audio-talk

The last of the components came from Ali Express.

Lash up just to test whether the I2S would be compatible:



Apart from one capacitor popping due to incorrect polarity and an issue with the bias on the I/V stage it sounds totally amazing. Blows the socks right off the AKM, just so much more dynamic and containing extra detail. A lovely sound. And in one channel only so far. I just need to build the rest of the boards and wire everything up properly and I think I'm going to be very happy!

Nice work, Tony. Looking good and great you are happy with the sound.

Thanks Greg! It's just like old times for me. Stop me if you see any "mad box" of motor run caps appearing!

Final result coming soon...

This project is still ongoing. A few trials and tribulations along the way - it wouldn't be a fun hobby unless there were challenges, eh?

I found an interesting fault that I believe I have now diagnosed and have a fix for: the Logitech Transporter designers, in their wisdom used the non-I2S compatible mode for the AKM DAC, hence the data bits are offset by one bit position. All is well (seemingly) with the signal digitally attenuated but hit the top 2 bits and the game is up! So now I need logic to delay the data by 1 bit and it _should_ be good. I'll prototype on breadboard and then if it works, I'll get a new PCB made.

Other than that, all the other modules are built and tested and awaiting this last wrinkle to be ironed out and then I can fully assemble.

Considering that the DAC is running on an attenuated 14 bit signal and on mostly shared bench power supplies it sounds surprisingly good. There is something just a lot more exciting and realistic about the sound compared to the AKM. The AKM sounds fine, provided you don't do direct comparisons but when you do you realize just how much excitement has been lost. I've heard the term "veils" and it sort of fits. A "boring veil", if you will...

Who knows, when it's done and I take it to a meetup, it might sound like sh*te.

Well, the breadboarding exercise to test the one bit delay of the data was successful and I have now received my new pcbs with the proper implementation of that logic.

I have gone for a 2 level approach where the lower "motherboard" will be attached to the I2S lines and hold the logic and its psu and then the actual DAC board will plug in using SIL headers. That way I can try out different DAC modules (with say different layout, decoupling caps, A, S1, S2 version of chip, etc) and have an easy way to change them over quickly to do side by side listening tests.

Hopefully this weekend I will get time to populate these new pcbs and do some initial tests.

I'll post more photos as it progresses...

Ooh, it works.



This was an initial test of whether the data delay logic worked, hence the rats-nest wiring and lack of proper power supplies. (plus even capacitors on the DAC!) The logic is SMD and hidden on the lower board in this photo.

It sounds remarkably good considering the supplies are on bench types and all mixed up with the digital and analogue. I did have issues when the delay logic was on the breadboard but now using SMD and proper track layout and ground planes it is fine.

Next up to remove the board and complete the power supply for the logic and also finish and test all the other supplies and get them wired in.

Almost there!



One small miscalculation so I need another small pcb making up but it works, that's the main thing!

You could do with a ‘Madbox’ for that lot

I feel sorry for the SMPS left cowering in the corner.

Nick wrote: ↑Tue Jul 06, 2021 10:53 pm I feel sorry for the SMPS left cowering in the corner.

Yes, he would've been for the chop except that the displays take 1.4A so no way I could fit a linear in there. He's safe for now.

Greg wrote: ↑Tue Jul 06, 2021 10:07 pm You could do with a ‘Madbox’ for that lot

It is the evolution of the old "Madbox" for an ever madder me!

I made the DAC board a plug-in module, via SIL headers so that I could try different things with the DAC itself without having to waste more support circuitry than necessary.

I have a plan to try different decoupling caps and also the DEM reclocking mod that "sort of" worked on my main DAC when we played it at Richard's house many years ago. It was almost impossible to try different things with the other DAC as space was so limited and it required almost dismantling to get to the main DAC pcb. With the hindsight of that I thought I'd make it easily accessible and think of as many future experiments as possible.

Up next: A new DAC PCB with Philips "Orange Drop" caps, DEM reclock and I2S attenuation...

I'm trawling the web and trying to find the best "bang for buck" tips for implementing the 1541.

I have to say V2 of my DAC pcb is sounding wonderful. I am useless at describing how I hear stuff so I won't even try.

Hopefully at some point in this century we'll have a meetup again and I will bring it along and people can make up their own minds....

And...new DAC carrier board fitted. This one has the famous orange drops as the decoupling caps. Best sound so far but too hot to wear headphones right now so more testing will have to wait.

I realized that I'd not posted a "final" photo - so here goes: