line to optical ---> optical to analog

[dirtbiker1824]

I'm sure this issue has been talked about before but I'm new to this (as of today) and I'm not sure what exactly to search for.
I made a small 2ch 40W/channel amplifier using the TDA7376B IC. It accepts ballanced or unballanecd audio signals. I'd like to make a circuit to take 3.5mm jack, 2 channel, and convert it to an optical signal then on the other end take that optical signal and covert it to a 2channel analog signal that the TDA7376B can use. I have to make over 80' runs and balanced audi signals and unbalanced audio signals are too noisy. I think the blanced noise is comming from my poorly constructed op amp circuit. But in any case I have pleanty of fiber optic cable so I want to go the fiber rout.

I have a reciever that takes the optical signal and coverts it to electrcial and I know I will need an DAC but i'm not sure what specfic IC's I will need. I saw some S/PDIF codec chips, do I need this? Also the optical decoder circuit will be on the same board as the amp so I don't have to run ballacned signals to the chip so what ever is easiest to cover to (bal / unbal) is what I'd like to go with.

Any and all help would be a apriciated.

[twest820]

SPDIF, whether optical or coax, would be the most straightforward option but the cable length's generally limited to 5 meters and 10 meters is the max allowed by the spec. You could look into wireless, sending I2S via LVDS over cat5e or cat6, or even SPDIF via LVDS. But it might be easier to improve the differential op amp receiver---what are you measuring for CMRR? 0.1% or better chip resistors are cheap so if you don't have a tight layout and aren't seeing 60dB or higher CMRR that's where I'd start (I'm assuming you've measured the ground offset between the two ends of the run and verified it's small enough clipping's not occuring). If the mic cable you're using isn't shielded twisted pair changing to that might help too (usually it's shielded untwisted pair).

[Pano]

I agree that it may be easier to improve the balanced analog signal. And then you can use CAT5 if the budget does not allow "fancy" balanced cable.

[dirtbiker1824]

Thank you very much for your reply. I am not as well versed in this mater as you might have though. I don't have an o-scope or any way to measure any kind of gains or losses I just have a basic multi meter. I was under the impression that optical signals could be transmitted much much further than 10m.

I would like to send my audio signals over a Cat5 cable so maybe I should just work on my balanced to unbalanced circuit. Do you have any good sites you could refer me to for some circuit layouts? I was using the OPA2134PA op amps and just some basic 5% resistors and a .1uF poly film cap. Also my power supply is just a wall wart with two 1000uF caps between each the + & gnd, and - & gnd.

Again thank you very much for helping.

[theAnonymous1]

Sounds like you need to make a decent power supply for starters. You should at least think about adding a virtual ground circuit to your wallwart.

Some things worth reading.....

Op-Amp Power Supply Quality Considerations
Virtual Ground Circuits

After sorting out a useable power supply, you should try a circuit using proper balanced line drivers/receivers like the DRV134 and INA134/INA137.

[twest820]

SPDIF is not a long haul fiber protocol. Nor does it come with a long haul price tag.

If you're getting noise problems at distance but not with shorter interconnects that fairly well rules out the supply. TI has a good set of app notes on fully differential op amps that usually top search results. You can also look at pretty much any DAC datasheet or DAC eval board datasheet to see reasonable examples of single ended and differential output drivers---Cirrus and Analog are good starting points. You'll also get much higher quality feedback here on DIYA if you post at least a schematic, preferably the layout as well.

You can measure more than you might think with a DMM---for example, at 5% you can measure the resistors and calcuate the CMRR even if you don't have a voltage source available to measure it directly. But getting set up with a scope or an RTA will make it easier to see if the common mode input range of the TDA7376 is being exceeded.