TOSLINK RX all sold out?

Hi,

I'm looking for TOSLINK receiver modules like EAPLRAA4. It seems that everything in this category is sold out when I search DigiKey and OEMstrade.

Any suggestions for RX modules that are available in the hundreds without having to negotiate with traders and parts of unknown origin?

Thanks,
Børge
 
Thank you for the recommendation on Cliff! I was not aware of that vendor. It's harder to be discovered if you're not on Digi-Key.

The only useful Toslink RX module on D-K is the TORX1355(F), of which they have 50 stocked.

I'm prototyping at the moment but want to be able to make a few hundred units eventually.

Børge
 
Hi I experienced more or less the same some time ago. Toshiba does not produce the well known Toslink audio devices anymore since many years after a flood in the factory. So no original Toshiba versions available anymore and these were top quality. There were/are some introductions of new versions but I never was able to order these.

This leaves the clones with quite some differences. For instance that optical cables fall out as they do not comply to the official sizing etc. Then the electrical properties some of them being limited to 24/96 etc.

Then the jitter which is often near 1 ns. It is really better to go coax SPDIF. Measures and sounds better.

If it really has to be Toslink for "convenience" that always wins then make sure to add various pads on the PCB for the various types there are. Some have extra metal pins (better!) for mechanical strength, some have plastic pins at a different spot. Shutter (terrible design), no shutter, 3.3V/5V etc.
 
Last edited:
Toslink is SPDIF! Also add a coax RCA/BNC connector for electrical SPDIF and simply good audio when using a streamer. The difference with modern Toslink versions is often striking. I am experiencing this with Elac DS-S101-G.... Toslink is simply unusable with this device.

The best circuitry and DAC chip/implementation won't suffice to polish away the faults of Toslink clones. Anyway I found a few NOS TOTX147/TORX147 from various sources and swear by these when it must be Toslink.
 
Last edited:
Yes, I'm designing a circuit now where I receive SPDIF with a transformer and comparator, and TOSLINK with whatever acceptable module can be had. Since WM8805 is hard to get, I use my own MUX and a WM8804. I have working code for the '05, so the '04 port should be quite feasible.

While jitter is a b***c, my goal is to have valid data and buffer it for playback.
 
Yes, it is a good idea to use 3 different pinouts 🙂

I searched Digikey for suitable Toshiba Toslink receivers but the available ones seem incompatible with previous versions as they are 5V only and 10 Mbit max. ?! I would not use those but choose 16 Mbit 3.3V versions.

The document by Peufeu regarding SPDIF/WM8804 and best termination of Murata DA102C is attached. Make sure you feed the WM8804/5 with 2 x 3.3V power supplies all well decoupled. I never bothered to use 12 MHz low jitter XO's but I think it was also Peufeu that pointed to the merits of using a 12 MHz low jitter XO. That one will also need a separate 3.3V power supply just like the Toslink receivers and the USB chip! So, in effect, there will be at least 5 x 3.3V regs for the receiving end of the device if you want to do it right at once. I have a few products of brand "Y" on my bench that only use 1 x 3.3V to supply all sections and it is simply not OK certainly when bus powered via USB. A separate PSU is better. SMD coils/beads will improve matters. Please avoid the dreaded LD1117 or its cousins. They are noisy crap still being used by many designers.

So, that's enough for today.
 

Attachments

Last edited:
Hi Jean-Paul,

I found that document in the thread you pointed me to. A good read!

With the '05 sold out and with multiple SPDIF sources, I decided to use the DA102C with a comparator that feeds a mux into the WM8804. The trafo and comparator part of the circuit has been tested OK.

My prototype board now holds footprints for 3x different TOSLINK RX modules....

As for the power distribution network, there is one 3.3V regulator (ADP151) that feeds the whole digital RX side. Each point of load has a well-matched pair of decoupling capacitors and individual serial L-or-short feeding it.

I'm not very concerned with the regenerated MCLK as long as I have correct audio data on the I2S output.
 
The suggestions by Peufeu make things more than OK. I only tested experimental so no PCB but it was a pleasant surprise.

Only 1 x ADP151?! You will experience feedthrough of sorts. In my extensive tests with WM8804 it performs better when having 2 separate low noise power supplies. Costs are minimal with only 1 Euro (!!!) per ADP151 and there is direct improvement. This absolutely counts for XO's and USB IC's that should never have a shared PSU. If "excellent" is the goal that is.
 
Last edited: