trying to pull I2s from my Sa7310 chip directly to offboard Peter Daniels TDA 1543 and had some questions. If i remove the onboard chip and route the wires from vacated holes to my offboard chip, do i need to make any connections with the open holes on main board? Everything is wired properly and should be working, but nothing. THe DAC has its own PSU so i did not attach ground of seperate psu to ground pin on main board, assuming it was unnecessary, but have I broken a necessary link for the I2s lines. Normally they would be routed to ground through TDA chip present onboard.
Bck, ws, and data are going to pins 1,2, and 3 resoectively. The question is whether the ground pin needs to be connected to the ground plain of the source of this lines versus what is now its own floating supply. The internal diagram shows a ground reference, which i have, but it is referenced to battery powered regulator and therefore not connected in anyway to main i2s signal board except for three i2s connections
The signals require a return current path, I2S is realy designed for use on a single PCB with one contigous ground. How are the return currents going to get back to the source without a GND connection. The best way would be to have a gnd return for each signal kept as close as possible, both through the wire and where they connect on the board.
Otherwise your signals are going to be messy, especialy the clock signal.
Otherwise your signals are going to be messy, especialy the clock signal.
None taken. Onboard, all pins had a reference to the ground plane through pin 4. With that chip pulled, i am running wires from onboard holes for pin 1,2,3 and 4. Now the offboard chip has ground reference to main board, with all pins having access to this point through pin 4. All should be working correct? Having a standalone regulated chip supply using a battery supply should not be causing any problems. So why an error signal? Is the fact that the now open pin 8 onboard is without a ground reference a problem?
Ouch...
Have a look at this, it shows how the signal and its return path are closely related. For digital signals this close coupling between signal and return is critical, for clock signals it very very critical. Puting clock signals down wires is not good practice (and probably induces more jitter than an on board clock, but thats another story).
http://www.x2y.com/filters/TechDay0...log_Designs_Demand_GoodPCBLayouts _JohnWu.pdf
I would try doing it with either coax for each signal, with its own return (approx 50ohm impedence) or twisted pair for each signal, so that each signal has its own return path closely coupled to the signal.
Have a look at this, it shows how the signal and its return path are closely related. For digital signals this close coupling between signal and return is critical, for clock signals it very very critical. Puting clock signals down wires is not good practice (and probably induces more jitter than an on board clock, but thats another story).
http://www.x2y.com/filters/TechDay0...log_Designs_Demand_GoodPCBLayouts _JohnWu.pdf
I would try doing it with either coax for each signal, with its own return (approx 50ohm impedence) or twisted pair for each signal, so that each signal has its own return path closely coupled to the signal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.