Hi
yes I know about the RPI4 and one unit will be under testing shortly
From what I know from previous generations...every time a faster and more power hungry CPU comes out, the noise floor is raised .
Second, between a new RPI and a CM launch there is a delay from 3-8
months
Third , software wise it might take some time to figure all the bugs out
Forth and most important. The new Usbridge is made for audio, We are already achieving 330Mbs . more than enough for DSD 1024
I understand the need for latest and fastest...but , again, for audio we have plenty of horsepower and bandwidth.
Our Usbridge Sig/Audiophile SBC is going full ahead
Previous generations of CM were made compatible witth eachother...but frankly until I see the schematic I cannot guarantee it .
What I do see its USB3 and a USB 2 inputs (usb 3 input was not available on the previous CM)
I agree, with the configuration you have described for usbridge utilizing the compute module CM3+ you have taken care of usb/ethernet & i/o and many of the weaknesses of the Pi. The only area where the 4 wins out is the 2 or 4 Gb of ddr4 RAM, which should impact performance. I'm sure if CM4 is developed it should be backwards compatible and hopefully a potential upgrade path. It would be interesting to evaluate benefits of the Pi4 with Digione/Katana etc., if there is an advantage it should be an easy upgrade path for users.
Whatever the outcome we know that power supplies are a critical component of overall performance so we are eagerly waiting on Shanti & Nirvana.
Whatever the outcome we know that power supplies are a critical component of overall performance so we are eagerly waiting on Shanti & Nirvana.
No doubt RPI4 is an very good SBC
Will it be a better SBC for audio compared with 3B or 3B+ ?
No. (but most likely you will get DSD 512 out of it)
I measured 3A , 3B and 3B+ generation of PI and each jump in speed shows a small loss of THD+N on a DAC. (boss)
However I understand the need to have the latest and "fastest" . Heck get both
Seriously , RPI4 (like RPI3 before) is not meant for audio . I am pretty sure that comparisons between the USBridge Sig and RPI4 will pop up . I am also rather confident that we will win this head to head .. rather easy.
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)
Will it be a better SBC for audio compared with 3B or 3B+ ?
No. (but most likely you will get DSD 512 out of it)
I measured 3A , 3B and 3B+ generation of PI and each jump in speed shows a small loss of THD+N on a DAC. (boss)
However I understand the need to have the latest and "fastest" . Heck get both
Seriously , RPI4 (like RPI3 before) is not meant for audio . I am pretty sure that comparisons between the USBridge Sig and RPI4 will pop up . I am also rather confident that we will win this head to head .. rather easy.
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)
Hi
as per diyaudio forums rules :
COMMERCIAL SECTOR AND VENDORS BAZAAR RULES
These rules apply to any members of our commercial sector as well as anyone promoting anything on diyAudio. Repeated violations of these rules will result in termination of your account and all your threads without hesitation.
I am unable to open multiple threads for our products
as per diyaudio forums rules :
COMMERCIAL SECTOR AND VENDORS BAZAAR RULES
These rules apply to any members of our commercial sector as well as anyone promoting anything on diyAudio. Repeated violations of these rules will result in termination of your account and all your threads without hesitation.
- No promotion or sale of products infringing trademarks, regardless of the country of origin of the trademark.
- Only one thread per vendor in the Vendors Bazaar
- No thread "bumping" in the Vendors Bazaar
I am unable to open multiple threads for our products
Nirvana updates .
The path to Nirvana in Buddhism ...is long , full of questions and wrong turns.
When achieved , it promises tranquility and understanding of truth.
At a much lower level , Nirvana SMPS , has been a titanic undertaking full of questions and wrong turns ,we were ready to give up a few times . Yet , we persevered .
What we achieved , is rather spectacular . Noise in a SMPS that equals a very good LPS. We are talking about 150nV on 0-20Khz. First reading is at 3A then 2A and 1A. Perfect for RPI4, USbridge Sig , anything really .
Note also that those test/ numbers are in real world scenario (no isolations transformers were used)
Capacitance has been increased to 9000uF on DC side (low impendance )
Attached , the latest noise figures with the new PCB. 500 PCBs have arrived today , components in stock. alu casing (waiting) but not much delay.
Please note you have a SW that will make the DC side earthed (ground earthed)
I am expecting to have them ready for sale on July 1. Pricing for the unit will be 59$ ( 59EUR + VAT in EU )
Preorders will be open 1 week in advance.
The path to Nirvana in Buddhism ...is long , full of questions and wrong turns.
When achieved , it promises tranquility and understanding of truth.
At a much lower level , Nirvana SMPS , has been a titanic undertaking full of questions and wrong turns ,we were ready to give up a few times . Yet , we persevered .
What we achieved , is rather spectacular . Noise in a SMPS that equals a very good LPS. We are talking about 150nV on 0-20Khz. First reading is at 3A then 2A and 1A. Perfect for RPI4, USbridge Sig , anything really .
Note also that those test/ numbers are in real world scenario (no isolations transformers were used)
Capacitance has been increased to 9000uF on DC side (low impendance )
Attached , the latest noise figures with the new PCB. 500 PCBs have arrived today , components in stock. alu casing (waiting) but not much delay.
Please note you have a SW that will make the DC side earthed (ground earthed)
I am expecting to have them ready for sale on July 1. Pricing for the unit will be 59$ ( 59EUR + VAT in EU )
Preorders will be open 1 week in advance.
Today testing shows that USBridge draws about 550mA. So its the same, you can use your PSU.. Still note that on power on there is a spike of about 850mA for 200mS
As long as the forum can't offer you a vendor subforum they should'n t
block you from opening discussions around your new products. You've been asking for it more then once if I'm not mistaken.
From what I see Allo is one of the companies most active and most appreciated by todays crowd.
Most subforums in the Manufacturer area are idling around for years.
That'd at be a shame if the forum management would act as cdsgames suggests.
block you from opening discussions around your new products. You've been asking for it more then once if I'm not mistaken.
From what I see Allo is one of the companies most active and most appreciated by todays crowd.
Most subforums in the Manufacturer area are idling around for years.
That'd at be a shame if the forum management would act as cdsgames suggests.
@cdsgames
I disagree with you on several accounts related to RPI.
I do understand that you're facing tough times with the recent PI4 announcement.
You had to learn it the hard way already how tough it is to offer your own SBC board. Keeping up the pace on the HW is one thing. That's quite challenging already. Add the SW and driver part to it and you'll feel the heat slowly creeping up. Good luck on that part.
Looking back I don't agree with cdsgames. Pretty much every PI generation evolved substantially on numerous accounts. And the new Pi4 seems
to be a new milestone - for several reasons.
Who'd care if a little noise gets added. (which has to be shown first - btw)
I know cdsgames is fully driven by the noise-mania. IMO noise would not even be the crucial factor here.
To achieve these new RPI performance levels the whole SBC needed to be redesigned and improved. That's what the PI folks have done. And I'm sure that covers also noise EMI/RFI and the power rails.
Also keep in mind that the GPIO performance has been dramatically improved. That should have a slightly positive effect on I2S.
And there'd be more, much more.
Looking back.
In the past I managed to shutdown pretty much all parts of the RPI
that weren't needed. And that included the entire USB stack on the RPI 3B+.
That left me with a 220mA idle power consumption. If we talk about noise we need to be very specific about what setup we're talking about.
I could go on and on.
That simplistic "noise" view by cdsgames I simply won't share.
Not to forget. It's the audio interface which is in charge to get rid of the noise. And at levels <-110dB noise who'd care about it accept the manufacturer and some benchmarkers.
Enjoy.
I disagree with you on several accounts related to RPI.
I do understand that you're facing tough times with the recent PI4 announcement.
You had to learn it the hard way already how tough it is to offer your own SBC board. Keeping up the pace on the HW is one thing. That's quite challenging already. Add the SW and driver part to it and you'll feel the heat slowly creeping up. Good luck on that part.
Looking back I don't agree with cdsgames. Pretty much every PI generation evolved substantially on numerous accounts. And the new Pi4 seems
to be a new milestone - for several reasons.
Who'd care if a little noise gets added. (which has to be shown first - btw)
I know cdsgames is fully driven by the noise-mania. IMO noise would not even be the crucial factor here.
To achieve these new RPI performance levels the whole SBC needed to be redesigned and improved. That's what the PI folks have done. And I'm sure that covers also noise EMI/RFI and the power rails.
Also keep in mind that the GPIO performance has been dramatically improved. That should have a slightly positive effect on I2S.
And there'd be more, much more.
Looking back.
In the past I managed to shutdown pretty much all parts of the RPI
that weren't needed. And that included the entire USB stack on the RPI 3B+.
That left me with a 220mA idle power consumption. If we talk about noise we need to be very specific about what setup we're talking about.
I could go on and on.
That simplistic "noise" view by cdsgames I simply won't share.
Not to forget. It's the audio interface which is in charge to get rid of the noise. And at levels <-110dB noise who'd care about it accept the manufacturer and some benchmarkers.
Enjoy.
OK lets see what they say, I am just making sure that we fallow the rules.
DoctorMord , the IC itself has short protection , overcurrent protection , oveertemperature Creeping distances were taken into consideration on both casing and secondary side as per standards.
FCC/CE testing will happen in about 2 weeks (so yes we will have it)
Transients testing its not done , since frankly we need a very specific machine and we dont have it. Still most SMPS with feedback have good transient response. Of course the 9000uF should help in that regards .
DoctorMord , the IC itself has short protection , overcurrent protection , oveertemperature Creeping distances were taken into consideration on both casing and secondary side as per standards.
FCC/CE testing will happen in about 2 weeks (so yes we will have it)
Transients testing its not done , since frankly we need a very specific machine and we dont have it. Still most SMPS with feedback have good transient response. Of course the 9000uF should help in that regards .
@cdsgames
How does Nirvana look above 20kHz?
Will it come with cable and USB output? If I recall it correctly the cabling option on Shanti looked much more DIY friendly then Nirvana!?!?
Can the pot be used to adjust the output voltage? Range?
From what I understand there won't be a buffer at the end of the cable right?
The cable length would than heavily impact the transient response - wouldn't it!?!? (I pretty much expect though you'll continue not to mention anything related to the transient responses of your PSes.)
Thx.
How does Nirvana look above 20kHz?
Will it come with cable and USB output? If I recall it correctly the cabling option on Shanti looked much more DIY friendly then Nirvana!?!?
Can the pot be used to adjust the output voltage? Range?
From what I understand there won't be a buffer at the end of the cable right?
The cable length would than heavily impact the transient response - wouldn't it!?!? (I pretty much expect though you'll continue not to mention anything related to the transient responses of your PSes.)
Thx.
Hi Soundcheck
my simplistic view of noise , comes from testing
I dont think that RPI4 is a bad SBC , I think its great one . The IO additions , direct Gb ethernet , USB3 makes a lot of sense for normal uses.
Our USBridge takes a different approach to SBCs
We concentrated on noise (the simple one). Look at the USB hub from the RPI4...do you see any dedicated LDOs ? Look at the Ethernet , do you think they have a dedicated power rail . or maybe the use whatever 3.3V goes to the CPU ?
Look at the CPU is it feed by a switcher or LDOs ? How many layers RPI4 has ..minimum (4) or as Otto is advising (10 for best EMI mitigation)
USBridge is going full steam ahead for those of you place a premium on sound quality and not on how fast the CPU or Eternet is.
As I said , we have plenty of power and speed beyond DSD 512.
Soundcheck , our unit is not competing with RPI4 . We want to beat in SQ the SOTM , uRendu Ultra and others (that by the way have lower CPUs speeds and rather equal ethernets)
But frankly speaking , you have been around for a while . You still think that RPIs are good for audio ? What did the Kali , Boss and others shown...its crappy at best.
Simply put...RPIs are not designed for audio but for a general purpose SBC. (and they do brilliantly at that)
Faster is not better (especially in audio) take a look at the Ferrari vs Mercedes F1 this year (Ferrari are faster in the straight line and yet they are way behind in constructors race)
Usbridge Sig is going full ahead . We increased the order of PCB on the first batch. We are also in talks with most reviewers for head to heads vs best streamers on the market
Come on Klaus , you must joke when suggesting that RPI4 will be comparable to USbridge Sig . Mitigated EMI /RF ? Its still the same standard , they made sure that they pass it .
In any case , we will release soon and everyone can add their voices .
my simplistic view of noise , comes from testing
I dont think that RPI4 is a bad SBC , I think its great one . The IO additions , direct Gb ethernet , USB3 makes a lot of sense for normal uses.
Our USBridge takes a different approach to SBCs
We concentrated on noise (the simple one). Look at the USB hub from the RPI4...do you see any dedicated LDOs ? Look at the Ethernet , do you think they have a dedicated power rail . or maybe the use whatever 3.3V goes to the CPU ?
Look at the CPU is it feed by a switcher or LDOs ? How many layers RPI4 has ..minimum (4) or as Otto is advising (10 for best EMI mitigation)
USBridge is going full steam ahead for those of you place a premium on sound quality and not on how fast the CPU or Eternet is.
As I said , we have plenty of power and speed beyond DSD 512.
Soundcheck , our unit is not competing with RPI4 . We want to beat in SQ the SOTM , uRendu Ultra and others (that by the way have lower CPUs speeds and rather equal ethernets)
But frankly speaking , you have been around for a while . You still think that RPIs are good for audio ? What did the Kali , Boss and others shown...its crappy at best.
Simply put...RPIs are not designed for audio but for a general purpose SBC. (and they do brilliantly at that)
Faster is not better (especially in audio) take a look at the Ferrari vs Mercedes F1 this year (Ferrari are faster in the straight line and yet they are way behind in constructors race)
Usbridge Sig is going full ahead . We increased the order of PCB on the first batch. We are also in talks with most reviewers for head to heads vs best streamers on the market
Come on Klaus , you must joke when suggesting that RPI4 will be comparable to USbridge Sig . Mitigated EMI /RF ? Its still the same standard , they made sure that they pass it .
In any case , we will release soon and everyone can add their voices .
Soundcheck , can you show me any transient response in any commercial SMPS ?
I said already we dont have the test machine/data to show ...so its not possible to show it. Its not that we want to "hide the data" , its not something that its standard for power supply . If I can get my hands on one , I will gladly take a test report.
Our AP machine only goes to 30Khz , over that we have to use CRO (with far less precision) that shows maximum noise 20Mhz bandwidth at around 70uV.
Voltage output is 5.2V at minimum load , 5V at 3A
USB connector is included , POT is set in factory , any change you will not get full benefit but you might add up to 0.3V to the output.
Front plate is removable and new cables can be soldered...but its not something that we recommend for most people.
I looked at other vendors in VENDORS BAZAAR.
There are quite a few who have multiple open threads in this section.
Matt
I refereed you to PSes in the past. Every high quality lab supply has these parameters in the datasheet.
The only thing you deliver: "It'll be good enough. Trust us!"
How can I trust you, if you don't even know how to measure these important parameters!?!?
Eeeh. Every stupid scope could do.
I do get the impression.
Because I do think that you've got a bunch of capable engineers in your team.
BTW.
Comparing Shanti and Nirvana. What are these nasty spikes in the 5k+ range on Nirvana?