USB DAC isolation ?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Unfortunately you won't get around the 12Mbit limitation!
I've been in contact with AD earlier about it.

If you get along with the limitation, it can be more attractive than the Opticis or Rover.
Price wise it is obviously the winner.

Gotcha. I was mistaken, or a bit off...

I thought there were some higher speed options, but it does look like they are not fast enough for USB high speed, and apparently they are not available in the bidirectional versions as I had thought. I bet it won't be too much longer before they can make it work.

for my review, 2 questions:
- I don't know of much 24/192kHz material. You really can't live with 24/96 and 12MBps?

- isolating the I2S signals *after* the USB controller is not acceptable? This way the analog output is still effectively isolated. (and can be done with current ADuM parts) This seems like a good solution to me? And is much like the SPI/I2C examples mentioned in some of the docs? No?
 
Gotcha. I was mistaken, or a bit off...
- I don't know of much 24/192kHz material. You really can't live with 24/96 and 12MBps?

For sure you can.

98% of the tracks out there are 44.1/16 anyhow,

- isolating the I2S signals *after* the USB controller is not acceptable? This way the analog output is still effectively isolated. (and can be done with current ADuM parts) This seems like a good solution to me? And is much like the SPI/I2C examples mentioned in some of the docs? No?

I agree. I don't think it really matters at what point you decouple. You just need to do it.

I am wondering what Ayre is doing. They do optical decoupling.
 
Ok, I'm back to this topic again as I have hiss on my set-up. I put a scope on the output from my USB DAC, which has a PCM1793 DAC with differential Vout through a transformer. I attached the scope leads to the secondary side of the transformer & I see about 20 to 25 uS peak-to-peak waveform at about 8mV height coming from it but ONLY when I connect the USB cable to my PC (no music playing). I can power the DAC separately from an external supply & the scope is relatively clean when it is just powered on but as soon as I connect the ground through the shield of the USB cable, I get the waveform.

I have powered the Laptop (Dell Inspiron 6000) through an ungrounded SMPS & it makes some difference but I didn't note it as it's late here & these were quick test before bed.

So it looks like a different laptop is needed or better a USB isolator like the ADUM (I know it won't do 24/192 but clean 16/44 is better than dirty 24/192)

I'll do some more tests tomorrow but first a question:
- is there not a difference between doing the de-coupling at the USB end rather than the I2S end, in that jitter could be introduced into I2S but it's immune to it on the USB side?
 
Last edited:
Hi Jkeny,
Did you try battery powered laptop to see if the pick is there?

Adum4160 evaluation board has a delay or is in backorder at Digikey...maybe a couple of months delay...:(

Unfortunately, my laptop battery is gone & I'm reluctant to buy another one - I just might buy a new nettop instead! I'm going to find somebody with a laptop with good battery & I'll try that. The SMPS could well be the issue!

I order a $14 board + ADUM4160 from here: http://www.circuitsathome.com/store
 
To get rid of Hum/hiss from HDMI

By following the recommendation, now I can get HDMI connected to HTPC(using TV as monitor) at same time I am listening to the USB 1543 DAC. I used to have pretty strong hum and hiss because of the HDMI connection. After the modification, with HDMI connected, it even sounds better than before without HDMI.

Thank you!
 

Attachments

  • adum4160.jpg
    adum4160.jpg
    237.3 KB · Views: 405
Oops! I forgot to post here my findings and posted on -ECdesign's- "Ultimate NOS DAC" thread only...

I have built three units based on Oleg's PCB's from circuitsathome.com and I am using them for USB audio and it works great. Sound is better than ever, more detailed, more dynamic and extended, with better timbre of instruments...real bargain at US$14 :cool:

I use them with its own power supply which allows to be used as a cable extender.
I used an "audiophile" discrete Vreg but I had some troubles since it wasn't very good regulating when the USB receiver was bus powered due to current demand: to weak.
I suppose part of the good effect comes from isolation itself and part from the better power supply of the isolator compared to that of the source (PC or whatever), because at first, with a wallwart Tx and weak supply capacitance it sounded detailed but thin and when I added a monster supply, the bass, punch and overall fullness of the sound became awesome. Try for yourselves and tell me...

For the bus powered DAC I will try today a plain L7808 supply and then swap it forp a TeddyReg supply and see if the sound changes.

Hi 2A3SET,
Wow, congratulations, though I have to say it looks a little scary from here :D
I had to use a TVC to get rid of the hum on the main system number 1, due to the multiple sources, TV Cable, DVD, TV set, DAC etc...

Guys, try the chips, it is worth.

Regards,
M
 
I did all the modification designed by ECdesign posted on TDA1541 forum.

48M clock, feed async shift-register reclocker, 1/4 to feed PCM2706.
I2S attenuator and dynamic attenuator for BCK, LED regulator at 5.2V for 1543, LED shunt regulator for 1543 bias at 3.8V as I am using 910R I/V.



Sorry guys, I've nothing to report on the ADUM board as I haven't built it yet - it won't work with high-speed USB so I had no need!

2A3SET, what modification did you apply?
 

Attachments

  • nos1543-1.JPG
    nos1543-1.JPG
    60.7 KB · Views: 642
Hi 2A3SET,

I did all the modification designed by ECdesign posted on TDA1541 forum.
48M clock, feed async shift-register reclocker, 1/4 to feed PCM2706.
I2S attenuator and dynamic attenuator for BCK, LED regulator at 5.2V for 1543, LED shunt regulator for 1543 bias at 3.8V as I am using 910R I/V.

:worship:

Today I bought my Red LEDs for the bias of my many DIDACs. ;)

Cheers,
M
 
USB isolator - ready to use solution

Hi all,

We have an off the shelf USB isolator in a nice little USB stick box, with built in isolated power transfer also. Selling it cheap at USB 2.0 Isolator - Isolated USB

There is no clock recovery in the adum4160 - which is also used in the USB2ISO stick. There will be no extra jitter - the transfer of data is done via magnetic coupling - just introducing a constant delay as 1 hub/wire.

The ground loop noise is removed but the sound quality is the one of the USB DAC. Normal USB sound cards uses a 16bit AC97 compliant sound chip with integrated dacs.

When the USB2ISO module is put between, there is no galvanically connection from PC to amplifier. This also removes the risc of damaging the PC and Amp.

Hope this solves your problems, as it solved mine with 5.1 HD-DTS audio from PC to Surround system.
 
Two points:
- ADUM will NOT work with the Hiface USB 2.0 asynch transport but it seems some have reported that it works with the Musiland USB 2.0 asynch transport. Why?
- How much of the reported benefits are due to the external 5V supply that you provide with ADUM, rather than the galvanic isolation? Maxlorenz has hinted at this!
 
USB 2.0 and speeds - adum4160 - USB2ISO module

Two points:
- ADUM will NOT work with the Hiface USB 2.0 asynch transport but it seems some have reported that it works with the Musiland USB 2.0 asynch transport. Why?
- How much of the reported benefits are due to the external 5V supply that you provide with ADUM, rather than the galvanic isolation? Maxlorenz has hinted at this!

Hi Jkeny,

Thanx for the info, I will add the results to our support page. The reason is the BUS speed - please note that USB 2.0 does not imply USB bus speed - read more USB2ISO support page.
The Adum4160 is a fullspeed isolator running at 12Mbps=1.5MBps-protocol=1.1MBps. Both USB SPDIF modules run 192KHz*24bit*stereo (1MBps) so they are probably both hispeed 480Mbit. The one using the cypress chip (Musiland) also claims that they do not use the USB clock, which might indicate that the Musiland does? - so both a bandwidth issue and clock issue can be why the Hiface does not work and the Musiland does work with USB2ISO.
 
USB Audio problems with a Subwoofer

Hi,

I was wondering if the USB2ISO can help a situation I'm having, since nobody else has. I have a thread on diy, but heres a summary of the problem.


I have a PC based audio setup (Ubuntu /USB). The USB output feeds a Firestone DAC that goes to an Audiolab preamp. The preamp feeds a Behringer 2310 and from there to the power amps. A subwoofer (Dayton 315 with Oaudio 500) is connected to the pre-out on the preamp.

PC (USB) --> DAC --> Preamp --> Behringer active X/O --> Power-amps

Preamp --> Subwoofer

Pulseaudio is used in Ubuntu and the volume is disabled. The DAC volume is set to ‘shunt’ mode and the pot fixed at 12’o clock.

Most program material plays fine. If there is deep bass and/or the volume is turned up the USB stream gets cut-off and replaced with very loud white noise from the main speakers. Turning the volume on the preamp down stops this. Restarting the audio player in Ubuntu (Audacious) does not bring the music back completely. There’s noise riding on the signal & the audio bit is very feeble.

Switching the amps and DAC off brings it back online, till the next ‘white noise’ break. So far, I’ve been able to play the system without problems if the SUB is switched off completely.

Weird – I don’t know, why – I have no clue. Has anybody seen this before ?
Can somebody please help me understand how to fix it.

Thanks,
Sunil
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.