Looking forward to your comparision. What are your first impressions?
Thinking of getting a couple of these for a dual mono application. Has the latest pcb incorporated any of your suggestions? What pcb design features have been added to the latest version available?
Wow! seperate the ES9018 AVCC_R/L power supply provide a much more 3D sound stage. The location of every instrumentals is much much more clear
It's like listening the different of a vertical piano and a grand piano,
This is really a breakthrough even after many layout improvement since the first version
I finally know why many diyer want dual mono config or at less seperate AVCC_R/L power supply.
An externally hosted image should be here but it was not working when we last tested it.
this version will be available after the last few pieces ex-version is out
Last edited:
The location of every instrumentals is much much more clear
It's like listening the different of a vertical piano and a grand piano,
This is really a breakthrough even after many layout improvement since the first version
I finally know why many diyer want dual mono config or at less seperate AVCC_R/L power supply.
this version will be available after the last few pieces ex-version is out
How much is the cost of the new version?
diyinhk I just ordered a board and you sent me the old version. Why? I am quite annoyed at this point.
The location of every instrumentals is much much more clear
It's like listening the different of a vertical piano and a grand piano,
This is really a breakthrough even after many layout improvement since the first version
I finally know why many diyer want dual mono config or at less seperate AVCC_R/L power supply.
An externally hosted image should be here but it was not working when we last tested it.
this version will be available after the last few pieces ex-version is out
this version has some component value missed print, the whole batch is went to the rubbish bin
It should available soon
it shouldnt be, that would be a bad idea to separate the grounds like that, you would just be asking for ground loop currents inside the chip.... in general its a bad idea, if you want to shield delicate grounds from strong circulating currents, put a slot in the ground, but maintain a single ground.
funny that slowly some of the failings I suggested before the production boards were even made are being 'discovered' one by one as time goes on. I didnt discover them either, but I was criticized for being picky for suggesting them and fought every step of the way.
the dac chip has separate supplies, references, everything for the L/R channels; use them.
funny that slowly some of the failings I suggested before the production boards were even made are being 'discovered' one by one as time goes on. I didnt discover them either, but I was criticized for being picky for suggesting them and fought every step of the way.
the dac chip has separate supplies, references, everything for the L/R channels; use them.
IMHO I would believe that a complete separation of analog/digital ground with be a very positive mod. Due to the high frequencies working on the digital busses. And a separation from USB ground from computer. Which in total should reduce noise ??
Or is there better ways around this ?
Or isnt it an issue ? Then I would like a short explanation if anyone got it.
Thanks for reply
I believe your question has already been answered earlier in this thread, read here for more background - Grounding of Mixes Signal Systems
I gave you a reply, short explanation and a brief pointer as to a way around it, but it seems you didnt understand it.
both are still connected to and in the dac chip and if the 2 are separate and at slightly different potentials, it will create circulating loop currents/noise inside the chip, both directly and parasitically.
its a bandaid solution that doesnt even work, use proper grounding and decoupling, use slots in the ground plane to shield delicate parts if you like, but it should be one plane. section it if you like but all should be connected, otherwise it will seek out its own connections and the loops may not be where you think/hope
how exactly are you going to define analogue and digital parts anyway?
aside from it also being the suggested modern practices with much more delicate devices than this to avoid it, the ESS app notes make a specific point of recommending to NOT do it.
and yes Chris, it is a bit De Ja Vu all over again
both are still connected to and in the dac chip and if the 2 are separate and at slightly different potentials, it will create circulating loop currents/noise inside the chip, both directly and parasitically.
its a bandaid solution that doesnt even work, use proper grounding and decoupling, use slots in the ground plane to shield delicate parts if you like, but it should be one plane. section it if you like but all should be connected, otherwise it will seek out its own connections and the loops may not be where you think/hope
how exactly are you going to define analogue and digital parts anyway?
aside from it also being the suggested modern practices with much more delicate devices than this to avoid it, the ESS app notes make a specific point of recommending to NOT do it.
and yes Chris, it is a bit De Ja Vu all over again
IMHO I would believe that a complete separation of analog/digital ground with be a very positive mod. Due to the high frequencies working on the digital busses. And a separation from USB ground from computer. Which in total should reduce noise ??
Or is there better ways around this ?
Or isnt it an issue ? Then I would like a short explanation if anyone got it.
Thanks for reply
Last edited:
OOps ;-)
Thanks for your patience anyway. I certainly do understand Im walking far away on thin ice here.
My reason for thinking this, was that in ESS´s papers everything is noted as AGND and GND. I didnt know they both were connected in the chip. which I do understand that makes a serious issue. We dont want the main ground connection in the middle of the chip.....
Thanks again
Thanks for your patience anyway. I certainly do understand Im walking far away on thin ice here.
My reason for thinking this, was that in ESS´s papers everything is noted as AGND and GND. I didnt know they both were connected in the chip. which I do understand that makes a serious issue. We dont want the main ground connection in the middle of the chip.....
Thanks again
No problem, yeah well it depends exactly how you define analogue and digital supplies whether they are joined in the chip. having each supply separate but with all joined under the chip is also fine. The idea is to provide each signal with a low impedance return and shortest loop possible, so that fast slewing digital signals dont have anywhere they would rather be. with bad layout you could still have benefit I guess.
check the ES9008 app note, actually I remembered incorrectly, they dont specifically say its bad, but they mention doing extensive testing and finding no benefit.
check the ES9008 app note, actually I remembered incorrectly, they dont specifically say its bad, but they mention doing extensive testing and finding no benefit.
I am still puzzled and confused with some of the postings: glt you are talking about a need for 1.2V supply, in addition to 3.3V AVCC and 3.3V DVCC. However I don't see a need for a 1.2V regulator for a v2 board. I just got my v2 board and it has AVCC, DVCC and additional 3 pins marked +12, GND and -12V. I don't get it why does this board require +-12V PS? I am yet to figure out where to source AVCC power supply. For +-12V (if really needed) I will use a TP Placid bipolar PS and for DVCC 3.3V I will use ADP7104 eval board. Any recommendations for AVCC 3.3V regulator?
Thanks KlipschKid, that explains it. Since I will try direct output, I will skip using +-12V supply, unless I decide to install the OPAMPs.
Thanks guys for an excellent thread so far.
You have convinced me to purchase a board, and give myself the goal of my first DIY digital project. (see below for my parts so far.....)
I am currently plowing through the 300+ posts in this thread.
Before I ask something already mentioned/answered....I'm gonna ask something probably already mentioned/answered.
As a digital newbie, I wonder about some digital interface basics. So bear with me for a moment whilst I reveal my inner binary child.....
.......
1) I'm not sure I need USB input, but surely I want toslink and coaxial.
If I put USB in...(i2s board)....are there rules for it's clock, and the DAC board's clock ?....as in proper sync ?
(I plan on using a good on-board clock, so using the i2s clock may be a step down...if I understand correctly)
Links for a good i2s ?
2) I see that balanced analog output is available on this board, however I want Balanced and unBalanced outs. Is this an "off the board" solution?....or can both be pulled from the board ? (loading one or the other only) I suppose a front or back panel switch could accomplish this too.
3) can the 680 ohm's become 1k ?...ultra precision resistors are easier to source at this value. I already have some 100 ohm Vishay 0.2ppm....It would be nice if these paired resistors had the same tempco.
Thanks...I'll keep reading/researching to answer my other questions !!
=Randy=
-ebay DIYINHK’s ES9018 board
-one "Kubota low noise regulator full blown version kit !! "
-two TPS7A47 Eval Boards
possible clocks:
-ASV-50.000MHZ-E-J-T
-CCHD-950-25-100 mHz
-FXO-HC736R-50 mHz
p.s.....just thinking here....doesn't the quality in the audio chain also depend on the digital source before the DAC ?...or are 1's and zero's >>> just 1's and zero's ?
You have convinced me to purchase a board, and give myself the goal of my first DIY digital project. (see below for my parts so far.....)
I am currently plowing through the 300+ posts in this thread.
Before I ask something already mentioned/answered....I'm gonna ask something probably already mentioned/answered.
As a digital newbie, I wonder about some digital interface basics. So bear with me for a moment whilst I reveal my inner binary child.....
.......
1) I'm not sure I need USB input, but surely I want toslink and coaxial.
If I put USB in...(i2s board)....are there rules for it's clock, and the DAC board's clock ?....as in proper sync ?
(I plan on using a good on-board clock, so using the i2s clock may be a step down...if I understand correctly)
Links for a good i2s ?
2) I see that balanced analog output is available on this board, however I want Balanced and unBalanced outs. Is this an "off the board" solution?....or can both be pulled from the board ? (loading one or the other only) I suppose a front or back panel switch could accomplish this too.
3) can the 680 ohm's become 1k ?...ultra precision resistors are easier to source at this value. I already have some 100 ohm Vishay 0.2ppm....It would be nice if these paired resistors had the same tempco.
Thanks...I'll keep reading/researching to answer my other questions !!
=Randy=
-ebay DIYINHK’s ES9018 board
-one "Kubota low noise regulator full blown version kit !! "
-two TPS7A47 Eval Boards
possible clocks:
-ASV-50.000MHZ-E-J-T
-CCHD-950-25-100 mHz
-FXO-HC736R-50 mHz
p.s.....just thinking here....doesn't the quality in the audio chain also depend on the digital source before the DAC ?...or are 1's and zero's >>> just 1's and zero's ?
Last edited:
what you guys think of this oscillator ? less than 2 euros in mouser, very good jitter values.
http://www.vectron.com/products/appnotes/Typical_Phase_Noise_Performance_for_the_VCC1_Series.pdf
http://www.vectron.com/products/xo/VCC1.pdf
http://www.vectron.com/products/appnotes/Typical_Phase_Noise_Performance_for_the_VCC1_Series.pdf
http://www.vectron.com/products/xo/VCC1.pdf
- Status
- Not open for further replies.