I have to make this clear, from the beginning:
We were paid to do this, by another manufacturer. Part of the agreement was that the info is considered "proprietary". So, I can not post the results. What I can do, is give a rough assessment, of our findings. IOW, yes, it will be brief, but is enough that anyone who really cares will have something to go by.
The goal was to find if the new parts, made by Everlight, are anywhere near as good as the old Toshiba TOSLINK parts. Since Everlight is what is available, that is what we measured.
We looked at how much they raised the jitter. Not in terms of "x pSec", but in actual terms of phase noise, which is what we feel is much more important.
We used several low phase noise sources, to run these tests. One supplied a 44 kHz SPDIF signal, another a 192 kHz SPDIF signal, and a 12.288 MHz clock.
We ran these tests on some old TOSLINK parts we had, around the shop. New, and never used. We also used some "regular" Everlight parts, and their "high speed" version.
So, in a nutshell:
All of them raised the noise floor by at least 20 dB. The higher the signal frequency, the more the noise floor went up.
The Toshiba parts were clearly better. (Any surprise?)
The "high speed" Everlights were not horrible, but the gap widened, as the frequency increased.
The "regular" Everlights....................spend the extra few pennies and buy the better ones.
Wish I could go into further detail, but that is all I am able to supply. I hope everyone will understand, and use what info has been supplied to good purpose.
