Hi there.
Did I tell you that I switched to the Piano 2.1 in dual-mono + Kali on my main system for the time being.
After a little (actually heavy) SW tweaking I had the impression that I don't have to go back
to my highly tweaked Sabre 9023 DAC (which i still prefer over my Boss btw).
Optimizations on the OS side do make a difference in my case, even with the Kali in place and with Kali supplying the master clock.
The experienced differences are the main factor that let me stay with the Piano21.
I do also feed the DAC with synchronous upsampled material @ 352k8/384 to bypass the onboard filters.
And I bypassed the RCA jacks. Soldering the wires to the board is ,as usual, a must!
Bottom line. Just running the DAC "as is" unde Moode/Volumio/PcP/Rune is by far (in audiophile terms) not the
end of the story. There's more to gain.
Enjoy. I do.
Did I tell you that I switched to the Piano 2.1 in dual-mono + Kali on my main system for the time being.
After a little (actually heavy) SW tweaking I had the impression that I don't have to go back
to my highly tweaked Sabre 9023 DAC (which i still prefer over my Boss btw).
Optimizations on the OS side do make a difference in my case, even with the Kali in place and with Kali supplying the master clock.
The experienced differences are the main factor that let me stay with the Piano21.
I do also feed the DAC with synchronous upsampled material @ 352k8/384 to bypass the onboard filters.
And I bypassed the RCA jacks. Soldering the wires to the board is ,as usual, a must!
Bottom line. Just running the DAC "as is" unde Moode/Volumio/PcP/Rune is by far (in audiophile terms) not the
end of the story. There's more to gain.
Enjoy. I do.
Last edited:
@cdsgames
I just scanned through the thread once more.
Can you elaborate why you went from 9028 to 9038Q2M?
I've been sneaking around in iancanadas DAC thread.
He listed a chart saying that the 9028 has half the output impedance
and double current capabilities compared to 9038 - which is usually
not a bad thing!?!?
I just scanned through the thread once more.
Can you elaborate why you went from 9028 to 9038Q2M?
I've been sneaking around in iancanadas DAC thread.
He listed a chart saying that the 9028 has half the output impedance
and double current capabilities compared to 9038 - which is usually
not a bad thing!?!?
The question just popped up. Perhaps we discussed this earlier. I forgot.
Anybody out there having a clue about:
1.
The pcm5142 datasheet doesn't mention 352.8kHz as supported samplerate
Just 384kHz is listed.
That's obviously the reason why 352.8 is not to be found inside the codec driver. ( many of us patched it in though )
see e.g. DS page 48
I can feed 352.8 though (because of the patch and no complaints from the DAC side). There doesn't "seem" to be any hickups/sideffects.
What's happening inside the DAC then!?!?
Or are the DSs flawed?
2.
I am also wondering about the background of 0.8db "Analogue Boost"
On page 45 I just found something about "Analog Gain Control".
Which would would be the mixer control "Analogue".
It toggles the output voltage between 1V or 2V.
What is then meant by that "Analogue Boost" control?
PCM5142 DS
Thx.
Anybody out there having a clue about:
1.
The pcm5142 datasheet doesn't mention 352.8kHz as supported samplerate
Just 384kHz is listed.
That's obviously the reason why 352.8 is not to be found inside the codec driver. ( many of us patched it in though
)see e.g. DS page 48
I can feed 352.8 though (because of the patch and no complaints from the DAC side). There doesn't "seem" to be any hickups/sideffects.
What's happening inside the DAC then!?!?
Or are the DSs flawed?
2.
I am also wondering about the background of 0.8db "Analogue Boost"
On page 45 I just found something about "Analog Gain Control".
Which would would be the mixer control "Analogue".
It toggles the output voltage between 1V or 2V.
What is then meant by that "Analogue Boost" control?
PCM5142 DS
Thx.
Hmmh. It's because of the datasheet !?!? What does that mean ??
AND. It's because of DOP. That I understand.
Or do you just mean:
ESS offers DOP as a feature for the 9038. And you put priority to that over lower output Z and higher output current.
Last edited:
One more.
I just wanted to mention it.
Audiophonics made an i-Sabre ES9028Q2M available to me several weeks ago to do some ""DOP"" testing for Moode (Thx for that).
And... ...DOP works quite nice on the 9028.
The IMO greatest feature is the Sabre SW volume control, which still works
with DOP. SW Volume control on DSD material can get quite tricky.
(I integrated their codec&driver into the kernel - it's not commonly available - obviously this on-dac SVC feature won't work without custom driver.)
The other nice thing with that Audiophonics DAC is the single 5V supply option. No +/-15V for the output stage is required. If that's the best sounding option !?!? I don't know.
However. You can't separate the DAC supply from the PI supply rail for that DAC. Hmmh. After HAT DACs being around and discussed for years, they should have known better.
I then tried to improve the clocking/I2S side. Unfortunately I gotta be too stupid. I didn't manage to get that 9028 DAC working properly with Kali in my environment. And that's a pity.
****
What I really liked about my old Sabre9023 and Kali was the synchronous clocking from Kali. It was quite an improvement over the single onboard clock for the Sabre. Hopefully you played around with this feature for Katana.
I'm pretty sure you did.
Looking forward to Katana. Knowing you guys, I'm certain you will put all your experiences into the project.
****
After Katana, you then can focus on a 9038Pro HAT project for 2018.
I just wanted to mention it.
Audiophonics made an i-Sabre ES9028Q2M available to me several weeks ago to do some ""DOP"" testing for Moode (Thx for that).
And... ...DOP works quite nice on the 9028.
The IMO greatest feature is the Sabre SW volume control, which still works
with DOP. SW Volume control on DSD material can get quite tricky.
(I integrated their codec&driver into the kernel - it's not commonly available - obviously this on-dac SVC feature won't work without custom driver.)
The other nice thing with that Audiophonics DAC is the single 5V supply option. No +/-15V for the output stage is required. If that's the best sounding option !?!? I don't know.
However. You can't separate the DAC supply from the PI supply rail for that DAC. Hmmh. After HAT DACs being around and discussed for years, they should have known better.
I then tried to improve the clocking/I2S side. Unfortunately I gotta be too stupid. I didn't manage to get that 9028 DAC working properly with Kali in my environment. And that's a pity.
****
What I really liked about my old Sabre9023 and Kali was the synchronous clocking from Kali. It was quite an improvement over the single onboard clock for the Sabre. Hopefully you played around with this feature for Katana.
I'm pretty sure you did.
Looking forward to Katana. Knowing you guys, I'm certain you will put all your experiences into the project.
****
After Katana, you then can focus on a 9038Pro HAT project for 2018.
One more.
I just wanted to mention it.
Audiophonics made an i-Sabre ES9028Q2M available to me several weeks ago to do some ""DOP"" testing for Moode (Thx for that).
And... ...DOP works quite nice on the 9028.
The IMO greatest feature is the Sabre SW volume control, which still works
with DOP. SW Volume control on DSD material can get quite tricky.
(I integrated their codec&driver into the kernel - it's not commonly available - obviously this on-dac SVC feature won't work without custom driver.)
The other nice thing with that Audiophonics DAC is the single 5V supply option. No +/-15V for the output stage is required. If that's the best sounding option !?!? I don't know.
However. You can't separate the DAC supply from the PI supply rail for that DAC. Hmmh. After HAT DACs being around and discussed for years, they should have known better.
I then tried to improve the clocking/I2S side. Unfortunately I gotta be too stupid. I didn't manage to get that 9028 DAC working properly with Kali in my environment. And that's a pity.
****
What I really liked about my old Sabre9023 and Kali was the synchronous clocking from Kali. It was quite an improvement over the single onboard clock for the Sabre. Hopefully you played around with this feature for Katana.
I'm pretty sure you did.
Looking forward to Katana. Knowing you guys, I'm certain you will put all your experiences into the project.
****
After Katana, you then can focus on a 9038Pro HAT project for 2018.
The Audiophonics 9028 uses a FPGA for the DOP. The 9038 has DOP incorporated in DAC IC. As far as impedance I am not aware of any differences between the 9028q2m and 9038q2m.
Yes we took in consideration to feed it only one 5V...but we do make +- 15v on a separate PCB (last one). You need good voltage swing for best sound on those discrete opamps.
At last , I think that Audiophonics found the issues of the 9028 DAC and Kali...simply it cannot take mclk from clocks at 45/49Mhz , we made some custom Kalis with 22/24Mhz for them.
Great. That's very interesting information. It clarifies this or that.
But than I'm asking myself if the volume control is done on DOP converted to PCM before the DAC inside that AP FPGA !?!?
Ok. Ok. It's not any of your concern.
Where/how does your 9038Q2M do the DOP software volume control. Is there SW VC for DOP at all on the 9038Q2M??
Regarding Kali.
...and I was thinking I'm too stupid to get this AP DAC properly going.
Thx again for the heads up. Always nice talking to you.
Here the link where IANCANADA posted his impedance findings.
However. A guy who has been trying the 9038Q2M questioned the impedance numbers, because he made a transformer work. And that one didn't work on a 9018 with pretty much the same output specs as the 9038 -- at least as the ones communicated by IANCANADA. I guess there's some confusion in the air.
But again. From my perspective this won't be an issue at all. I was just curious.
But than I'm asking myself if the volume control is done on DOP converted to PCM before the DAC inside that AP FPGA !?!?
Ok. Ok. It's not any of your concern.
Where/how does your 9038Q2M do the DOP software volume control. Is there SW VC for DOP at all on the 9038Q2M??
Regarding Kali.
...and I was thinking I'm too stupid to get this AP DAC properly going.
Thx again for the heads up. Always nice talking to you.
Here the link where IANCANADA posted his impedance findings.
However. A guy who has been trying the 9038Q2M questioned the impedance numbers, because he made a transformer work. And that one didn't work on a 9018 with pretty much the same output specs as the 9038 -- at least as the ones communicated by IANCANADA. I guess there's some confusion in the air.
But again. From my perspective this won't be an issue at all. I was just curious.
Last edited:
From what I know...all DSD is converted back to PCM on Sabre ICs. DSD output its not native. So in that case SW volume control its the same as PCM.
I am checking with the team for impedance.
1. I think the 384K is the max pll freq, if you feed in another clock that's <Fmax the PLL will lock to it. I'm assuming you have configure for 1x bypass mode pg22. TI often refers the the tables as "examples" so other solutions are possible.
2. The analog boost, I think, is because of the different definitions of "line out" levels. Is it consumer line out or pro line out? There is a 2x difference.
2. The analog boost, I think, is because of the different definitions of "line out" levels. Is it consumer line out or pro line out? There is a 2x difference.
1.
Even if the PLL locks at 352k8, which would be nice, I'm wondering if the filters get bypassed - as with 384k.
2.
+0.8db voltage = +10%
Is this it? >> analog1=1V (-6db) & analog2=2V (0db) & analog3=analog2+0.8db = 2.2V?
I'm wondering if it is really recommended to use analog3 at 2.2V??
I always have the impression things get a little "bloated".
I have seen some sites though recommending to go for analog3 !?!?
Even if the PLL locks at 352k8, which would be nice, I'm wondering if the filters get bypassed - as with 384k.
2.
+0.8db voltage = +10%
Is this it? >> analog1=1V (-6db) & analog2=2V (0db) & analog3=analog2+0.8db = 2.2V?
I'm wondering if it is really recommended to use analog3 at 2.2V??
I always have the impression things get a little "bloated".
I have seen some sites though recommending to go for analog3 !?!?
I took a look at the Z of ess 9028/38. On the 38 the datashet says 774Ohm
In our design the impedance is in the range of few Mohms . Ideally it has to be under 5ohm and we are way below that.
At last note this...both thd+n and SNR on ess sabre are given ONLY with a IV stage , in voltage mode , both numbers drop by at least 8db (according to a question answered by tech support a few months back)
In our design the impedance is in the range of few Mohms . Ideally it has to be under 5ohm and we are way below that.
At last note this...both thd+n and SNR on ess sabre are given ONLY with a IV stage , in voltage mode , both numbers drop by at least 8db (according to a question answered by tech support a few months back)
Last night I finally had a real chance to listen to my piano + kali. I run the piano in 2.2 mode. Everything was going well and sounding good. I listened to an album all the way through, played single track from a second album then decided to listen to a third album, all of a sudden my system belched out really loud noise. I immediately pulled the plug on my amp but the damage was done, two fried tweeters. I'm using the unmodified Allo firmware on GitHub linked to from the Allo website (created 03/26/2017). My pi is a 3, I'm running Raspbian 4.9.59-v7+, my playback software is JRiver (which I run on two other pis with no troubles). I recall some early posts which indicated there were issues with clicks when changing tracks but nothing like what I experienced. Clearly, I'm not too keen on the piano right now but I wonder if based on my setup there is something I've done wrong. Any help or suggestions is appreciated. Matt
@designmule
I've used Piano 2.1/Kali with Moode and Volumio, mostly in dual mono and only briefly in 2.2 mode. Never had issues with noise, apart from some occasional quiet crackles with Volumio. Was there anything different about the third album, apart from the obvious? Did the noise happen immediately at the start, or some way in?
I've used Piano 2.1/Kali with Moode and Volumio, mostly in dual mono and only briefly in 2.2 mode. Never had issues with noise, apart from some occasional quiet crackles with Volumio. Was there anything different about the third album, apart from the obvious? Did the noise happen immediately at the start, or some way in?