Asynchronous I2S FIFO project, an ultimate weapon to fight the jitter

[andrea_mori]

Demian,

the price of the OCXO is high, probably cheaper than a Wenzel-Croven but they ask 290 GBP for a custom part, excluding VAT and shipping. Maybe a group buy of at least 5 people could decrease the price, but don't expect great reduction, OCXO are anyway not cheap.

BTW, as I said in a previous post, I'm working on a cheaper alternative, using a high performance cold welded AT-cut crystal.
The prototype works on the breadboard, but I have to build a wired prototype (or better an air wired) to go to a university lab and test its phase noise (again, simple and quick with an Agilent phase noise measurement system and the professor who know how to use it).
I expect good results, but not the same of the OCXO. An AT-cut crystal does not reach the Q level of an SC-cut, but I believe a bit better than a Crystek or a Tentlabs But in this case also, the crystal only is more expensive than the previous finished XO.
If the results will be good as I expect, I'll develop the PCB and then I'll share all the information (schematic, PCB, measurements and so on).
Unfortunately I am not very productive, since my job leaves me just a little free time (congratulations to Ian who is really very efficient).
If anyone was interested, please be patience, a pair of months I hope.

[qusp]

surely OCXO measurement and quick dont go hand in hand ever, wont best performance be reached after a significant warm up period?

[andrea_mori]

warm up < 8 minutes, that means quick, also for those like me who have little time to spend on their hobbies.
Don't care about long term stability, since we are talking about audio application.

[qusp]

Originally Posted by andrea_mori

warm up < 8 minutes, that means quick, also for those like me who have little time to spend on their hobbies.
Don't care about long term stability, since we are talking about audio application.

OK, it was an honest question, some OCXO take much longer to become stable to the point where they stop improving and then the PLL needs to become stable and the FFT takes time, not just after the element comes up to temp. still sounds like 30mins-1hr for a professional, who is being paid? may not be cheap, but again if they discount their time or are simply interested in the result, or you do not count time and the 20K diagnostics gear doesnt have to be paid for, then sure, all is easy, quick and free.

[nvduybom]

With Jriver 17 in Output mode is Drirect Sound or Wave out, my lossless system play normally but trouble with other Output mode, ex : Kernel Streaming, WASAPI output & WASAPI output-Event Style. Jriver 17 still playing but no sound and I see all the LEDs flashing in the DualXO clock board. Does anyone have this case ?. This error also appears when I use the Empirical Off-ramp 3.

[hochopeper]

Ian,

I've updated wiki page to include GBIV docs. Most significant changes to the wiki are:

Si570, FIFO isol board, battery management and TPS7A4700 are no longer 'under development'

Links to the new docs provided for each of the above are now included in wiki.


Cheers,
Chris

[iancanada]

Originally Posted by hochopeper

Ian,

I've updated wiki page to include GBIV docs. Most significant changes to the wiki are:

Si570, FIFO isol board, battery management and TPS7A4700 are no longer 'under development'

Links to the new docs provided for each of the above are now included in wiki.


Cheers,
Chris

Hi Chris,

You did great job to the community. Others can save a lot of time from your summarized information. I'm highly appreciated. Thank you so much. Just hope everybody could realize that.

Cheers,

Ian

[iancanada]

Originally Posted by nvduybom

With Jriver 17 in Output mode is Drirect Sound or Wave out, my lossless system play normally but trouble with other Output mode, ex : Kernel Streaming, WASAPI output & WASAPI output-Event Style. Jriver 17 still playing but no sound and I see all the LEDs flashing in the DualXO clock board. Does anyone have this case ?. This error also appears when I use the Empirical Off-ramp 3.

Do you have a oscilloscope? Can you find out what's wrong with the I2S signal?

In your normal playing, does dual xo clock board switching between 44.1K and 96K without problem?

Ian

[necplusultra]

concerning the PCB of TPS7A4700

I tried to solder 2 of my prototype boards but they do not work. It may be because of the solder or other reason.

Is there a macro picture so that we can see the chip's orientation.

regards.

[hochopeper]

The pic Ian attached to the GBIV is probably the best one for it, there is a dot on the top corner of the TPS chip.

Note there are a few subtle changes to the layout of the board below, the chip orientation doesn't appear to be one of them though

At a glance the changes are something like: Extra output cap for 6 x10uF rather than 5x10uF; change of position for LED+resistor; Larger value input caps (now total of 200uF rather than 100uF total previously); there may be a few other tweaks Ian has done too, regardless, I don't think the prototype TPS7A regs had any inherent problems.