TOSLINK, S/PDIF input selector design help

Jump to Latest
#23
The DAC101C is added to the schematic, in addition there are some minor changes like pull down resistors and so on. The Board is 4 Layer, 3rd Layer is gnd-plane, 2nd Layer has a copper region for +3.3V. Should i use metal fill on the top an bottom layer? Some other things to remark or should i go to production?

1740994359871.png
1740994436075.png
1740994521739.png


Michael
 
#24
Why 4 layer? This is s.l.o.w. electronics.

Why 2 caps in series for DA101C?

Suggestion is to adapt to 25 years established SMD technology parts. Soldering is not difficult at all of large parts like 0805, SOIC etc.
 
Last edited:
#26
FWIW I only do manual routing and this non complex device can be on single layer even but 2 is standard. It is up to you. Not familiar with Teensy but does it maybe require the same DA101C in between for galvanic isolation?

I would give the Toslink receivers a separate 3.3V uLDO in SOT23-3 or SOT23-5. Had a few issues in the past with certain Toslink receivers that polluted supply voltage (forgot brand/type). Extra cost is negligible.
 
Last edited:
#27
jean-paul said:
Not familiar with Teensy but does it maybe require the same DA101C in between for galvanic isolation?
Click to expand...
Not sure, i had tested the spdif_in to I2S_out with a arylic spdif input board that i have in stock. It was working well. But the board for the teensy is still "to create", booth will have the same power supply. At the coaxial input i want use DA101C for electrical separation between the source and the thing I am building. What is the advantage of seperation between ic output an the teensy, can you explain, please.

With the seperated LDOs, i will follow your expertise and will try to add. This means also with decoupling Cs? I am a little bit scared, handling SMD, but i will have 4 trials, number 5 will be bridged😉.
 
#29
Nothing special. Practical approach fed by prototyping. No simulation, no numbers game, just stabile devices that do only 1 thing very good for a longer time without requiring users choices (except on/off).

Dinosaur design 🙂 My concrete mixers mix concrete.
 
Last edited:
#33
A regulator for each Toslink receiver is a bit overdone. One regulator for 4 x Toslink with each a coil and cap is OK.

Probably TS1117-3.3 is well decoupled with 10 uF electrolytic/tantalum (choose 16V versions) caps both at input and output. Don’t use ceramic caps in that value.
 
Last edited:
#37
1117 family is not always tolerant to the ceramic at the output, there is a risk of self oscillation.
Very depend of a manufacturer, for example - Diode's AP1117 may work with ceramic cap at the output without any problem, but AMS1117 - better to use with tantalum.

Alex.
 
Last edited:
#40
Thanks for the inputs, and i have some more questions. From handling point of view, i would prefer SOT23-3 or through hole, i think to solder SOT23-5 reliable is a competition with the equipment i can use, means TS or AP 1117.

One question is, what will be difference of the outcome, between the suggested LDOs. Is it audible in this application? If not, i would prefer the most reliable and practical solution.

How do you know, which combination of ldo and capacitors works well and which not? is there a rule or some something like that, or is it experience?

Michael
 
You must log in or register to reply here.
Top Bottom