LM49450: Onboard amplifier audio cuts out

Another to note, if the speaker output pins are too close (surrounded) to a ground plane on either side (top or bottom of the boards), could that trigger the overload protection and force the audio to go out of the aux headphone output instead? 

Hi John,

Have you inspected the voltage on HPS pin? Is it close to the threshold rather than being strongly pulled up or down?

Best regards,
-Ivan Salazar
Applications Engineer

Ya, it is measuring 0.02V. With a headphone plugged in then it goes up to 2.49V as expected. When the amplifier cuts out audio to the speakers on it's own, the HPS pin still stays at 0.02V until headphones are plugged in.

I can replicate the problem 100% everytime if I ground a speaker output pin, which is why it's leading me to believe that since they are surrounded by ground planes that it could be triggering the short circuit protection in the amplifier. We are considering adjusting the ground planes so they are not immediately next/close to the speaker pins and give it a bit of breathing room.

Another question. We have all the recommended caps close to the pins as described in the documentation. When powering the system direct with 1 cable, either from a 12v source which drops it down to 5v, or 5v directly, the cut out happens as described. However when adding an additional cable between the power cable and to the board with the lm49450, then the audio never cuts out from the speakers. 

Hi John,

Can you share a schematic to better understand how the power is connected?
Perhaps you have a ferrite or choke with limited current spec before reaching the amplifier, and it helps when you add a direct connection to the supply bypassing those components.

Best regards,
-Ivan Salazar
Applications Engineer

Hi John,

Can you specify where in the schematic the cables you mentioned are connected to?

Best regards,
-Ivan Salazar
Applications Engineer

On the 5V. There are some pads to connect a cable at DTR / RX / TX / 5V / GND / GND

I am thinking of adding 100nf caps next to the three 5v pins on the lm49450 (which is located on the far left in the image). I think something may be triggering the circuit protection.

Ivan,

A few changes were made, 100nf caps were added close to the three 5V pins on the lm4950 (pin 16 / 18 / 24) and removed the copper pour that is close to the speaker pins.

Do you think the C8 capacitor for pin 10 and 11 (DVDD and IODVDD) should be changed to two 1.1uf caps as per the datasheet or the one 2.2uf is ok? We have seen setups similar to this and they work with no issue. THe same with the BYPASS, 

Hi John,

I'll review this and get back to you by the end of the day tomorrow.

Best regards,
-Ivan Salazar
Applications Engineer

I believe you are correct about the choke effect of the cabling. Using longer cabling makes the system run longer before cutting out, which we have been able to test repeatedly. The extra decoupling caps should do the trick. Will test it out soon.

Hi John,

On a quick glance I noticed that the input barrel jack connector does not seem to have any decoupling caps, you may want to add some bulky capacitors there as well.

Thanks for the updates let us know how the new test goes.

Best regards,
-Ivan Salazar
Applications Engineer

Thanks, will do.