The mixed clock domains and long wires might be part of the noise problem. Maybe noise on the I2S bus is causing false triggering in the FPGA inputs or something. You could try looking at I2S signals into and out of simple DSD converter with the scope to see if they look like they should. Also check power supply voltages and noise. Other than that it only makes sense to clean up the layout and clocking. Its bad enough right now that it could be causing various problems, or possibly making other problems even worse.
EDIT: Agree with PJotr25 to try removing Chronus. It may be doing more harm than good under the circumstances.
EDIT: Agree with PJotr25 to try removing Chronus. It may be doing more harm than good under the circumstances.
Yes, it's connected Amanero (clock slaved from Cronus 22/24Mhz) to pcm2dsd to Cronus to LPF.
Cronus DSDON output is DCK output?
I guess is better to connect directly with the 20 pin connector the Amanero to the Simple DSD Modulator & by software come back to unslave the Amanero (Slave_1081_DSDSWAPPED e firmware_2006be11 PIN6 MCLK input per selezionare il CLOCK devi cambiare il PID_071A in PID_0a23 e abilitare il selettore PIN 1 o PIN 11) someone remember with soft to flash for unslaved Amanero?
IIRC Chronus DCK is BCLK (data-clock / bit-clock). Easy to check clock frequency with a scope to make sure. D1 and D1 would be either LRCK and DATA for PCM, or else DSD_L and DSD_R for DSD mode. Check with scope if not sure which one (D1 or D2) is LRCK.
EDIT: I would still recommend to clean up layout and wiring. Its way too disorganized for RF. Transformers/inductors are too close to RF, etc. Think of a clean layout in terms of signal flow, keep I2S wires as SHORT AS POSSIBLE. Move boards around to make I2S and MCLK wire lengths very short and direct. Also, twisted pair wires have high characteristic impedance than 50-ohms. Probably closer to 100-ohms. You may need to add some damping resistors (maybe try 50R metal film) in the I2S lines where the impedance in changing to achieve best waveform fidelity. Ground scope probes carefully with short ground leads and use a fast enough scope. Otherwise the scope and probes will distort the waveforms you see so what you see will not be accurate.
EDIT: I would still recommend to clean up layout and wiring. Its way too disorganized for RF. Transformers/inductors are too close to RF, etc. Think of a clean layout in terms of signal flow, keep I2S wires as SHORT AS POSSIBLE. Move boards around to make I2S and MCLK wire lengths very short and direct. Also, twisted pair wires have high characteristic impedance than 50-ohms. Probably closer to 100-ohms. You may need to add some damping resistors (maybe try 50R metal film) in the I2S lines where the impedance in changing to achieve best waveform fidelity. Ground scope probes carefully with short ground leads and use a fast enough scope. Otherwise the scope and probes will distort the waveforms you see so what you see will not be accurate.
Last edited: