TPA3116D2 Amp

Yes, for the hiss either a 5K potentiometer, a 4K7 by-pass of the initial potentiometer outputs to ground or a reduction of the gain to 20dB.

Check if the two chips are not already synchronized.

The big black components are likely to be decoupling (capacitors) of the power supply rail. They look like foil types which means the board has no big electrolytes for decoupling. If one is about to go loose, it indicates rough handling.

True, thermally conducting glue is annoying.
 
Disabled Account
Joined 2010
Hello, how did you did the "deadman switch"?
Thank you :)


The deadman switch uses the pre-fade listen signal that feeds the master vol pot - thus it is independend of master vol setting.
This signal is connect to an analog comparator input of an Attiny25. Any peak exceeding 20..30mV triggers an interrupt that restarts the timer.

With that microcontroller it is easy to achieve timeouts of 15minutes or even more.
And this makes sense in my case as the Attiny acts as a power-manager including SLA-battery management, so this useful function requires little additional circuitry.
 
Member
Joined 2018
Paid Member
Thanks for that.

I can't follow the board to see which resistors are for the syncing.

It is only the black plastic that has come loose on one leg. It just looks like a piece of medium fuse wire underneath!

The postman had thrown it over the gates!


I may just leave the pot at full tilt. Is that a solution if volume control isn't required.?
 
Member
Joined 2018
Paid Member
So I've killed the aforementioned board.

TPA3116D2 DC24V 2x 100W two-channel stereo Class D digital audio amplifier board
Online Shop TPA3116D2 DC24V 2x 100W two-channel stereo Class D digital audio amplifier board | Aliexpress Mobile to clipboard

Had another play with it last night after a quick test to prove it worked. Well it didn't go to plan. One channel dead silent, the other is pulsing the driver in and out every second or so.

I'm no EE but I put a DMM over various places in vein hope. I do have fluctuating 16v DC out of the pulsing channel and about 4vac.This is with 19v supply. Other channel seems to have nothing.

All I have done is pop off the heatsink to try to figure out the gain resistors.

Do you think I may have lifted the chip/s whilst doing so.? Any other tests with simple DMM?
Bit miffed as I have now blown my first 2 d class boards!
 
So I've killed the aforementioned board.

TPA3116D2 DC24V 2x 100W two-channel stereo Class D digital audio amplifier board
Online Shop TPA3116D2 DC24V 2x 100W two-channel stereo Class D digital audio amplifier board | Aliexpress Mobile to clipboard

Had another play with it last night after a quick test to prove it worked. Well it didn't go to plan. One channel dead silent, the other is pulsing the driver in and out every second or so.

I'm no EE but I put a DMM over various places in vein hope. I do have fluctuating 16v DC out of the pulsing channel and about 4vac.This is with 19v supply. Other channel seems to have nothing.

All I have done is pop off the heatsink to try to figure out the gain resistors.

Do you think I may have lifted the chip/s whilst doing so.? Any other tests with simple DMM?
Bit miffed as I have now blown my first 2 d class boards!
Hi jimk04-San,


At first measure the voltage between GND and each speaker output terminals in DC range.

If theses voltage are not around the half voltage of mains, thus means something wrong about that chip(s)....
 
Just tried 24V on a 16 ohms dummy load, feeding it a square wave and cranking the amp all the way... so the output should be some 30W and the chip/board remained cool as a cucumber. Measured 47ºC with one of those infrared thermometers. Ambient temp is 31ºC. I don't think I'm even going to bother installing the heatsink
 
I would bet the MLCCs were less than 1 uf, maybe 0.1 uf. It takes time for the caps to come up to the half Vcc bias to support the AC signal on a DC only system. By using 10 uf caps, that time is increased considerably, more than the TPA3116's built in power up delay. And that may be the cause of your power on pop.


Yesterday I could found out removed MLCCs, I measured it's capacitance. It was 1uF 2.0mm x 1.25mm MLCC, This board has 26dB gain, thus means calculated cutoff frequency was 1.6Hz. But In this case biased input voltage decreases MLCC's capacitance. So I need change to polyester 1uF Cap or few times bigger capacitance (2.2uF - 3.3uF) MLCC anyway.
 
I got another Bluetooth receiver to test with my TPA3116 amp. It produces a loud buzz through the speakers when connected but with no audio file (source is MP3 on laptop) is playing. As soon as I hit Play.the buzz disappears.

If I use the same receiver with my earpods there is absolutely no noise when no file is playing, even if I wind the output volume up to max.

Is this an input impedance problem and can it be easily solved or is it something else that I'm not aware of, such as a ground issue with the amp?