But samadi81 tried re-driver board?
"The re-driver has improved sound quality depending on the setting of the signal conditioning. The setting value varies depending on the nature of the cable, but it had a good effect with medium-intensity signal conditioning for cables over 1 meter."
...
BTW
You said that You measured the loss of USB datas that is not related with listening
Same here, never had any problems with USB data integrity even in extremely long measurement sessions, loopback recording at high fs, for hours.
Using Win10 and ASIO for RME ADI-2 Pro, you can monitor CRC and other errors on the USB driver level and I did actually try to provoke errors. I had to cut open an USB cable, single out of one of the data lines and place a large ferrite close to that conductor, to totally screw up line impedance (CM and DM). At some point of proximity communication would break totally and re-enumeration would also fail. So a binary go / no-go scenario, basically. Backing off again a little bit, once it worked the connection was reliable as tested for about 6 hours.
The only way I could provoke errors with a normal certified USB cable was equally brutal, with an ESD pulse (from a piezo lighter) creating a ton of peak current in a wire parallel to the USB cable. I used a fully isolated section between two USB isolators to reduce the risk of frying the RME's USB input...
This is exactly related!
For good sound performance, there are only 2 things required from the digital interface:
1) no errors in data ("bit perfect")
2) no errors in timing (jitter) at the point of D/A conversion (this is provided by async. sync. from the goog oscillator, located close to the DAC chip).
Eye diagram is a good thing, but on a picture above I didn't saw a problem. It changed, but still inside the requirements.
BTW, I tried to switch 2 USB Hosts (PC and Media Plyer) to one USB DAC port with a regular signal relay (Omron G6K).
Using this type of relays for USB High Speed switching is totally against the rules, but it works.
I did not measure the eye diagram, but with 1-1.5m cables there were no problem with "listening".
(USB PHY was Microchip's USB3318).
I'm not well versed in all of this digital wizardry. If I wanted to buy an ADUM3165/3166 eval module to use with a USB oscilloscope, which one should I get? The upstream/downstrean clock thing confuses me.
You want to connect you scope ("device") to PC ("host")?
Maybe I'm wrong, still didn't used these chips, but- see page 23, "Host Applications".
I thing you should use 3165 as at fig.19
Maybe I'm wrong, still didn't used these chips, but- see page 23, "Host Applications".
I thing you should use 3165 as at fig.19
I have seenADUM3165 is much sensitive to the output USB cable - tried few USB-A/Type-C cables with Cosmos ADC.
Some cables shows very high packet lost level, which make this setup unusable.
The same cables with the TI's ISOUSB211 based isolator works perfect (~ the same PCB schematic and layout for both versions, just different chip, and ISOUSB211 - with external 1.8v regulators ant both ends).
Some cables shows very high packet lost level, which make this setup unusable.
The same cables with the TI's ISOUSB211 based isolator works perfect (~ the same PCB schematic and layout for both versions, just different chip, and ISOUSB211 - with external 1.8v regulators ant both ends).
Thaks for the experiment 🙂
Maybe to try usb redriver, from TI, after isolator to handle the cables?
But they are in very small packages 🙁 also this feature makes these redrivers good to implement in connector body?
Maybe to try usb redriver, from TI, after isolator to handle the cables?
But they are in very small packages 🙁 also this feature makes these redrivers good to implement in connector body?
I tried, it helps.
But - the only reason I'm using the ADUM3165 is because ISOUSB211, which works great without any tricks, is not available right now.
Not "very", just "small" 🙂
I use it an another projects.
Aaaa Yes
Thanks
I have some TI redivers but my eye vision is getting vorse yeat after year... I think that Im not capable to solder them 🙁
Thanks
I have some TI redivers but my eye vision is getting vorse yeat after year... I think that Im not capable to solder them 🙁
The only thing left is the ground conducted interference that perhaps could act negatively on a not well thought out analog parts of a downlink unit. But I agree of the above list and it is complete wrt. digital L3 transfers. But sh*t happens either on ground or radiated disturbances and may impact - thats the only reason aI can think of if there still is a difference.... I always they to use opto if possible. Opto dont sound dull - it's the sound of "correct" - change other things of one is not happy 😉
//
I’m 60, I don’t even see what is written on the 0603 resistors anymore.
But I have Hot Air Gun and binocular microscope 🙂
(Hm, sound like "I have a beautiful daughter, but I also have a gun, a shovel and ALIBI !" 🙂