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

Disabled Account
Joined 2002
Hello,
So once the two caps in series have been charged from a completely drained condition. The balancing circuit should make the caps work like identical twins no matter how many charging and discharging cycles there will be during their lifespan.
Of course not all applications will give the same number of cycles. But if you use them for your Italian clocks 24/7 , with 6? Caps in series better use them according to the Canadian master. Greetings Eduard
 
Before you start the caps fully drained, measure the caps first.
Otherwise you (can) make the same fault with the low voltage.
First make sure what is the problem what you have now.

Regards,
Rudy
tried a different power supply, same problem, caps are drained.
What I also don't understand is that the overvoltage LED just stays on for another 10 minutes if there is no voltage on the input.

will measure the voltage separately from the caps on saturday.
 
Disabled Account
Joined 2002
Hello,
I got my parcel from Ian a few days ago.
Today i got some Eaton coke cans from Mouser.
Plus the transformer to make a choke input supply followed by a 5 volt Guido Tent shunt for the 5 volt Q7 side.
I have the purepi for the Raspberry pi. This one will get choke input too followed by a regulated supply made by Doede Douma who makes things that work for most people.
I hope i won't run into trouble with my supercaps supplies after studying the manuals. Will do it on a weekend without any hurry.
Greetings,Eduard
 
  • Like
Reactions: 2 users
Disabled Account
Joined 2002
Hello,
This thread is less popular than the group buy thread while people being busy buying new boards should actually spend a few hours here on this thread. After all you want to end up with things that are actually working.
It is true that some " issues " appear because people misunderstood or not carefully read the bunch of manuals.
I remember when building the dddac there was much more going on amongst readers / owners on the dddac thread and Doede would usually share his thoughts very frequently.
Of course the dddac thread was more about installing your own modifications before the prints got updated with all the worthwhile improvements. It was a motherboard with one or more identical little boards on top. There were some power supplies that could be improved if you were willing to spend 100 % more. Output caps or transformers, load resistors different people different ideas.
Much more fun because usually it was about improving something that was already good. Here it is like people struggling to make it work in the first place.
Greetings,Eduard
 
  • Like
Reactions: 1 user
Hi,

Would the parts below be the right set for the currently best possible DAC setup without power supplies using Ian's parts? Would adding ReClockPi make any noticeable difference?

#1D. FifoPi Q7
#4A. 9038Q2MPi
#23C. ShieldPi Pro MkII
#35B. StationPi Pro Fully finished

I'm a transient expat here, so please take my comments with a grain / pinch of salt. Once my Summer Project is over, I'm likely done but will try to drop in a few times a year for a pit stop. My comments are trust, but verify since I'm rusty.

There have been multiple reports that there is absolutely zero need for a ReClockPi. So it's a easy pass / punt. A less is more scenario.

The only reason I can see a need for a ReClockPi+Q7 is the tallness of the SC-Pure clocks. Since the ReClockPi acts like a ShieldPi+GPIO Extender combo, it will give the proper headroom for SC-Pure clocks. But otherwise, big 'No', ReClockPi should make zero noticeable difference as commented by Ian and others. Of course, the comments were YMMV in nature but in theory I think it's stated no difference.
 
I still love my Flagship LiFePO4 power supply for my State of the Art clocks @16.5V. Still going strong and can be easily re-configured and repurposed in the future. Solid Core Silver 18G w/ a Switchcraft 2.5mm DC plug. That's why the cable stays up on it's own in mid-air, it's stiff. Noise getting into traditional wiring via Air is a thing. There are reports in theory that transformers can transfer noise into Ian power supplies via tradtional wiring via the Air. Solid core is impenetrable so I doubt noise can get into these wiring. That's why I stick with battle-tested Medical Grade Mean Well DC to run Ian power supplies + Solid Core Wiring.

bat.jpg

My Summer Project is mostly complete. The max size for USB cables is 20 gauge for traditional wiring. I tried my best to implement 20 gauge solid core wiring, but eventually settled for 22G. So one USB cable down, another one to go and then I think I'm set. I'm comfortable now with making 22G Solid Core Cooper USB cables. The following is 18G Silver (Blue Wire) for Q7 clean side @3.3V. I'll try to take photos of USB Solid Core in a week once I wrap up. Left side is UcPure MKI w/ UcBalancer @3.3V. Right side is UcPure MKII w/ UcBalancer @5V. Ignore the blue wiring, not hooked up to anything yet. No connection between the UcPures:

ultra.jpg

I'll eventually with to JST Solid Core cabling for the rivets at the bottom of the UcBalancer.
 
Last edited:
i did a quick measurement, cap H is 0.524 volts cap L is 2.669 volts.

so cap H is not charged and cap L (over voltage) is normal.
so there is an imbalance.

I will unload them both and then do the first start procedure again.
Something is very wrong there. Please share pictures showing all your wiring and PSU/Transformer user for charging. I have a few UcPures here with no issues so will try to help you. To fully discharge the caps I use an 8ohm 100 watt resistor connected to the always on UcPure output J10
 
Disabled Account
Joined 2002
Hello,,
The problem is happening with an ucconditioner.
But just like an ucpure there are two caps iin series and both boards contain some circuitry to keep the caps at equal voltage and to keep things working safely.
Today i showed the datasheet of the 3000F . He told me to wear safety glasses and no jewelry during handling. These old mechanics/electricians always have impressive stories about batterie accidents..
These 3000F caps are not like a popping to low grade cap on a pc motherboard
So better wait untill the problem is fully investigated
Greetings Eduard
 
  • Like
Reactions: 1 user
I have flashbacks of Ian mentioning decent powered soldering iron and double-cooper anytime UcConditioner is mentioned.

Could be Mr. Obvious here, but that's how far I would backtrack. I believe I upgraded the soldering iron for UC boards back in the day due to double-thick cooper boards. So I would double-check the soldering on the UC that's not filling up.

Double thickness 2oz PCB copper layers especially designed for high current and ultra-low ESR
applications.

Please make sure using solder iron with 80W or higher power for durable connections of this high current
conditioner board.

I think draining only applies to UcPure, not UcConditioner, but I never had to troubleshoot the UcConditioner. I believe Ian recently mentioned draining shortly before UcBalancer was available when unbalanced Ultra Caps was a thing.
 
Canadian master

I've been enjoying a early 2000s Television show based in Toronto from our Canadian masters called "The Border" here:

https://watch.plex.tv/live-tv/channel/ion-streaming

It shows up a few times a week for about six consecutive hours. If the timing is right, I'll try to remind you to try to catch it.

The television show is based on Canadian Immigration and Customs, so if my Ian package is stuck in Canadian Customs, I know where to go.
 
Hello guys!
Just wanted to share my build using Ian's toys, but not only. Started doing it almost two years ago..

On a StationPi I have put 2 x RPIs, FifoPI Q7, SinePi. Plus a custom PCM1794A DAC.

Now let's break it down:D:

RPI:
I'm using 2 x RPI 4 running symphonic-mpd software in AoE config. Both RPIs have upgraded clocks on them. Power for RPIs is supplied by AMB Sigma 11 5v + 2 x 2.7v ultracaps for filtering. Power for RPI clocks is supplied by 4 x LTO 2.8v batteries in 2S-2P config + 2 x 2.7v ultracaps for filtering.

FifoPI Q7:
3.3v/5v is supplied by UcPures 3.3v/5v

SinePi:
clocks are Andrea Mori's TWTMC-DRIXO 11/12mhz. Power for clocks is supplied by UcPure 15v.

custom DAC:
PCM1794A based DAC running in NOS mode with passive I/V stage using Ivan Byiak's D60 transformers and balanced output. Currently using 2 chips per channel. DAC requires 3 different voltage rails, 3.3v/5v/8v. All of them are supplied by UcPures.

Everything is put together on Ikea bamboo chopping boards. Neotech UP-OCC wire used all around. Samsung tablet as screen. Sounds wonderful since a while back.(y)

Next steps would be to finish DAC part as it's in testing phase at the moment, final version should have 10 dac chips per channel. And maybe try an active I/V stage. What about a case for the whole thing?:unsure:

Technical advices from my friend Toby, Ivan Biyak, IanCanada and Doede helped a lot! Similarities were obvious, DAC part is inpired by the DDDAC.:geek:
Designing pcbs for DAC would not have been possible without my buddy, Vlad.:coffee:

IMG_20230430_232659_208.jpg 20230501_190845.jpg 20230511_194319.jpg 20230607_172213.jpg
 
Last edited:
  • Like
Reactions: 7 users
I have flashbacks of Ian mentioning decent powered soldering iron and double-cooper anytime UcConditioner is mentioned.

Could be Mr. Obvious here, but that's how far I would backtrack. I believe I upgraded the soldering iron for UC boards back in the day due to double-thick cooper boards. So I would double-check the soldering on the UC that's not filling up.





I think draining only applies to UcPure, not UcConditioner, but I never had to troubleshoot the UcConditioner. I believe Ian recently mentioned draining shortly before UcBalancer was available when unbalanced Ultra Caps was a thing.
thanks for the tip.

I immediately inspected the soldering points off the caps and as it turned out, one of them was not properly soldered despite the fact that it seemed that it was firmly attached.

I also bought an extra 80 watt soldering iron for this.

all soldering points went through again and hoppa it works well again.

thanks again:)(y)
 
  • Like
Reactions: 1 users
Disabled Account
Joined 2002
Hello,
VERY probably there are people who decided to not buy the right soldering equipment because they will only need to make a few soldering joints. Unfortunately you only need a single bad joint to get into trouble. There are some joints out there that could work so so for some time and then suddenly problems will appear when the chassis is well tucked away.
That is why Ian wrote a manual.
 
  • Like
Reactions: 1 user
UcPure mk-III- Eaton 2.7v. 3000 Farad

2 more weeks since last listening impression, and I must say that the ultracap psu obviously has settled down.
I write theese words as I am playing right now, and the realism is spooky. The musicians are in the room with me.

Bass is deep and tight. Everything has take another step up the ladder to hifi nirvana.
I have only one UcPure, and that is for the Accusilicon 338 clocks on FiFoPi-Q7.
I strongly recommend Ian’s UcPure psu to all who is considering buying one.

As I have said before, Ian is a gift to the diy community.
 
  • Like
Reactions: 2 users
Hello,
VERY probably there are people who decided to not buy the right soldering equipment because they will only need to make a few soldering joints. Unfortunately you only need a single bad joint to get into trouble. There are some joints out there that could work so so for some time and then suddenly problems will appear when the chassis is well tucked away.
That is why Ian wrote a manual.
it's not that I didn't have the right soldering iron, but a soldering tip didn't flow properly ...... was my own fault. but it works again.