Some devices will not work, because Vbus is used not only for power but also as a signal for "connect/disconnect".
To put a resistor at GND is not a good idea.
yes some devices use vbus for detection or even powering the usb section
some are happy with only gnd and even this isnt "really" needed in most cases but just so your dac initally connects
connecting ground over a resistor gives you basicly the same potential and a way so your dac connects normally but isolating gnd "further" without galvanic isolation
if you use a external power supply you need to directly connect its ground to the dac, the resistor needs to sit between dac/power supply and the pc
as far im concerned there is nothing wrong with this method, the designer of ifi`s idefender told me that its basicly what the idefender does (tho the idefender is more "fancy" since it switches automaticly between pc and external power)
Please could you expand this?
Here we're talking about USB Isolators, and they come before the DAC or any other USB device.
So the negative or ground of the external (single) power supply has to be connected to the GND of the Isolator in order to the Isolator itself works.
i was purely talking about a modified usb cable not about connecting trough a isolator
regarding a isolator, it could be potentially beneficial (just a guess tho) to also use a resistor on ground before isolating, while galvanic isolation is "good" i dont think its perfect (but i guess this could be easly tested by a modified cable before the isolator)
but with this you would lose the ability to use power from the pc
if you use a isolator and external power use the resistor on ground between pc and isolator, you can directly connect the external power supply to the isolator and therefore the dac
the resistor value isnt "that" important but at some high value point the dac isnt connecting anymore .... 100ohm uses the ifi idefender, 100-300ohm should be safe and everything above is experimental
Last edited:
Please note that cable or isolator the point does not change because the 5VDC coming from the PC (or other) are still needed from many devices (Isolators, DAC, USB SATA Bridge, etc).
I've more than one USB audio device powered by an external power supply and only Singxer F-1 XMOS Digital Interface does work with the Vbus wire detached.
In my experience all ADuM chips I've (3160, 4160, 4166) need the 5VDC wiring in order to work.
I've more than one USB audio device powered by an external power supply and only Singxer F-1 XMOS Digital Interface does work with the Vbus wire detached.
In my experience all ADuM chips I've (3160, 4160, 4166) need the 5VDC wiring in order to work.
Last edited:
I installed a simple 2 way switch. Start up is with USB power. Once the device is recognised I flipp the switch to feed 5v LT3045 with 175F SuperCaps (2x350F). Can be easily automated.
But why bother with USB audio anyway if Dante Monism boards that sound better only cost €130 and give us an additional channel.
Aoip is the more sophisticated way to get audio out of a Computer.
IME its possible for USB to sound fine. Could it be Dante Monism sounds better than your particular USB for some unknown reason?
Yeah, but on the DAC side of that isolation transformer is yet another computer creating noise that travels everywhere. OK, maybe not in every case, but how do you know?
Singxer F1 with Lifepo4 on the clocks, dedicated powersupply with LT3045 into Maxwell 350F caps is not a sloch. I have yet to feed any clean signal into the Monism boards and have not fed them with external dedicated clock and they beat the mentioned usb solution in sonics by a small margin. I know multiple people who have swiched to Aoip and they will never go back. What boggles my mind is that hardly any of them use the multy channel posibillities of these 8 or 16 Channel RedNet boxes. Go figure.
Ulogon, I do not see a quote button in the post you asked how to switch automatically to external ps. A simple relay would do the trick. On my Topping E30 II and DX1 I had put a Sanyo 330 uf SEP on the USB input to bridge the interupt time of the switch. Otherwise it woud loose connection.
Okay. Would be helpful to understand a little more about the USB it is being compared to. Would be willing to describe the dac and the USB interface, as well as the overall clocking system? Is the USB board galvanically isolated? How it it powered? Is the system entirely synchronous? What clocks are being used? How are the clocks powered, bypassed, buffered, and switched? Where is the USB board in proximity to the dac board? Are there pics of the setup?
Sorry for asking so many questions. Its just that I have found a lot things can matter. I try to do everything right here, at least insofar as I know how. Don't have much idea how your USB implementation compares. Maybe there is something that could be causing problems?
It's a Forum's feature applying to the actually last post in the thread page.
However, you can just select the text from a post and then the Quote balloon does appear.
No idea at all how to reach what you described starting from scratch, but it seems to be great!
Hi Mark,
I admit that usb has come along way.
Been using it for almost 20years starting with a diy version of John Svensons 2706 that he introduced over at AudioAssylum.
My Topping E30II with clean power to the relevant sections is sounding mighty good.
However I look at systems as a whole. What good is it to get a component after alot of fiddling to great performance only to loose much of the gain in other areas like passive crossovers. The beauty of Monism/RedNed Dante systems is the multiway possibilities.
Crossover, phase correction, room correction, time delay for any of my speakers at the click of a Mouse Button.
If you want to achieve this with USB stereo you need to intoduce a separate device like the Behringer DCX etc. that needs to be tweked and reclocked.......,........
Every extra item with features/variables adds complexity and yet another maintenance issue and fault possibility. Component tolerances become critical as well just like channel (level) differences.
The human factor is that things need time and self education, right interpretation and dependance on the person that takes care of this all.
Then aging comes around. One might misinterprete stuff or simply loose knowledge or the interest. No one else in the family is interested as it was your thingie. No one even understands what it is and what the purpose is let alone how to solve a simple issue. The mouse clicks are reserved for social media
The passive cross-over might look like a big problem to some but these don’t ask for attention nor a learning curve.
The human factor is that things need time and self education, right interpretation and dependance on the person that takes care of this all.
Then aging comes around. One might misinterprete stuff or simply loose knowledge or the interest. No one else in the family is interested as it was your thingie. No one even understands what it is and what the purpose is let alone how to solve a simple issue. The mouse clicks are reserved for social media
The passive cross-over might look like a big problem to some but these don’t ask for attention nor a learning curve.
Last edited:
AoIP may well be better than USB in terms of functionality and flexibility. However this all started with the dubious claim that AoIP-to-I2S board sounds better than USB-to-I2S. While it is possible that Dante Monism boards sounds better than a hacked Singxer F1 that does not change the fact that AoIP clocking has flaws. Also there seems to be a misunderstanding about the number of channels supported by USB. Even if Singxer F1 only supports 2 channels nothing prevents using USB for 16 or 32 channels with TDM output or e.g. 8 channels with I2S output.
Yet another theorizer.
There are numerous RedNet user over at the German Aktives-Hoeren-Forum. They had fancy usb chains with Ether Regen, Intona, expensive clocks ect.ect. The only reason they do not go back to USB is sonic performance. If the clocking issue that has been brought up was such a big deal then why does Aoip sound this good?
That USB can theoretically be use in multichannel is of no benefit for people who are not experts in the field.
I have plenty of tweaked USB devices and non has reached the performance of the Monism boards as of yet. I don t have the money or time to do all that has to be done to USB to reach Aiop. I am not saying it can t be done all I am saying is that it is to much work for me to achieve what I already got.