I managed to carefully pry it off, without taking off the chip with it(happened once)
Many thanks, Jared. I trust your readings and need no photo of the chip.
Turning to the datasheet ( http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf ), Table 1 on page 14. R4 corresponds to R2 in Table 1 and R5 to R1 in table 1.
R4 = 1 Ohm and R5 = 39K is no valid combination. But, a value of R2 much smaller than the value of R1 means we are in the lower part of the table where the operation is "Slave". "Slave" means the TPA3116 chip has to be controlled from an external clock. But, there is no such external clock and from previous experience with another TPA3116 board having a similar resistor flaw, I expect that the carrier frequency is some 130KHz while it should be 400KHz.
We need "Master" operation and that is possible by removing R4 such that we have a situation like in the first line of Table 1. The operation should change to "Master" and the gain be reduced to 20dB. 400KHz carrier frequency.
Do you know how to remove an SMD resistor, Jared?
You normally need not pull it with tweezers. The fine tip on the soldering iron, heat each end of the resistor in turn and soon the resistor will move on the PCB or stick to your soldering iron tip.
Check afterward that the two PCB pads for the resistor are not connected by solder.
Watch out for wrong values of the TPA3116 gain and clock setting resistors!!
In more cases we have found TPA3116 boards, stereo and mono, with wrong values of the gain setting resistors (datasheet http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf , Table 1, resistors R1 and R2). This does not only affect the gain but worse if the TPA3116 operates as "Master" or "Slave".
The symptoms if Master/Slave is wrong is excessive heating of the chip and the filter chokes, and hiss that will not disappear.
Recent production philosophy is apparently that as long as there are resistors, who cares about the values.
Try to identify R1 and R2 on your board, read or measure the values and compare with Table 1 to see if you have a valid combination and your TPA3116 is operating as "Master" or "Slave". If there is only one TPA3116 chip, it has to operate as "Master".
In more cases we have found TPA3116 boards, stereo and mono, with wrong values of the gain setting resistors (datasheet http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf , Table 1, resistors R1 and R2). This does not only affect the gain but worse if the TPA3116 operates as "Master" or "Slave".
The symptoms if Master/Slave is wrong is excessive heating of the chip and the filter chokes, and hiss that will not disappear.
Recent production philosophy is apparently that as long as there are resistors, who cares about the values.
Try to identify R1 and R2 on your board, read or measure the values and compare with Table 1 to see if you have a valid combination and your TPA3116 is operating as "Master" or "Slave". If there is only one TPA3116 chip, it has to operate as "Master".
Checked, it's set to master 36db.Recent production philosophy is apparently that as long as there are resistors, who cares about the values.
Try to identify R1 and R2 on your board, read or measure the values and compare with Table 1 to see if you have a valid combination and your TPA3116 is operating as "Master" or "Slave". If there is only one TPA3116 chip, it has to operate as "Master".
At first I thought it was set to slave 26db as are the same resistors but inverted.
R4 is indeed 010 and R5 is 393. R1 and R2 are both 222 (2.2kOhm), which is not in the table you linked.
Just as in Jareds amplifier. Also for you, the solution should be to remove R4 as I explained to Jared. It is very sloppy of the manufacturer to insert wrong resistor values when the guidance in the datasheet is so unambiguous. If you remove R4, please let me know if it solves your hiss problem. Perhaps you can remove the 4K7 resistors afterward.
I removed the R4 resistor and...no more hiss. Absolute silence. I then removed the 4k7, but that brought back some of the hiss, so I reattached them. Now the max volume is a little lacking (below my usual listening volume), probably due to the lower gain? Still, I'm very happy I got this far, and can't say enough how grateful I am for the help I (we) have been getting in this thread.
Many thanks for reporting back. It tells other members with the same problem that this is a solution.
It is a solution only if the amplifier operates correctly Master/Slave-wise. In case of Slave operation when it should be Master, the amplifier operates poorly in general due to the much lower carrier frequency.
finally my college exam is over.
so, a couple days ago i use my tpa3116d2 board, i noticed a dc offset in the right channel. rising from 0v to a certain voltage until it trigger the fault mode (protect mode) -> fault mode turn on -> dc offset comes down to a certain voltage -> fault mode turn off -> dc offset rising again -> and it keeps repeating like that, that's my thought.
the wierd thing is, the more i use the board (not at one time, "i use it, turn it off and use it in another day" kinda like that) the slower dc offset rising.
and then yesterday the dc offset still rising until it stop rising around 4.1~4.2 volt. the highest dc offset was 7.5~8 volt.
and the more wierd thing is, the dc offset comes up and down if i playing with the input trimmpot, the dc offset is lowered 150~250 milivolt if the input trimmpot at around 50%.
any thoughts?
thanks.
Could it be a leaking capacitor on one of the two inputs for the right channel? Are the input capacitors ceramic, electrolytic or foil?
yeah, it could be. but i've checked all the capacitor in the front (near the input side) of the amp/chip a couple weeks ago and it seems nothing wrong about it.
I think, i need to re-check that again.
I'm not quite sure about what is the type of the capacitor, but it's all in smd parts, except for the output filter caps (i think it's a "trough hole" film caps, red colored) and the supply filter (trough hole electrolytic caps).
any thoughts?
thanks.
I actually did get back to measuring this board balanced in and out while playing around with a new sound card. Compared to unbalanced as in the first measurement I did, it cancels out a lot of the even harmonics with balanced setup. Still the composition of the distortion spectrum is varying a lot when the amplitude is changed.
It is not that the capacitors are open or shorted. All you need to create DC-drift is uA leakage currents.
Did you check the values of the gain-master/slave setting resistors?