANK YOU VERY MUCH TO ALL OF YOU FOR TRYING TO HELP ME. I`M GRATEFULL

 

[catalin123m]

here is the pcb layout. Everything is like in the datasheet, but there is no output on the speaker.
When I connected Pin1 and Pin2 of the amplifier (Shutdown and Bypass) together, the amplifier started to work. I could hear sound coming out of the speaker.
But now, as I said, the amplifier isn`t shuting down anymore, not even connected directly to ground. and it`s getting hot when I connect the speaker to it, even with no input.

 

[73's de Edd]

Sir catalin123m . . . . . . .

In your last submission . . . . . of your PCB layout . . . . .along with the mentioning of the chip running hot.
How was your thermal cooling treatment of the DAP aspect of that chip. I don't even see that it is being incorporated.
I include info, along with the minimal thermal implementation aspects of the parent chip manufacturer.

Tech Info 1:
http://www.neurochrome.com/soldering-ics-with-exposed-dap/

Tech Info 2:



A work-a-round tip:

In the past I have been known to get some copper foil from a crafts supply . . . . . . used for copper tooling of small art murals.
It is easily being the thermal equivalency of 1 oz copper board cladding . . . if not far exceeding.
Then I form a calculated sizing, small right angle, double bend copper bracket like this [
The central part gets soldered to the DAP tab and the two edges are made of such a length as to rise a bit above the height of the I.C. chip proper,
then I solder my heavier gauge "slab" of copper heat sink to the top of those two thermal risers. Bent up end fins on that, also spread out the cooling area
Carried off heat then radiates upward from the mass, all sitting just above your IC.


73's de Edd

 

[catalin123m]

I am a beginner, so I missed that part of the datasheet . But I am planing to put a radiator on the chip, using adhesive non-conducting thermal paste.
But I also have to mention that I only want to play short "melodies", like applause effect, boo-ing effect, or something like that (2-3 seconds), few times a day, so the heat generated from such a short use, shouldn`t be a problem even without cooling.
The problem I am facing right now is that the chip get`s hot and it`s not shuting down, not even when there is no sound playing and the shut down pin is pulled to GND. And with class D amplifiers being 90% efficient, this shouldn`t happen.

Thanks again for your help!

 

[Alec_t]

It's not shutting down because, as per the datasheet "the LM4871's micropower shutdown mode (I Q = 0.6µA,typ) is activated when V DD is applied to the SHUTDOWN pin."

 

[catalin123m]

And I'm doing just that by seting the microcontroller's output pin 8 (the one connecred to the amplifier's Shutdown pin) LOW

 

[CDRIVE]

And I'm doing just that by seting the microcontroller's output pin 8 (the one connecred to the amplifier's Shutdown pin) LOW

No, you're doing exactly the opposite. To be clearer ...

Shutdown Pin High = Amplifier OFF
Shutdown Pin Low = Amplifier ON

Chris

 

[catalin123m]

Yes , you are right, CDRIVE. I misread the datasheet. Thank you very much! Now all seams to work, but I haven't tested it enough.