XMOS-based Asynchronous USB to I2S interface

Sorry for offtopic. I have Xmos XU208 board from Aliexpress and have encountered an interesting problem, it works good with 5V external supply from KIA7805 but absolutely dead with LT3045 (0.5A) or TPS7A4701 (1A). With those LDOs it even not recognized by PC. I've measured current, whole this board eats 110-120mA at boot and when idle and voltage is ideal 5V... But if I boot up TPS7A4701 first and then connect it to XMOS it works. Any ideas what is wrong with these LDOs and XMOS ?
Hello probably too slow to rise to the desired (5V) voltage at start up. I had a similar problem with the SPDIF board I made software controlled with Arduino.
I exaggerated with some Electrolytic Caps in the PSU and I was not able to initialize the board , because the micro controller (Arduino) was already trying to communicate with the board, which had not yet reached the power up voltage.
This can explain why if you boot first the power supply and then connect it to the card, it works.
Anyway WAVEIO card reach more than 0.5 A current absorption with high sample rate files.
 
  • Like
Reactions: Anatolii_A
@ mizohalen: Sorry for my late response! I managed to send you a PM with instructions on how to change the firmware but it seems that you already did this!
@ arteom: I really hope the isolator is not damaged due to ESD discharges! Moreover, regardless of the DAC you're using make sure that it does accept 5V I2S signals on its inputs! In fact, that's the main criteria to choose the voltage supply wired between the V+ and Isol.GND pins. What I'm saying is that a 3.3V compatible chip will not be happy with 5V signals...
 
Anyway WAVEIO card reach more than 0.5 A current absorption with high sample rate files.
well, if this is happening then it's somehow bad! At first, WaveIO was designed to be powered form USB and there, the current limit is around 0.5A. I say "around" because I saw many aberrations... Somewhere in the past I decided to measure the current consumption of WaveIO under different sample rates up to 384 fs and none reached 0.5A. The measurements were made using my Agilent/Keysight 34461A multimeter in a controlled environment @ 25 °C with histogram enabled and each sample rate was checked for about 20-30 minutes. In fact, the highest value I got at that point was 0.4 A. But, I don't say that it cannot be possible to be higher than that! (I saw parts changed over time, more or less, especially after Covid19 even though they had the same serial number - all bought from Mouser or DigiKey!)

Is it possible to setup the waveio to playback in 16bit 48khz? Only options showing in sound settings are 32bit.

Sadly not with this board revision! All samples are 32-bit wide.
 
  • Like
Reactions: arteom
Yes maybe I was too generous saying absorbtion is much more then 0.5 A...
What I measured with a multimeter was if I remember something like 0.42 A. Anyway just other 0.4A. Since my DAC has not 384kHz capability I linearlly extended in the higher values other 0.5A the current consumption. I just use a Studer type power supply.
 
Hello probably too slow to rise to the desired (5V) voltage at start up. I had a similar problem with the SPDIF board I made software controlled with Arduino.
You're right ! To slow to rise, I've made temporary solution with RC delay + relay, it runs very quick, may be 0.1-0.2 seconds, but here is full 5.0V already and XMOS like it. The question is how to friendship them together in best way
 
@ arteom: as can be seen above, WaveIO will run properly with a 5V @ 0.5A PSU. Anything above that would be considered a bit overkill BUT 1A is not a bad solution if you like to have some current "reserve". In my opinion, anything other than USB power is good alternative - with very few exceptions! This Traco PSU seems to be a nice approach excepting the fact that is a SMPS (Switching Mode Power Supply). I would preffer to be linear instead but you can add some filters at output just to lower the ripple noise.
 
  • Like
Reactions: arteom and dddac
Hello Luciano

I will try to be concise and forgive the bluntness
I've been waiting for the new WaweI since September, writing to you privately, lately I haven't had any answers
I write here
Is there a reliable release date for the updated WaweIO?
Is he really carrying out the project?
I've been using an interface on loan for almost a year and I have to return it soon and I would be left without music
I definitely have to evaluate this
In Italy there are 3 of us who are interested but we want to clearly understand if the new version is progressing and the waiting times
Thank you and I hope for a sincere response
Marco
 
As of now, almost all of the changes I made to WaveIO will not be "visible" for end user. I had plenty of issues in the past years, some of them acted like a nightmare to me so I decided to fix those first. Here I'm speaking about ESD protections to I/Os and USB. Aside of that, some of the parts hit the EOL limit so I have to find alternatives. Since I'm here, I want to change a couple of parts with better ones while try to keep the same price range. Frankly I can't say I'll succeed because I remember when I started that XMOS chip I'm still using had a price of around 7 Euros in quantity of 90. No that price varies but it at least double and Qs is around 180 pcs. Same for all parts involved.
And, the most important changes are related to assembly and testing processes of these boards.
Ah, not to metion jumpers: I hate them from the bottom of my soul. If the one used for USB input switch may be preserved, the other one will be gone forever!
There were many PSU issues because of that so I really want it gone. Perhaps a relay and a fuse will be a nice replacement.

So, this is a part of what I'm "fixing" right now to WaveIO. As I said, most of the changes are "transparent" to the end user but I hope will add up to the reliability of this board.
 
  • Like
Reactions: Sagen and dddac
Lorien, nice to hear you are working on the board 👍 After I broke two boards with ESD from bad USB port on PC ( one board from you and the adder I got from Doede with the DDDac ) I did the trick with the Bulgin USB cable to J1 and switch jumper on J8, I can play music again.

What Is the advantage / disadvantage compare to use the oem USB connector on the board?

Frank
 
What Is the advantage / disadvantage compare to use the oem USB connector on the board?
Both paths are the same! I placed the second USB input (the one you're using now) to avoid mechanical constraints added by the on-board USB connector and to allow using different USB cables (I'm assuming there are out there cable adapters that are using different connectors than USB-B). Later on I saw it was a good idea to temporarly bypass all problems generated by the USB multiplexer chip.
Yes, pin jumpers and additional length of the adapter's cable may pose a problem but if you get a 'reliable' product then all should be fine!
 
  • Like
Reactions: Sagen