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

Jump to Latest
A

A123

Disabled Account
Joined 2020
yunyun said:
Like my post below, I can provide AS318-B with free international shipping. If more people buy together, the unit price can be lower. $5 you see on the Internet is AS318, it is not AS318-B, the performance gap is very big. 45.158400 and 49.152000, MHz is available for sale, 90.316800 and 98.304000 is not now until 2 months later.
Click to expand...


Hi @yunyun. Thanks. Usually when things are too good to be true, in this case low priced AS318s, I need time to do due diligence on the product to verify that they are likely indeed too good to be true. Appreciate you pointing out they are way different. That it's AS318-B, NOT AS318. Saves me time from the research.

I'll keep your group buy in mind for my next FiFoPi build. I want to focus on a more portable FiFoPi build first, so want to keep things at a minimum footprint and also take minimum risk in the build. Purchasing from Audiophonics will be almost no risk and ease my learning curve when swapping clocks. If things don't work, the problem is human error on my part. I don't want to troubleshoot the clock and second guess my implementation at the same time. Once I get the hang of clock-rolling, I'll be more comfortable group buying clocks for my second build. My second future build I'm planning a separate independent battery-powered RPI -> i2S -> FiFoPi stack powered by LiFePO4 MKIII/MKIV.

So since this will be my first attempt, I have to play it safe this time around. It's also a timing issue since it's unknown when the group buy will transpire. For my second build, the risk and timing is not a factor. I consider your group buy very, very low risk, but Audiophonics is close to zero risk. If the group buy is short 1-2 more posters, I'll try to jump in in the future.

wlowes said:
You are better with 45.49. Clocks typically have less jitter at lower frequency. Lower jitter equals better sound. You could even go down to 22.24 if that is available. Ian has a table in the FIFOpi doc that shows what you can play with each frequency.
Click to expand...

Thanks! That's a solid option. 22.24 would of been amazing. I would agree that suggestion would of been the most optimal for my case. Unfortunately, it doesn't seem 22.24 is produced by looking at the AS318 PDF and Audiophonics list of items for sale. I've noticed 22.24 Crystek on Mouser is available.

I went through the manual a few times, but overlooked the table. I guess I was concerned about how to power at that time. Great table though, it really helped target the range I want to focus on optimizing.

Well, with your confirmation of 45.49 and the helpful table in the manual, I guess it's time to place an order.

JCMcNeil said:
Thanks!

I hope the Accusilicon works well for you. I don't have any experience with it yet but chose it largely on the strength of this review.

I don't feel I need the 41kHz family version because MiniSharc requires (or resamples) everything at 96/24.

Chris
Click to expand...

Thanks, I hope Accusilicon works well for your project too. That's a fantastic article. I made sure to save the web page and save as PDF. I'll be re-reading that article several dozen times. It even mentions that it's a straight drop-in to replace the stock clocks and some troubleshooting comments at the bottom. Couldn't imagine a better article on this niche product mid-2020.

I just love the concept of LiFePO4 powering audio products and clock-rolling (w/ LiFePO4 power) with solid SQ hardware in the RPi ecosystem. Now I have to grasp the concept of taking the RPi out of the equation (ReceiverPi) which I never considered. Mind blown.
 
Last edited:
A

A123

Disabled Account
Joined 2020
yunyun said:
A123,

andrea_mori remind me that AS318B must be more competitive, so the price of 45 and 49MHz is directly cut in half now, only 20 US dollars, including international postage fee.
Click to expand...

Hi @yunyun,

Thanks for the offer, but it's too early to make a decision until I succeed (cross fingers) with the first batch via the Audiophonics.fr order.

The Audiophonics.fr order is already being processed.

Once the AS318B (Audiophonics) + FiFoPi Q3 runs smoothly then I'll be open to the idea of additional AS318B (YY GB). I'll keep an eye out on your AS318B GB thread.

It's just like I'm going to start off with one FiFoPi Q3 first instead of ordering two FiFoPi Q3 to start. I want to focus on just one project at a time to make sure it all goes smoothly before attempting further projects.

Cheers
 
A

A123

Disabled Account
Joined 2020
ATM, I am planning a FiFoPi Stack + LiFePO4 MK III Stack.

This will enable me to experiment with a separate RPi w/ ConditionerPi to connect via i2S to the FiFoPi Stack (RPi-less) w/ 12 inch u.fl cables.

But now, I'm starting to wonder how I can use an almost decade-old NUC to spit out i2S if I add a HDMIPi Receiver into the mix. Do I need a special mini-HDMI to i2S cable or Ethernet to i2S? Can I connect a standard mini-HDMI to HDMI cable direct to HDMIPi Receiver?

I was hoping to power the NUC via MK III then have it output to the FiFoPi Q3 via i2S.

I had horse blinders on for RPi, so now trying to branch out since the FiFoPi Stack allows for so much flexibility into other possibilities.
 
Last edited:
L

linh0983

Member
Joined 2016
acnt
I have i2s splitter as shown :


HDMI I2S LVDS to I2S Input Module HDMI I2S LVDS to I2S Output - Audiophonics



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

acnt
 
H

hosoludi

Member
Joined 2018
Hey @iancanada

I bought a Fifo pi and a Transport pi. And i use a NT-50HQ-09 From Thel audio.
Audiophile Super-Netzteile, High-End Analog Netzteile für DA-Wandler; Raspberry Pi
[FONT=Arial,Helvetica][SIZE=+2][/SIZE][/FONT]
I've installed ropieee on the raspy and use it for roon streaming.
At the moment i use COAX shielded. But im planning to use i2s later with my new DAC
Can you tell me what i can upgrade to get the best possible performence out of my streamer?
Kind Regards

https://www.diyaudio.com/forums/members/iancanada.html
 
J

jimk04

Member
Joined 2018
Paid Member
Another question from myself if I may.

Yesterday was playing around with my 2nd full Ian system. I have acquired some of the Aliexpress output ttansfotmers and so was testing the passive IV stage.

Also new in the system was my diy UC pair I was trying on Fifo 3v3 rail...all working well although I have some uncertainties about the trafo output...but that is another story.
Anyway I hadn't realised the UCs weren't plugged into their LT3042 charger and went flat. I was changing the SD card from Moode to Volumio and it took me a while to realise why I had no music.....'empty ' led and tracking dots on the display.
Recharged the UC for Fifo clean but I still have the above errors. Can I have damaged something running the system whilst it went below usable voltage....and what to test ?

Regards
James
 
J

jimk04

Member
Joined 2018
Paid Member
Some additional diagnosis...so the power pin had snapped off one of the clocks. I have removed the pin from the board and left it unpopulated for now. Only one clock in there at the moment. Confirm psus for clocks are OK as 3v3 present over pins 2 and 4.
Clock is in X01 but the green led for X02 is lit and that is empty. Red empty led lit and tracking dots remain.

I thought maybe the case of the clock has shorted over other pins but I would have thought the little plastic bumps on the bottom of the clock case would prevent that.
 
P

pz1

Member
Joined 2013
jimk04 said:
Some additional diagnosis...so the power pin had snapped off one of the clocks. I have removed the pin from the board and left it unpopulated for now. Only one clock in there at the moment. Confirm psus for clocks are OK as 3v3 present over pins 2 and 4.
Clock is in X01 but the green led for X02 is lit and that is empty. Red empty led lit and tracking dots remain.

I thought maybe the case of the clock has shorted over other pins but I would have thought the little plastic bumps on the bottom of the clock case would prevent that.
Click to expand...




Hi

I have the same problem for a week or 3, suddenly after restarting my gear after a few days. I used a regulator 3,3V and that regulator is still accurate. Also red led empty and 3.3. V over the XO pins.
 
J

janho12345

Member
Joined 2016
Hi,

I have a FifoPi Ultimate (not the newest version, the one before) .
The FifoPi has two(3) i2s outputs.
1. u.Fl connectors
2. J7
3. J2
The number 1. and 2. are both isolated and re-clocked after the isolator (by a not isolated master clock).
Number 3. is non isolated non re-clocked Output.

If one believes that this old wisdom is still true (yes it's been a long time, and much has changed, but probably not the fact that isolators add jitter and a Flip Flop is used to re-clock this):

iancanada said:
NOS DACs such as TAD1541, they don’t use MCLK for the internal timing, instead, they use SCK and WS driving the internal state machine. That means those NOS DACs are sensitive to the jitter on the signals other than the MCLK. In this case, reclock the I2S signal by the same master clock is a very good idea. And after doing that, the total jitter on the I2S signal=master clock jitter + additive jitter of the reclocking flip-flop. At is meaning, just as you said, the intrinsic jitter of FIFO was eliminated. But instead, the new flip-flop jitter is introduced.

On those clock boards I’m cooking right now for my FIFO project, the reclock circuits have already been included. The only thing is we need select the flip-flop with careful. Normally the flip-flops will also introduce quite a few amount of additive jitter. The best flip-flop is the PECL series, which the additive jitter is just around 1 ps level, unbelievable! The normal TTL or LVTTL flip-flop have the similar or little bit worse jitter level than the CPLD on the FIFO board. Some new generation flip-flops, such as AUP, AUC series, have better jitter performance. I’m not recommended using the S,AS,F series TTL IC doing the reclock job, not only because the speed is not as good as the new technologies, but also, the high switching current will introduce big ripple on the power supply which will increase the clock jitter rapidly.
Click to expand...

is it a good option to connect my dam1021 (which already has i2s input isolators on board, and I'm not going to remove them) directly to the non isolated non re-clocked i2s output on J2?

The question in hand is: is the actual FiFo i2s output jitter higher than the jitter after the isolator and re-clocker? If not, it would make sense to use the non isolated out in this particular situation with an dam1021 (with on board isolators and [?eventually?] a re-clocker on board.

The question is, has someone tried this with a dam1021?

Thanks and greetings,
Jan
 
Last edited:
You must log in or register to reply here.
Top Bottom