Check voltages across the ceramics as well.

Well, so far I have traced voltage from analog outputs of the ESS chips (there are two, for left and right channel) to the nearest ceramic caps located above it to the right. For the left channel there is 2.2V coming from the chip pin to the cap and after the cap there is approx. 5V. The same applies for the right channel with a difference that after the cap there is 12V. Something is producing this huge voltage spike and I cannot yet figure out what. It doesn't help that I have to measure the card while inserted to a motherboard. Would be much easier if I had extension cable for it.

Paar wrote on 2020-12-12, 12:22:

Hi everyone. I have a problem with my ES1868 based sound card, mainly with it's amplifier part. The right channel is crackling all the time, but only when amplified. When the card is set to LINE OUT, both channels work fine. I have replaced the amplifier chips with new ones but the problem persists (chips are two LM386N1). When there is only right amplifier installed the channel works fine but when there are both, the left is fine but the right crackles. Made some measurements - left amplifier outputs approx. 6V but the right one outputs 12V.

12V output is no good. The amplifier is driven by a +12V supply, so with 12V it's driving the output hard agains the positive rail. So somehow the right amplifier is overdriven. As you only have the problem when the left amplifier is installed, it is possible that cause of the problem is leakage from the output of the left amplifier into some pin of the right amplifier. With the left amplifier missing, its output is open, maybe grounded through the decoupling cap and the speaker (if connected). With the left amplifier installed, the left channel output is at 6V, which can contribute a very high DC bias for a line-level input signal.

Paar wrote on 2020-12-12, 12:22:

The weird thing is that bot GAIN pins on the left amplifier have approx. 1.3V but with the right amplifier one of the chips has 5V-7V and the second one has around 3V and is changing constantly. Is it some voltage noise?

1.3V on both gain pins seems correct while the amplifier is idling. If the GAIN pins are at such a high voltage, it indicates that the input stage is completely off its designed operating point. Please also check input voltages to the right amplifier. Both input pins should be around 10 to 15 millivolts if the amplifier is used in the typical application circuit. If the inputs are correct, you need to check for leakage into the BIAS pins. If an input pins carries a voltage way above 500mV, something is wrong at the input pins.

I've made some new measurements and now I am really confused. All voltages now seem fine on both amplifiers, even though I have done nothing to the card since the last testing. Both channel work in full volume even though there is still some noise on the right channel, sounds similiar to the noise of wind blowing to a mic. It disappears after awhile, randomly gets back. I should be glad that the card is closer to a working state but I'm still sure there is something wrong with it and that the previous problems will come back eventually.

What I want to try is to desolder speaker out jack and check if there is no solder crack as I have encountered that before. Input voltages on the main ESS chip are fine, left amplifier outputs 6V and the right 5V. Both GAIN pins on both amplifiers have 1.3V now. I have not been able to measure INPUT pins yet as I have not enough room to work with.

EDIT: Well, it seems like a success! I have reflowed some coils but that didn't change anything. Then I have reflowed voltage regulator on the bottom, the connector for 3D sound expansion (the top most pins on the card near the upper amplifier) and speaker out pins and now the card works flawlessly. Seems like it was bad solder joint after all. Thanks for help!