Sure, Leo, you know my approach to PSI'd go for the XO area first JohnI compared things in stages with this one and found the XO reg and upgraded XO gave quite a noticeable boost in SQ, to my ears anyway.
Give the bat feeding the stock XO a try first, it should be a good starter. I used a Tent XO mainly because I had it sat spare but there may be cheaper or even better XO alternatives .
& yes the clock PS is always the first area I visit as it's usually neglected in most implementations. In fact, I believe that most of the clock upgrade package benefits are as a result of the better clock PS.Yes you will find it very worthwhile!
Last edited:
To make the most of the ESS jitter reduction, you need to use a high frequency clock.
The maximum the ES9022 will take is 50MHz, so you should use one wither 50 or 49.xxx, right next to DAC, and fed by its only regulator. I use LT1761-3.3.
Once you have done that, the actual 11.2896MHz clock for the QA becomes much less critical. Yes, you will still get some improvement with a good clock fed by separate batteries and low-noise regulators, but it will not match the improvement by the high frequency clock close to the ES.
If you want to do sychronous clocking, you need to :
a) find an XO with 45.1584 MHz and place it close to the ES, e.g.
OSC 45.1584 MHZ 3.3V HCMOS SMD - FXO-HC730R-45.1584
b) remove the TCXO of the QA,
c) use some logic to divide the 45MHz XO by 4 and fed that to the QA.
Me. I'll stick to asychronous.
Patrick
.
The maximum the ES9022 will take is 50MHz, so you should use one wither 50 or 49.xxx, right next to DAC, and fed by its only regulator. I use LT1761-3.3.
Once you have done that, the actual 11.2896MHz clock for the QA becomes much less critical. Yes, you will still get some improvement with a good clock fed by separate batteries and low-noise regulators, but it will not match the improvement by the high frequency clock close to the ES.
If you want to do sychronous clocking, you need to :
a) find an XO with 45.1584 MHz and place it close to the ES, e.g.
OSC 45.1584 MHZ 3.3V HCMOS SMD - FXO-HC730R-45.1584
b) remove the TCXO of the QA,
c) use some logic to divide the 45MHz XO by 4 and fed that to the QA.
Me. I'll stick to asychronous.
Patrick
.
Last edited:
Patrick,
I believe that others have said synchronous clocking sounds better than asynch - I still have to completely prove this to myself yet, I thought so before but I had some other issues that needed addressing.
Yes your synch scheme will work & I believe that addressing the QA clock like this is still going to be more beneficial than the Sabre asynch re-clocking.
I believe that others have said synchronous clocking sounds better than asynch - I still have to completely prove this to myself yet, I thought so before but I had some other issues that needed addressing.
Yes your synch scheme will work & I believe that addressing the QA clock like this is still going to be more beneficial than the Sabre asynch re-clocking.
Obviously, I don't know the full workings of it - only Dustin & development team know that. I have read about turning off asynch & my initial tests seemed to be show a benefit but as I said there were other factors that needed to be sorted out. Maybe now I can begin the tests again?
In a lot of instances, I feed the MCLK through from the transport & it doesn't suffer as a result, as long as the MCLK is low enough jitter. As I said, I believe jitter can be found on all I2S lines, not just MCLK.
Just compare this to the original PCB :
http://i273.photobucket.com/albums/jj220/kingbusoms/P1090354.jpg
Patrick
http://i273.photobucket.com/albums/jj220/kingbusoms/P1090354.jpg
Patrick
I used the PH shunts for the XO. You can use batteries or linear, your choice
I have some Lifepo4 batteries but they was way too big to fit in the supplied case, the micro shunts was a snug fit. Of course you can build a bigger enclosure etc and just optimize it for I2S so making some of circuitry redundant .
You can just use a better external main supply like low noise linear one or batteries. Many things to try with these units
I have some Lifepo4 batteries but they was way too big to fit in the supplied case, the micro shunts was a snug fit. Of course you can build a bigger enclosure etc and just optimize it for I2S so making some of circuitry redundant .
You can just use a better external main supply like low noise linear one or batteries. Many things to try with these units
An update!
I haven't been able to run an update on the card. I have the qls file on the card's root directory with no other files. It's called UpdataB8.qls. I insert the card, press the SW1 & hold while I turn on the device. All I get on the screen is "Start" & then "No music" message. Am I doing something wrong?
I connected a ES9022 DAC directly to the I2S out & it sounds great. The other news is that this can be a portable device, all run off one 3.3V battery (my favourite LiFePo4 types). I lifted the Vin leg of the AMS11127 3.3V reg & fed it from my battery. No need for the external PS. Powering this way doesn't have any screen display (will this be missed? ) & I doubt if it has any dig out - I haven't tested this.
The ES9022 DAC directly runs almost all headphones I have thrown at it - the difficult ones not to ear splitting levels of course. Using headphones would require a vol control.
Nice little portable unit - I might just look into doing an SD Card DAC based on this.
I haven't been able to run an update on the card. I have the qls file on the card's root directory with no other files. It's called UpdataB8.qls. I insert the card, press the SW1 & hold while I turn on the device. All I get on the screen is "Start" & then "No music" message. Am I doing something wrong?
I connected a ES9022 DAC directly to the I2S out & it sounds great. The other news is that this can be a portable device, all run off one 3.3V battery (my favourite LiFePo4 types). I lifted the Vin leg of the AMS11127 3.3V reg & fed it from my battery. No need for the external PS. Powering this way doesn't have any screen display (will this be missed? ) & I doubt if it has any dig out - I haven't tested this.
The ES9022 DAC directly runs almost all headphones I have thrown at it - the difficult ones not to ear splitting levels of course. Using headphones would require a vol control.
Nice little portable unit - I might just look into doing an SD Card DAC based on this.
You say you have no other files in the root directory. Do you have any music files on the card?
What you describe has worked OK for me. Try re-formatting the card maybe, or try a different make of card?
I for one would be interested in an ES DAC with I2S for the QLS.
John
I'm trying to update the firmware as per the QA550 instructions so I left any other files off the SD card only the update file. Ah, I just got it - the switch for updating was at the front - I misinterpreted the QA550 diagram.
I have some other ideas for this - a balanced analog out using two ES9022 DACs might be nice for balanced headphone users or straight to a balanced system?
I don't think this chip has the processing power to handle high-res files - pity!
An update posted here :
http://www.diyaudio.com/forums/digi...ng-new-ess-vout-dac-es9022-2.html#post2458187
Patrick
http://www.diyaudio.com/forums/digi...ng-new-ess-vout-dac-es9022-2.html#post2458187
Patrick
Referring to :
http://www.diyaudio.com/forums/digi...card-16-44-1-wav-transport-5.html#post2391648
I have to give John this one.
The synchronuous version is implemented and tested.
And I am totally converted now.
http://www.diyaudio.com/forums/digi...ng-new-ess-vout-dac-es9022-2.html#post2484647
Patrick
http://www.diyaudio.com/forums/digi...card-16-44-1-wav-transport-5.html#post2391648
I have to give John this one.
The synchronuous version is implemented and tested.
And I am totally converted now.
http://www.diyaudio.com/forums/digi...ng-new-ess-vout-dac-es9022-2.html#post2484647
Patrick
> What is the purpose of the diode at the input when it is feed with dc input? Can it be bypass when it is supply with battery power?
Prevent wrong PSU polarity; can be bypassed (at own risk).
I wire my supply directly at the cap after the power switch.
> Has anyone manage to bypass the output transformer?
Never used the SPDIF. Always been using I2S.
> Any improvement to the SQ when the above are bypass?
Suggest you try I2S first. Cannot be easier with a TDA1543.
Patrick
Prevent wrong PSU polarity; can be bypassed (at own risk).
I wire my supply directly at the cap after the power switch.
> Has anyone manage to bypass the output transformer?
Never used the SPDIF. Always been using I2S.
> Any improvement to the SQ when the above are bypass?
Suggest you try I2S first. Cannot be easier with a TDA1543.
Patrick