Dropouts have been seen to be related to DPLL bandwidh (too low causes dropouts) and "cold" dac. It takes 15 minutes to 1 hour for the DAC to "warm up".

The input voltage is further regulated down by the local (trident) regulators

Hi,

Seems I have the same problem now with my BIII setup. Using the default 5.2V gives a lot of clicks and sputters. When I touch the chip body of the comparator IC (not any pins!), the clicks are getting much worse and then the sound is gone. I have this with all sample rates, 192k is the worst. Only input 1 is connected to the 4 input SPDIF board ("4-Channel S/PDIF (or AES-EBU) Level Converter Kit"). The board is connected to the BIII with a 5cm (2 inch...) long flat cable.

But (and thanks for the tip!) when I lower the voltage below 4.8V, then the DAC works perfectly, all the way up to 192k/24. No clicks at all.

Tried 4 different DVD players and 2 computers (I have a company...).
Some are better than others, but all have at least some clicks at 5.2V, but even the 'worst' source is working perfectly at 4.7V.

As experiment, I connected the S/PDIF Transceiver Module to the BIII trough a Sidecar (I2S) and that works great with all my sources at all sample rates. Seems the BIII is OK.

Seems there is some kind of problem with the 4 input board. What can I try? Or just let it run at 4.7V?

Thanks for reading.

Problem kept coming back and got worse. Removing all inputs to the BIII except no.1 helped. Turned out the 1.2V from the regulator was closer to 1.16V in reality. Correcting this helped against the occasional ticks in the sound at 192k.

But now the input stopped working altogether. Think the IC on the input board might be bad now after too many experiments from my side.

Tried the Toslink module on input 1 and this works ok (even 192k/24 worked, although not recommended), no dropouts if someone switches on the light etc. But Toslink is not a good solution for high-end audio.

Now use XMOS I2S USB module with DSD f/w. Works great, but of course no SPDIF input then.

Think the best solution will be a combination of remapping of the input (so I2S uses different pins compared to SPDIF) and a SPDIF to TTL design that is more robust against power spikes and other interference. I'm sure the original design of TPA works fine for many people, but my environment has lot of interference apparently.

Will try that later and let you know the results if you're interested.