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

Hello friends of good sound.🙂
Thank you for this thread and I'm drawing information! YouTube @Gabster 2000
https://www.youtube.com/@Gabster1/videos
I recently finished my first raspberry audio player with sw Volumio.
  • Rpi4 4GB RAM
  • IAN CANADA UCPI Universal Ultra-Capacitive Power Supply Board for Raspberry Pi
  • Power: Toroidal Transformer 2x 9V 5.5A
  • Power supply so far with one branch from IAN Canada 5V
  • LCD Display: 7.9-inch capacitive touch LCD display, 400×1280 LCD Waveshare
  • Custom-made aluminum box CNC milling and screen printing. ModuSchop
A new sound card has recently arrived:
- IAN CANADA TRANSPORTPI DIGI II Ultra-Low Jitter Digital Audio Interface Module.
https://www.audiophonics.fr/en/dac-and-interface-modules/ian-canada-transportpi-digi-ii-p-19555.html

After the New Year I will continue to improve my equipment (but there is not enough space there ) 😀
- I want to use: IAN CANADA ISOLATORPI III Galvanic Insulation Module
https://www.audiophonics.fr/en/accessories-and-cases/ian-canada-isolatorpi-iii-p-19222.html
+ IAN CANADA DOP DECODER DoP to Native DSD Decoder Board
https://www.audiophonics.fr/en/dac-...-dop-to-native-dsd-decoder-board-p-17014.html
and I will power it with:
- IAN CANADA LIFEPO4 MINI Battery Power Supply Module 3.3V
https://www.audiophonics.fr/en/batt...-battery-power-supply-module-33v-p-15831.html
tmp-1734814391231_copy.jpg

tmp-1734814391283_copy.jpg

tmp-1734814391311_copy.jpg
 
Last edited:
I have just finished making the Gabster td1 dac, with a hell of a lot of help from Gabby. This was my first soldering project other than fixing pipes and I made tons of blunders, especially when trying to solder those ufl connectors. Anyway, have it up and running now and am frankly very surprised at how well it performs. My previous streaming dac was a Meitner 3a, which is a pricey piece of equipment. While I no longer have it on hand to directly compare to the td1, I clearly am hearing a different sonic experience. The td1 has more attack, the drums especially have more zing to them, are more real sounding. I’ve been listening to favourite tracks for the past few days and the td1 has simply working up my speakers. Female vocals are outstanding. Thank you Gabby.
 
Dear @pkonstantinidis and forum colleagues,

As I promised, I revert with my impressions on TDA1 DAC. Before, I dive in, I need to state the following: the comparison will be made using as benchmark the duo Ian Canada streamer – Holo May KTE DAC. My opinion, as every other opinion, has a level of subjectivity. In my case, because of a sensitivity I have in my ears, I can’t stand harshness, and this is a determinant of my verdict.

Ian Canada streamer – Holo May KTE DAC: Initially, I streamed using my windows PC. The improvement was substantial when I removed the polluted PC signal from the signal path by replacing the PC with Ian Canada streamer. In fact, @iancanada in one of his helpful emails argued about that. But I had difficulty to believe is since Holo May was marketed as immune to jitter. Anyway, there was a roll off in high frequencies that I noticed immediately. The sound became more pleasant as if all unpleasant harmonics were rejected (in fact they were just upsent). The soundstage was increased. There was more air among the instruments. Sound from all instruments, and particularly the high frequency ones, became more refined and focused.

Ian Canada Streamer - TDA1 DAC: Since I was so satisfied with the Ian Canada streamer – Holo May duo, I saw building the TDA1 dac as a challenge to beat the Ian Canada streamer – Holo May duo. With the Ian Canada streamer- TDA1 duo, the sound stage opened even more. I confirm that the 1) the addition of Ian Canada I2s to PCM converter contributed on that and also 2) that I saw improvement when changed the first TDA1541 (non A chip) with a TDA1541A that I took from a Proton AC 620 cd player. I also confirm that there was even more air among the instruments. Though, I did not listen that refinement I heard from the streamer – Holo May duo. This is reasonable for a legacy chip not being that detailed.

The negative, and a rather determinant factor in my assessment, was that I listened harshness. I could assume that if I had available spectrum analyzing equipment, there would be 5th and 6th harmonics in the spectrum. The TDA1541A sounded better than the non-A chip but still the outcome was far from the threshold of being pleasant.

I wrote to @Gabster 2000 requesting his advice. Since for my setup I used the bill of materials Gaby used and the fact that the Ian Canada streamer together with the Holo May achives great results, I assume that either 1) this is the sound identity of TDA1541A Dac or 2) the sample of TDA chips that I used is way below the average standard of these chips. To that end, I would be keen to try some other TDA chips.

I invite opinions of forum colleagues, particularly the ones who have already built and enjoying listening the TDA1 Dac. Further, during the next vacation days, I intend to invite friends and critically listen the TDA1 dac and share their impressions.
 
Last edited:
  • Like
Reactions: JMD3254 and Clausen
It will be interesting to hear your report of what other listeners think. Harsh is not a word I associate with the 1541a sound. I am wondering if you do not like the sound of a NOS DAC. Perhaps you are sensitive to the high frequency noise. You can try a small cap across the output to filter frequencies above 20k and even run it at 2 or 4 times over sampling and see if your perceived harshness disappears.
 
  • Like
Reactions: TheodorM
Dear forum,

I am the happy owner of an Ian Canada #93A project, it includes a PurePi II, a FiFoPi Q7II, a MonitorPi Pro and a TransportPi AES. After the setup I started using it with roon and ropieee, and had an issue: roon was not detecting the device quite often. I switched to DietPi as roon bridge and it never happened again, recommending it to everyone with similar difficulties. I am burning through the Audirvana evaluation month with an RPI4 4GB added to the project at the moment, and it sounds just great.

I have one question though: the kit did not include the cable to connect the PurePI II J10 to the MonitorPi Pro J5 to coordinate the sync charge function. I acquired some cables from Aliexpress, and after installing one of them the function started working, I heard the relay clicking and saw the current going up when the system went into MUTE. This stopped the next day, and even replacing the cable with a new one does not make it work again.

Anyone have any hint, tip, suggestion, spell or incantation which might help me getting closer to understand what happened, why this worked for a day and stopped working afterwards? I measured resistance on the cable and it is OK. I can absolutely live without it, especially as I am planning to add a 3.3V Supercapacitor device, UcConditioner Pro sonner or later to the system, just to push it to the limit, but still kind of interested what is going on with the sync charging?
 
that clicking is a relay.
As soon as no music is played, the Pure output LED turns off and starts charging. As soon as the music starts again, the Pure LED turns on and stops charging. he will also charge it for a moment with the next number. it's all in the manual too.
When playing for a long time when it is almost empty, the pure output LED turns off and as soon as it is charged again (orange LED) the Pure output LED turns on again.
 
Last edited:
Thank you again for taking the time to collect these. My assumption is at the moment that clicking might happen when switching from pure output mode to protection mode. When the battery power is lower, it will run on protection mode continuously, and no clicking. Hopefully will be able to reproduce it tomorrow. As I listen to music continuously almost when working from home I wonder how long it will take to drain the batteries enough to go into protection mode. I will try to measure that also.
 
I think I was right in my assumption: whenever the pure led goes on or off the click comes. Although when there is not enough charge in the batteries for pure I suppose the system is working in protection mode continuously, so no clicking when going to mute or coming out of it. There is an interesting trend in the current consumed: right now playback is going on in pure mode, the system takes 0.52 A at 5V. If I stop playback, click comes, pure led goes off and the current consumption goes up to around 1.2 A. (These are not calibrated numbers, I use my hand built linear PSU to feed the PurePi II, and that has a simple current and voltage meter in the output line, so no calibrated numbers but the magnitudes are still interesting). When it was doing playback without the pure led on, it was consuming roughly 0.6 A regardless of playing music or not. That was kind of surprising that even when not having enough juice for pure mode when mute was initiated the high current charging did not kick in. After a while consumption drops down to 0.67 A and the charged battery led comes on.
So the only question remaining in my head: when playback is happening without the pure led on, why power consumption does not jump up for charging when the music is topped? Would be very interesting to know.
 
  • Like
Reactions: Clausen
Dear Forum and Ian,

I'm using UcConditioner Pro 5V for 2 months approx with great result feeding topping E30 Dac. Today, after 2 weeks absence for some holidays I switched on again and after a while the 4 usual leds lit on, plus a fifth led : D6, wich is overvoltage above 2.7V on one cap !!! UcConditionner is fed by Ian's LinearPi MkII

I measured incomming voltage from linearPi, and output voltage from UcConditioner : same 5.07V. Then I removed the UcConditionner from my setup and mesures voltages cap1 : 2.85V, cap2 :2.04V. There is indeed a problem.

Is there something I can do, or is the UcConditioner definitively defective ?

Update : with adequate resistor I discharge one cap to bring it to same level than second cap. This brought the system to normal condition again. But certainly this will happen again after long switch off.
 
Last edited:
Hi @greinedo,
Just to share my own experience. I initially purchased one 5V UcConditioner. I fed it with 5.07V from LinearPi MkII, D6 led was on for a long time. Please see below a photo, I sent to audiophonics while I was experiencing the issue with the 5V UcConditioner.
Audiophonics assured me that so long as the input is not more than 5.3V there is no problem. I decided to allow some grace period. After about a week of use, which was adequate for the UcConditioner to break in, D6 led went off.
Afterwards, I purchased another two 5V UcConditioners. There was the same problem. When I let them break in their leds also went off.

From my experience, I would advise you to leave it for some it and see, so long as you do not feed it with more than 5.3V.
 

Attachments

  • 13.jpg
    13.jpg
    231.2 KB · Views: 76
Dear @TheodorM,

Thank you for sharing your experience ! I indeed noticed again this morning same problem, and a breaf and weak D6 led blinking on another UcConditioner (I have 3 in total) that led me thinking it's may be not a defective unit but something there by design. The annoying thing is I measured 2.85V wich is exacte overvoltage limit for this ultracapacitor. Was certainly a bit higher because it took me some time to remove it and measure.

I forgot to indicate earlier my setup :
N100 with Lyrion server (LMS) on linux & Qobuz
Pi 3B+ with ShieldPi Pro MkIII+5V UcConditioner Pro&LinearPi MkII, TransportPi Digi II +3.3V UcConditioner&LinearPi MKII Pro and PiCore player,
Oyaide Spdif cable,Topping E30+5V UcConditioner Pro&LinearPi MkII
Home made tube preamp and modified TPA3251EVM board.
 
  • Like
Reactions: TheodorM
I had this on one of my UCConditioners and emailed to Ian about it.
I am using LinearPi II's, and he said it should be no problem and that after a few runs D6 will go back to normal.
Mine did do that.

Also asked him to further describe D3 and D6 in the manual, instead of that they are to indicate overvolting without any further information.
But I didn't receive an answer to that.
 
I'm new to DIY builds and need some advice.

I'm trying to extend the Monitor Pi Pro from the FiFo Q7 card to the case front plate, but I don't have a 50A Universal Raspberry Pi GPIO extension kit with 6" and 12" FFC cables.

While FFC cables are easily available locally, could you please suggest any alternative options for the extension connectors that we can consider, with related links.

Additionally, are there any other precautions I should focus on?

I also read that to use the IR remote control with a GPIO extension, adjusting Pin 25 at the software or hardware level is mandatory. Is this correct?

Thanks in advance.
 
I had a very strange but unfortunately not unseen issue this morning with my FifoPi Q7II and TransportPiAES system (porject #93A) this morning. They are fed by a PurePi II and sitting on the top of an RPI 4B 4GB running DietPi. I abandoned Ropieee and roon because the output device became inactive and roon could not open it like every other day after power on and I had to reboot and reinstall roon bridge and do other black magic to make it work again but could never intentionally reproduce or fix the issue. Adurivana and DietPi were working ok this morning, when the same happened: Audirvana could not open the audio device, even after several reboots and power cycles. When I plugged in the DAC also on USB into the RPI to see what might be broken magically started working again. I suppose the usb plug might have kicked off some hw scan or restarted some part of the kernel which initiated something which made the card visible again. Any idea anyone what could be done to avoid these unpleasant incidents happening again and again?