Yes, but I would like to:
- have three inputs USB (PC upsampled to DSD), Toslink (TV), Coax (Tidal from Rpi3).
- I would like skip asynch mode (100mhz clock turned off)
- all inputs to be isolated by hermes, then reclocked by cronus then to buffalo working in sychrnouse mode.
Such board with those three inputs would be great addition to your product portfolio.
You could have different inputs USB, Toslink, Coax, all isolated, all reclocked and Buffalo working in sychronous mode.
Having used these chips longer than probably most anyone on the planet My opinion is that the very best solution for SPDIF will always be to run async. In fact for that setup you will be doing yourself a huge favor not to bother with sync mode at all
There are several impediments to success in such a scheme. First and most important is that real success with Cronus requires its sources be able to accept it's master clock - your SPDIF sources can't... You would need a FIFO (which will incur a lag - not good for live audio or watching video)
So one cool option is that the ES9028/38 allows you to assign GPIOs to SPDIF input. The default firmware could be altered to allow for that - then you could have 2 SPDIF sources (one consumer level coax, and one TTL/CMOS) without needing any external switching.
If anyone is interested I can create a branch that does that.So in short - if you are expecting SPDIF by all means you can still use Cronus. Cronus will provide you a super clean jitter free source - even when you run the DAC itself Async!!! This is not a bad way to go at all!
Cheers!
Russ
Hi there, this is a humble report from a user quite satisfied with this new B3SE 9028pro which I'd like to say just amazing!
After receiving the package three days ago and finishing the soldering work on the next day, the old B3SE board was simply replaced with the new 9028pro board without any alteration of the PS units to the BBB, Hermes/Cronus and IVY-III other than checking the supply voltage to the 9028pro was 5.04V (my LCDS did not allow to adjust the voltage to 5.00V as Russ suggested).
After setting the SW1-1 to serial, I turned on the DAC and confirmed that it flawlessly and beautifully played the DSD512 sources upsampled by HQ player, which were transported by the BBB/Botic system working as an NAA for HQP and Hermes/Cronus boards (90/98 clocks).
Though merely 24 hours after starting music listening with this B3SE 9028pro, I'd say that the quality of sound is superb in terms of the range, depth, resolution and balance.
I said "flawlessly", but the first attempt with the default DPLL setting went to cause some stuttering in playing music. This was quickly solved by changing the bandwidth to the highest (0b1111). As Russ wrote somewhere recently, strict setting of bandwidth may not be a mandate for 9028/9038 family any more.
Though I rarely use SPDIF, I had prepared a button for SPDIF/serial switch on the front panel of my chassis designed for the old B3SE and thankfully it still works well by wiring to A0 and GND on the external I/O connector as suggested by Russ.
That said, I'd like to thank Russ and Brian with a great respect for designing this new, wonderful DAC board with perfect compatibility with the old one and smart emphasis on the characteristics of the new DAC chip.
(BTW, the red PCB in the picture is a dual mono AK4497 board, quite comparable to 9028/9038pro family but been sadly waiting for being replaced with B3SEpro
).
Regards,
After receiving the package three days ago and finishing the soldering work on the next day, the old B3SE board was simply replaced with the new 9028pro board without any alteration of the PS units to the BBB, Hermes/Cronus and IVY-III other than checking the supply voltage to the 9028pro was 5.04V (my LCDS did not allow to adjust the voltage to 5.00V as Russ suggested
).After setting the SW1-1 to serial, I turned on the DAC and confirmed that it flawlessly and beautifully played the DSD512 sources upsampled by HQ player, which were transported by the BBB/Botic system working as an NAA for HQP and Hermes/Cronus boards (90/98 clocks).
Though merely 24 hours after starting music listening with this B3SE 9028pro, I'd say that the quality of sound is superb in terms of the range, depth, resolution and balance.
I said "flawlessly", but the first attempt with the default DPLL setting went to cause some stuttering in playing music. This was quickly solved by changing the bandwidth to the highest (0b1111). As Russ wrote somewhere recently, strict setting of bandwidth may not be a mandate for 9028/9038 family any more.
Though I rarely use SPDIF, I had prepared a button for SPDIF/serial switch on the front panel of my chassis designed for the old B3SE and thankfully it still works well by wiring to A0 and GND on the external I/O connector as suggested by Russ.
That said, I'd like to thank Russ and Brian with a great respect for designing this new, wonderful DAC board with perfect compatibility with the old one and smart emphasis on the characteristics of the new DAC chip.
(BTW, the red PCB in the picture is a dual mono AK4497 board, quite comparable to 9028/9038pro family but been sadly waiting for being replaced with B3SEpro
).Regards,
Attachments
Hello Russ,
I thought the ES90x8 was pretty much immune to source jitter thanks to its ASRC. So a clean source is still beneficial? Would something like a mutec mc3 be of interest between a jittery source and a BIII spidf input?
Thank you
Mislabeled Tridents in kit
Checking the 9038PRO kit today I discovered the Trident reg in the 1v2 bag is marked as 3v3!?
I pulled one of the 3v3 regs out of the BII board and plugged in each of the new kit Tridents to measure the outputs. One of them is 1v2 and the others are 3v3.
So, all ok, just so you know.
I pulled one of the 3v3 regs out of the BII board and plugged in each of the new kit Tridents to measure the outputs. One of them is 1v2 and the others are 3v3.
So, all ok, just so you know.
Checking the 9038PRO kit today I discovered the Trident reg in the 1v2 bag is marked as 3v3!?
I pulled one of the 3v3 regs out of the BII board and plugged in each of the new kit Tridents to measure the outputs. One of them is 1v2 and the others are 3v3.
So, all ok, just so you know.
Huh. Okay, you are the second person to report this. I suspect there may be another couple of folks with the same issue, as I build then in panels of 5.
Having used these chips longer than probably most anyone on the planet My opinion is that the very best solution for SPDIF will always be to run async. In fact for that setup you will be doing yourself a huge favor not to bother with sync mode at all
There are several impediments to success in such a scheme. First and most important is that real success with Cronus requires its sources be able to accept it's master clock - your SPDIF sources can't... You would need a FIFO (which will incur a lag - not good for live audio or watching video)
So one cool option is that the ES9028/38 allows you to assign GPIOs to SPDIF input. The default firmware could be altered to allow for that - then you could have 2 SPDIF sources (one consumer level coax, and one TTL/CMOS) without needing any external switching.
So in short - if you are expecting SPDIF by all means you can still use Cronus. Cronus will provide you a super clean jitter free source - even when you run the DAC itself Async!!! This is not a bad way to go at all!
Cheers!
Russ
Ok. Got it. Thanks Russ. I have similar setup right now:
- some 35$ ebay board with toslink, coax, i2s input based on es9038q2m (asynch mode)
- rpi3 + hifiberry digi+ pro connected to this board by i2s. Working on Volumio.
I just thought that putting spdif conversion to I2S off the sabre chip would benefit and also making everything working in synch mode (skipping DPLL) would bring further improvement.
Hi Russ
For those of us looking to possibly use an external controller via i2c, is there any non NDA information available regarding the chip registers/programming?. I understand that the IO pins could be readily used for many functions but being able to use GPIO connections for inputs as well as getting data back from the chip would be fun to play around during with.
Thanks
Martin
For those of us looking to possibly use an external controller via i2c, is there any non NDA information available regarding the chip registers/programming?. I understand that the IO pins could be readily used for many functions but being able to use GPIO connections for inputs as well as getting data back from the chip would be fun to play around during with.
Thanks
Martin
Take a look at the header files here:
GitHub - twistedpearaudio/Buffalo-III-SE-Pro-On-Board-Firmware
The ES9028/38 register mapping is nearly complete. I will be updating it soon to make it more complete.
Actually I believe with ~100Mhz clocks you can get up to DSD1024 - though I have not tried that myself - the datasheet says you can
And you can also do up to 768Khz PCM.
You are limited to DSD256 via the DoP decoder if you use it.
Cheers!
Russ
But, why?
Buffalo III SE (Stereo Edition) Pro
When using output stages like the Mercury, Ivy, Legato that mount directly to the Buffalo - it is also recommended you solder the female analog output headers to the bottom of the board and the male (long tail) headers to the output stage.
Buffalo III SE (Stereo Edition) Pro
When using output stages like the Mercury, Ivy, Legato that mount directly to the Buffalo - it is also recommended you solder the female analog output headers to the bottom of the board and the male (long tail) headers to the output stage.
Preference, I have just found it easier to line up the modules when assembling that way. I have built a lot of these, and that's what I have found to be the easiest. There is nothing that will go wrong if you don't follow the advice. Do what you find easiest for you, this is DIY.
There is nothing that will go wrong if you don't follow the advice. Do what you find easiest for you, this is DIY.
Last edited: