Thanks for your answer. I know your USBridge Sig is near in production, so major changes in PCB is not feasible. But I think you should consider to apply some shielding around the RPI Compute module to prevent EMI radiation from polluting other component, especially clock.
@cdsgames
[/QUOTE]
Nirvana...was delayed for at least 4 weeks.
Shanti lunch date is July 8. We will open preorders a week in advance. Mass production has begun , each unit is tested individually . Hopefully I can post pics in the next 48h (casing just landed). At 3A (3.9A on burst) it can easily power the RPI4
or USBridge Sig/ Audiophile SBC
Price is 159$ (EU is slightly more+ VAT)[/QUOTE]
Any updates on Shanti and Nirvana?
[/QUOTE]
Nirvana...was delayed for at least 4 weeks.
Shanti lunch date is July 8. We will open preorders a week in advance. Mass production has begun , each unit is tested individually . Hopefully I can post pics in the next 48h (casing just landed). At 3A (3.9A on burst) it can easily power the RPI4
or USBridge Sig/ Audiophile SBCPrice is 159$ (EU is slightly more+ VAT)[/QUOTE]
Any updates on Shanti and Nirvana?
Last edited:
@Skip Pack,
If you are referring to the Katana, I've found that powering the stack all at the same time works very consistently. I am running my stack with 4 power supplies... RPi/Isolator 1.2 input, Isolator 1.2 output/Katana DAC board, Katana output board, and Katana Microprocessor board.
I've been running the Katana since it first was released into the wild and have found this provides the most consistent power-ups.
Greg in Mississippi
If you are referring to the Katana, I've found that powering the stack all at the same time works very consistently. I am running my stack with 4 power supplies... RPi/Isolator 1.2 input, Isolator 1.2 output/Katana DAC board, Katana output board, and Katana Microprocessor board.
I've been running the Katana since it first was released into the wild and have found this provides the most consistent power-ups.
Greg in Mississippi
cdsgames,
Thanks for all of your work on Allo's new USBridge Signature. I've been waiting for a product such as this almost as long as I've used RPi's as I2S streaming sources.
Your test results confirm what I've heard since early 2009. That's when I first tried a linear supply to power a motherboard-based cMP/cPlay computer music player, based on Peter Daniel's great work. That was a revelation-level improvement in the SQ of that setup and I continued building on that success, ultimately ending up with nearly 15 separate linear supplies powering the various ATX-24/P4 rails on the motherboard, the Juli@ soundcard, the I2S-connected DAC, and various peripherals such as the mouse, keyboard, and SSD Op system/music drive. The only thing I didn't try before I moved onto other setups was modifying the motherboard, where some had replaced the onboard DC-DC converters powering the memory with linear regulation to good effect.
The same techniques worked just as well when I went first to modifying a Sony HAP Z1-ES (which uses a small processor-based board as the built-in music server) and then to RPi's. While I got very good results moving from basic SMPS supplies like the iFi to better linear supplies and ultimately to a dual Uptone Audio LPS-1.2 setup paralleled using one of MPAudio.com's dual 3||LT3045 boards, the SQ took a significant and decisive jump when I modified a couple of RPis 2Bs to use the LDOVR.com RPi Mezzanine board which uses good linear regulation to provide not only the main 5V, but also the 3.3V and 1.8V, replacing the on-board DC-DC converters. Those are now my reference RPi sources.
You've taken that several levels higher in your USBridge Signature ALONG with fixing some of the inherent HW issues in RPis 3 and earlier. AND your test results confirm what I and many others hear, that providing the processing board with the best power produces the best SQ from the connected DAC.
AND in this realm, the recently announced RPi4 (and previous RPi3) are not even in the game as they still use noisy DC-DC converters for their various rails.
I know that some claim that the current SOTA galvanic isolation plus reclocking interfaces (such as your Isolator / Kali combo) eliminate any influence from the source setup affecting the SQ of the output. BUT in the real world and at the current state of the art, that does not happen. I know of no implementation of that type of setup that achieves that level of isolation. It is definitely NOT about just delivering 1s and 0s.
I posted more info on this from the perspective of I2S interfaces here:
IanCanada's Latest RPi GB Goodies Impressions... and your tweaks, mods and hints...
AND Markw4 followed with his view, providing some additional mechanisms, but with the same results here:
IanCanada's Latest RPi GB Goodies Impressions... and your tweaks, mods and hints...
I know of some development efforts in audio attempting to develop new interfaces that WILL eliminate any upstream influence. The first application of this I know of will be in the soon-to-be released Uptone Audio EtherRegen. While it is a historic and heroic attempt towards this goal, even they say "we'll see", though in my conversations with them they've been a lot more positive about the results.
But the EtherRegen isn't available yet. AND it is designed to provide Ethernet isolation, which still leaves us using an RPi-based endpoint to produce the I2S stream going into our DACs. Ultimately I expect Uptone will extend that to their USB devices, but AFAIK they have no plans to do so for an I2S-connection that would be useful for us DIY'ers.
SO until an entirely new generation of I2S interface becomes available, the USBridge Signature will likely be the best RPi-style endpoint for an I2S-connected DAC. I know Allo is also intending it to be the best USB endpoint and I hope you achieve that too. That is where most of the audio market is today. I will say I've been watching the development of the latest Sonore optical Ethernet modules and OpticalRendu and it seems they've made significant strides with their latest offerings. Still, where I believe you will price the USBridge Signature will make it VERY attractive in comparison!
But since my DAC setups all use I2S feeds, that's what I'm focused on. AND it will be the only game in town for improving that realm! I can hardly wait to get my hands on one!
Finally, to @Soundcheck, with disrespectful, demeaning, and denigrating comments such as:
- "The new question if the SBC suddenly should now be in charge to improve THD-N on the audio interface is nonsense."
- "That's nonsene. These "testers" are simply clueless."
- "I'm not commenting on Gregs wild fishing exercises."
Why would we even engage with you? I think cdsgames is a saint for doing so... me I have no time for someone who leads with such dismissive comments. AND I've been working with power upgrades to improve my computer audio as long as you have on the Op System and SW sides (AND I haven't ignored those areas in my various implementations as they ARE important, but can't fix inherent issues with poor and noisy power).
Greg in Mississippi
Thanks for all of your work on Allo's new USBridge Signature. I've been waiting for a product such as this almost as long as I've used RPi's as I2S streaming sources.
Your test results confirm what I've heard since early 2009. That's when I first tried a linear supply to power a motherboard-based cMP/cPlay computer music player, based on Peter Daniel's great work. That was a revelation-level improvement in the SQ of that setup and I continued building on that success, ultimately ending up with nearly 15 separate linear supplies powering the various ATX-24/P4 rails on the motherboard, the Juli@ soundcard, the I2S-connected DAC, and various peripherals such as the mouse, keyboard, and SSD Op system/music drive. The only thing I didn't try before I moved onto other setups was modifying the motherboard, where some had replaced the onboard DC-DC converters powering the memory with linear regulation to good effect.
The same techniques worked just as well when I went first to modifying a Sony HAP Z1-ES (which uses a small processor-based board as the built-in music server) and then to RPi's. While I got very good results moving from basic SMPS supplies like the iFi to better linear supplies and ultimately to a dual Uptone Audio LPS-1.2 setup paralleled using one of MPAudio.com's dual 3||LT3045 boards, the SQ took a significant and decisive jump when I modified a couple of RPis 2Bs to use the LDOVR.com RPi Mezzanine board which uses good linear regulation to provide not only the main 5V, but also the 3.3V and 1.8V, replacing the on-board DC-DC converters. Those are now my reference RPi sources.
You've taken that several levels higher in your USBridge Signature ALONG with fixing some of the inherent HW issues in RPis 3 and earlier. AND your test results confirm what I and many others hear, that providing the processing board with the best power produces the best SQ from the connected DAC.
AND in this realm, the recently announced RPi4 (and previous RPi3) are not even in the game as they still use noisy DC-DC converters for their various rails.
I know that some claim that the current SOTA galvanic isolation plus reclocking interfaces (such as your Isolator / Kali combo) eliminate any influence from the source setup affecting the SQ of the output. BUT in the real world and at the current state of the art, that does not happen. I know of no implementation of that type of setup that achieves that level of isolation. It is definitely NOT about just delivering 1s and 0s.
I posted more info on this from the perspective of I2S interfaces here:
IanCanada's Latest RPi GB Goodies Impressions... and your tweaks, mods and hints...
AND Markw4 followed with his view, providing some additional mechanisms, but with the same results here:
IanCanada's Latest RPi GB Goodies Impressions... and your tweaks, mods and hints...
I know of some development efforts in audio attempting to develop new interfaces that WILL eliminate any upstream influence. The first application of this I know of will be in the soon-to-be released Uptone Audio EtherRegen. While it is a historic and heroic attempt towards this goal, even they say "we'll see", though in my conversations with them they've been a lot more positive about the results.
But the EtherRegen isn't available yet. AND it is designed to provide Ethernet isolation, which still leaves us using an RPi-based endpoint to produce the I2S stream going into our DACs. Ultimately I expect Uptone will extend that to their USB devices, but AFAIK they have no plans to do so for an I2S-connection that would be useful for us DIY'ers.
SO until an entirely new generation of I2S interface becomes available, the USBridge Signature will likely be the best RPi-style endpoint for an I2S-connected DAC. I know Allo is also intending it to be the best USB endpoint and I hope you achieve that too. That is where most of the audio market is today. I will say I've been watching the development of the latest Sonore optical Ethernet modules and OpticalRendu and it seems they've made significant strides with their latest offerings. Still, where I believe you will price the USBridge Signature will make it VERY attractive in comparison!
But since my DAC setups all use I2S feeds, that's what I'm focused on. AND it will be the only game in town for improving that realm! I can hardly wait to get my hands on one!
Finally, to @Soundcheck, with disrespectful, demeaning, and denigrating comments such as:
- "The new question if the SBC suddenly should now be in charge to improve THD-N on the audio interface is nonsense."
- "That's nonsene. These "testers" are simply clueless."
- "I'm not commenting on Gregs wild fishing exercises."
Why would we even engage with you? I think cdsgames is a saint for doing so... me I have no time for someone who leads with such dismissive comments. AND I've been working with power upgrades to improve my computer audio as long as you have on the Op System and SW sides (AND I haven't ignored those areas in my various implementations as they ARE important, but can't fix inherent issues with poor and noisy power).
Greg in Mississippi
Exactly,
even with EtherRegen and opticalRendu the source upstream of these devices will have an impact on the SQ at the output of the DAC.
The nice thing of the USBridge Signature is that it delivers a high quality output for both I2S and USB.
Matt
"disrespectful" - Oh yeah.
very disrespectful!!!Try to understand what's been said. If you have a problem, simply ask.
And don't just copy snippets and put them out of context to support your wailing.
Again. If you have a question - ask. And again don't put snippets wildly out of context.
Oh dear.
Somebody got a hanky. What do you think you're doing?
Swapping countless parts and boards spending thousands of $ and verifying the results by ear to me is nothing else then fishing.
Or do you want me to call this engineering? Or perhaps even scientific research?
No. Greg, what you do is IMO - fishing.
That's what most of us do - btw. You're not alone!
Look. I also have offline talks with cdsgames. I even support them on improving or solving issues on their products. Just happened last week.
I do have an engineering background. If I think somebody has it (slightly) wrong I'll address it.
Enjoy. Just call if you need some more hankies.
Oh dear. Sometime I get the feeling I'm surrounded by toddlers.
The latest attack on me by Greg was not about Allo or cdsgames. This was all about Gregs ego. Once in a while he gets his frustration out this way.
He doesn't realize that he's the only one running these attacks. Look at his ridiculous quotes (taken out of context) and what attributes towards me he ties to them.
If somebody around here has a problem. It's him.
Regarding Allo.
I very much appreciate what Allo is doing and I'm also supporting them on that. And that for a much longer time then many of the people around.
All I do is calling out Allo/cdsgames on their statements which I think doesn't make sense, are inconsistent or which I simply consider wrong.
And I usually do that in a constructive manner. I do give examples and references. I even write testcode for them.
But if I get a "it's good enough - trust me" and later on I get a "I don't know how to measure transient responses or rise times" from somebody who's designing highest quality power supplies, I'll respond to that. And I'll continue to do so.
Enjoy.
Soundcheck , I know you are trying to help .
You are asking me for transient times and rise times . I don't see any manufacturers testing their units (IFI or Sbooster ) .
As I said , I will look for a tester and provide them when possible . At last , can you show me any data correlating THD and transient time ? There is none I believe. Transients must be dealt with locally by using enough capacitance on PCB or output of PSU.We did that.
We are all here to learn and improve..right ?
My test data (and I have more) shows to anyone that THD+N changes depending on the i2s (and USB) source. (using same jitter and PSU)
What I found is that harmonics change significantly (while overall THD+N changes slightly). That might explain the difference in SQ heard by a few people (since we know that harmonics should be weighted differently , Mark knows about it..)
Do you (or anyone ) has a theory about that ?
Lets put our heads together and try to solve one of the many mysteries of audio. I for one , I will experiment with i2s line impedance into the DAC IC and see if I can find the answer.
You are asking me for transient times and rise times . I don't see any manufacturers testing their units (IFI or Sbooster ) .
As I said , I will look for a tester and provide them when possible . At last , can you show me any data correlating THD and transient time ? There is none I believe. Transients must be dealt with locally by using enough capacitance on PCB or output of PSU.We did that.
We are all here to learn and improve..right ?
My test data (and I have more) shows to anyone that THD+N changes depending on the i2s (and USB) source. (using same jitter and PSU)
What I found is that harmonics change significantly (while overall THD+N changes slightly). That might explain the difference in SQ heard by a few people (since we know that harmonics should be weighted differently , Mark knows about it..)
Do you (or anyone ) has a theory about that ?
Lets put our heads together and try to solve one of the many mysteries of audio. I for one , I will experiment with i2s line impedance into the DAC IC and see if I can find the answer.
USB-C as a power adapter may not be the best way to go....Raspberry Pi admits to faulty USB-C design on the Pi 4 | Ars Technica
The Pi 4 just uses the port as a way to connect to a power supply, but its not configured as the official spec so can present problems. Does USbridge sig use usb-c as power connection?
The Pi 4 just uses the port as a way to connect to a power supply, but its not configured as the official spec so can present problems. Does USbridge sig use usb-c as power connection?
Certainly USB-C per se is not the problem but the correct wiring according to the official specs.
I have a Khadas Tone Board which uses power via USB-C as well. It seems to work without any issues so far.
Matt