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

Jump to Latest
B

bigpandahk

Member
Joined 2009
necplusultra said:
Hi Ian

I tested the isolator with power on clock board at 5v.

the input voltage on fifo side is 3.3v which is good.

the voltage on the clock side is 1.7v which is not good.
so I unsoldered all the regulator circuit on the clcok side and the voltage now reads 2.3v on the L pad of the regulator circuit.

everything else is still soldered in place. just the clock regulator circuit is removed. the L pad should read 5v.

any ideas?
regards.
Click to expand...

Measure the resistance between ground and the resistor array, it should be something higher than 300Kohm, if not, than resolder the resistor. sometime the resistance will drop after the isolator is powered for several minutes. This was my experience and I have to resolder it several before get it work. Use good quality soldering flux is very important.
 
B

bigpandahk

Member
Joined 2009
qusp said:
that wont work, fifo runs asynchronously, even with sync clock out, it will be running with no relation to the amanero board timing, how could it be synchronous? also the fifo needs to choose a samplerate and thus clock, based on the input. with no input it wont do anything, bad enough with the dual clock board, even worse with the Si570 which chooses among many different clock speeds
Click to expand...

I also planning to use the OTTO to bypass the FIFO when playing DSD with the Amanero. If I am running BIII in sync mode with FIFO, Can I use the Arduino to control the clock of the i570 when playing DSD?
 
iancanada

iancanada

Member
Joined 2009
twitter.com/iancanadaTT
Paid Member
ed linssen said:
to Ian and all,

I want to share my initial impressions about using the Spidf-Fifo-Clockboard combination. Right out of the box I connected my SBTouch, which was hardly used in the past months because of unfullfilling soudquality, and it was frm the first second a relevation! It sounded like music, which it did never before.
Not connected to my BIII buffalo, but with the nice U-fl connectors and cabels into I2S inputs of Doede new 1794A Dac.
WaneIO connected to the I2S backdoor of the Spdif-board.For playing my canned CD's through XXHighEndPlayer on my PC. This was already a splendid combi but it improved. The crazy thing is, I am talking about the generic clocks that where provided with the set by Ian, just for checking purpose. I look forward to tell you more after implementation of all the proposed refinements to come. Such as better clocks, (I ordered Chrystek957). And waiting for parts to get the Isolator and the managementboard complete... And Lipo4 batt. for the clock?
I will for shure come back on this.
Happy holydays:hohoho:
Click to expand...

Thanks ed linssen for your feedback. Your result is a good approve of mine testing result. S/PDIF FIFO really improves spdif sources. If you try better clocks, better power supply and cut the ground noise, you will still have some space to improve the sound quality. Yes, the xo from the kit is not that bad, based on my own testing, the rms cycle to cycle jitter was around 5ps. As compare Si570 is 2ps, CCHD957 could go even lower. But I don't have equipment testing the phase noise.

I'm looking forward to your future progress.

Thanks again for sharing the experiences, have a nice holiday too

Ian
 
N

necplusultra

Member
Joined 2011
Greetings

I removed and resoldered the isolator chip. I reinstalled the regulator circuit and all appears to be working. Only issue is that the music turns into distortion but I think it is related to my 5v connection. When I wiggle it slightly, the music returns. I have 2 final questions:

1) what is the digikey part for the 5v connector on the clock board?
2) can I feed the board 5.25v directly from one of the buf3 psu?

kind regards to all that helped.
 
iancanada

iancanada

Member
Joined 2009
twitter.com/iancanadaTT
Paid Member
necplusultra said:
Greetings

I removed and resoldered the isolator chip. I reinstalled the regulator circuit and all appears to be working. Only issue is that the music turns into distortion but I think it is related to my 5v connection. When I wiggle it slightly, the music returns. I have 2 final questions:

1) what is the digikey part for the 5v connector on the clock board?
2) can I feed the board 5.25v directly from one of the buf3 psu?

kind regards to all that helped.
Click to expand...

housing 455-1165-ND
pin 455-1318-1-ND

Already included in the isolator kit of the GBIII

This connector could also be found from many used electronics like dvd player and etc.

Ian
 
A

Adrculda

Member
Joined 2009
What about a SMPS supply ??
I know its noisy, but with a well designed filter section should be able to get clean power from any other source than batteries.

Dont take me wrong batteries are fun but they have a limited shelf life, as most of us will want to keep this piece in our "arsenal" for a long time to come and we need a more definitive way to power this thing
Medical Grade SMPS supplies come to mind which even on digikey are quite affordable

Whats the total current draw for the works ??

If noise is an issue why not house SMPS in a billet enclosure
 
Last edited:
A

Adrculda

Member
Joined 2009
CUi Inc Has 2 very affordable versions, VSKM-S10-5U or VMS-20-5 supplies come to mind which even on digikey are quite affordable. Another option is to go to a slightly higher voltage into a reservoir of caps and use your own regulators...
this way audiophile grade caps and film caps can be used to decouple them from the supply side while a CRC filter will keep noise in check
 
Last edited:
R

regal

Member
Joined 2004
Adrculda said:
What about a SMPS supply ??
I know its noisy, but with a well designed filter section should be able to get clean power from any other source than batteries.
Click to expand...

As much as I like LiFePO4 batteries I don't see the advantage of driving a board connected to a 600mhz computer with a battery, probably a SMPS would work fine? Unless the batteries are just for the reclock past the isolators but I think even those pass a lot ofhigh frequency stuff, I mean they have to pass the clock speed right? So really what are they doing? Sorry just feel like playing devils advocate today cause we didn't get our pcm boards before Christmas ;)
 
C

crobbins5421

Member
Joined 2007
Battery Supply Question...

Got the boards Ian. Thank you. They look very good.

I'm especially grateful this Christmas that you did "the heavy lifting" and had the resistors on the isolator board pre-soldered....

After reading through the thread and the documentation on the FIFO and Dual XO, I saw where you give a recommendation for battery power for the FIFO. I was wondering...should I choose to not remove the two LDO's on the XO board (implementing the isolator) what would be your recommendation for reasonable LiFePo4 battery power to the XO board?

if I missed that recommendation earlier I apologize.
 
M

MisterRogers

Member
Joined 2011
Hey Ian. I'm working to install the Isolator board. I tested it with double clock board powered by FIFO, and it's working fine. I'm working to power it directly now. According to docs, I removed L11. 44.1 & 88.2 leds light up now, but now I'm not getting lock or sync. I'm digging through the thread, but what am I missing?
 
B

Blitz

Member
Joined 2002
1. bigpandahk
2. tagheuer
3. hochopeper
4. qusp (of course)
5. AR2 - definitely!
6. wktk_smile
7. hirez69
8. CeeVee - you bet!
9. number9
10. analog_sa - GB maniac
11. edbk
12. atom6422
13. misterrogers X2 - Of Course!
14. NicMac - as usual!
15. Zoran
16. PET-240
17. Coolhead
18. Slartibartfasst
19. SYklab
20. Regland
21. Neb001
22. SPWONG
23. Greg Stewart (also of course!)
24. Vitalica
25. spm
26. Fridrik
27. ccliu
28. makumba1966
29. lindamar
30. Finaxe
31. Odysseas x2
32. palmito
33. crazikid
34. deanoUK
35. Julf
36. DUC985
37. rsotirov
38. kvl
39. bkdog
40. necplusultra
41. Nikola Krivorov X2
42. nvduybom
43. Popolvar
44. TNT
45. TUBO
46. Blitz
 
M

MisterRogers

Member
Joined 2011
Hey Ian - thanks for the followup. I removed the Isolator and all was well, so it's clear I have an issue there. It worked fine, so I suspect a cold join. That's on my chore list for this afternoon - I love having a break, with plenty of time to play :)

Also, I build a couple TPS7A4700's at 3.5v this morning for use as AVCC regs in my BIII. Interesting :) Not sure what I expected; I'm running Paul Hynes there so it's already pretty damn good. Just a short listen - seems to have a touch more bottom end clarity and more high end sparkle. Thoughts on what to expect?
 
Last edited:
qusp

qusp

Member
Joined 2009
I really dont think that many of us are interested in SMPS for one of the more vital analogue sections, i'm certainly not and I dont mind them elsewhere. the fifo doesnt use much current really, so linear or shunt low noise regulators are pretty easy to implement, I see no need to push for SMPS

@ MisterRogers:

the Paul Hynes regs are superior, maybe even by an order of magnitude, but certainly significantly
 
Last edited:
You must log in or register to reply here.
Top Bottom