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

Did You take mechanical interference / loopback into consideration in this project? You need to decouple the clock source (oscillator crystal) from the playback room. Tap on the (or play loud music close to the) crystal and measure the jitter!


You are right. Mechanical viberation is one of the three main sourceses to the close-in phase noise. I designed optional rubber ring suspensions on the corners of all clock boards. But there might be other solutions.

Ian
 
You are right. Mechanical viberation is one of the three main sourceses to the close-in phase noise. I designed optional rubber ring suspensions on the corners of all clock boards. But there might be other solutions.

Ian

I am sure it was mentioned somewhere before, but for a late bloomers.... is there any source for rubber rings that you guys could recommend?
Thank you
 
yeah I know, but I didnt have any medical tubing lying around, but i've got plenty of heatshrink, as does just about any DIYer. havent had any problems since. I also used some fine 10000grit abrasive cloth to get rid of the sharp edges on the PCB where they attach. the heat wont be a problem, just leave the clocks off when you do it, remembering they were probably soldered in a reflow oven
 
that depends, that covers complete ******** sourcery, to wrong, to reasonable, but how exactly were you planning to use that sort of thing inside the case? remember we want to decouple the clock board from the rest of the build, but we want to keep it close to the dac board, which emits vibration of its own along with the PSU. isolating the chassis assuming what you have isnt a silly spike that doesnt actually spread energy at all, you will still be left with at least half the problem.

why bother? the clock board is perfectly set up to use its own mechanism that actually works and costs virtually nothing in parts or time; why would you substitute something of dubious benefit?

but what are you talking about? i'm always wary of things labelled 'audiophile'
 
There are numerous anti vibration products available such as AV mounts fromthe likes or RS and other more bespoke companies. These unfortenatly are not audiophile grade*, but used for such things as mounting control boards and other electrical equipement in tanks, helicopters on the side of generators etc. The advantage of these solutions are that they can be tailored to certain frequency ranges.
*they are designed using science and physics, and work without reams of psudoscientific write ups:D

Having done work with high vibration situations, I would worry at the volume levels required to vibrate a crystal on a board in a case enough to cause problems. its the same with MLCC caps, tapping them does indicate microphonic action (X7Rs) but have found it hard to create the vibration levels in the real world where this becomes a problem. Though if the equipement is mounted on a surface that resonates with the music then you are going to introduce vibration.:D
 
yep for sure, perhaps nattonrice would like to show some pics? hes gone a bit nuts with vibrational decoupling in his dac, built with his usual skill and enthusiasm, but all stuff that will actually reduce vibrations; no Less Loss Blackbody's here. the dac boards are mounted to thick teflon base boards and then mounted with vibration mounts similar to this

An externally hosted image should be here but it was not working when we last tested it.


I actually do think there is something to this given jitter is at its most harmful in the very low frequencies, more worthwhile than some other popular tweaks anyway, when we are shooting for uber low jitter and given the fifo removes so many fun tweaking areas from the playing field.
 
I agree, if you are going all out for the ultimate then go all out. though mounting the boards on standard Av mounts would do the trick, with a very heavy base plate for mechanical mass is reccomended. Even if I personaly think some may be overkill, I'd rather see over engineering for the full on audio system than what you do seeoften on this site, ahalf arsed attemp at the ultimate.
I would like to see a DIY project where everything employs the ultimate engineering, 10-14 layer PCB's AV mounts, etc etc (similar to what i do during the day realy).
The majority of the tweeks I see regarding clocks on this site are more detrimental to the clock signal than leaving the dam thing alone, this is one of the few threads on clocks that does emply engineering and class components, insted of wires everywhere, impedance mismatches and general bodging.
 
I went a little overboard by taking a page out of Erno Borbely's book.
The base board is 6mm teflon, stands are 20mm rubber machine mounts.
This pic is a shot of it all unfinished with missing stuff and looks like junk so no laughing.

Edit: I tried to minimize the general bodging.


dacstack.JPG
 
Last edited:
Haven actually got to install stage for this yet, once I ask around I reckon silicon will be the go, finding the right diameter/circumference will be the challenge. That and mount the external board to small vib mounts, vibration is a force that affects things, even a cd Player on an isolation platform changes when springs are mounted between the cdp and platform.
Tis all very strange.....
 
Haven actually got to install stage for this yet, once I ask around I reckon silicon will be the go, finding the right diameter/circumference will be the challenge. That and mount the external board to small vib mounts, vibration is a force that affects things, even a cd Player on an isolation platform changes when springs are mounted between the cdp and platform.
Tis all very strange.....

a perfect reason for using SSD and fifo, so you dont have to worry about that nonsense. then again I suspect you will just find something else to worry about :p
 
Wow...Thank you guys. Just one small question and two pages of ideas. You guys are awesome. Here is what I found. First I was thinking about some rubber seals for plumbing, than I remembered this place that I use for screws and bolts when needed, They have everything. Official title to search for - O Rings:
McMaster-Carr
What would really help is if anyone has original Ian's rubber rings, to share aprox diameter. Ether ID or OD would work.
 
Last edited:
No brass? Strangely it passes vibrations steel does not, with the lid on, see it you do not.
And qusp, yes I know it's tongue in cheek, but it really does help, even my little Alix board running mpdPup and a WaveIO benefits from damping, springs and whatnot, it runs from ram, and does nothing but serve music, my little TerribleDAK(learning experiment) does the same, this is all solid state, when I get time I'll take a pic and show you, better yet, you can come hear for yourself..... and I got the clocks today, Many Thanks! Though I got two 45. MHz ones so 3 in total? My score!