Thanks for the information and making a place for making the case for the single board approach.
I doubt I would be alone in hoping you can justify making some of the CCS boards. I know I would like to have a pair.
Would using surface mount resistors save enough space to make it worth the soldering difficulty?
I am confused - we can correct for DC offset with Pin 20 CCS?
I would have thought this could only be done with a pot on whichever side was "higher".
I must be missing something ... Though this is really neat if that is true.
Another reason to install the CCS!
this is the Basic function from pin 20. It sets the Output bias current. Now on the I/V resistor (Ra//Rb value) this results in a DC value. My measurements Show, that 2,7 Volt is a good desired value for Maximum Output head room at Minimum d2 distortion...
dddac
I get dc offsets of 1.1mv and 3.2mv when I adjusted the CCS to get
exactly 2.7V. This isnt a problem for my transformers but it would be nice to have
a way to get to zero dc offset. Are you suggesting that zero offset can be achieved
by adjusting the CCS? If so how ? Or is there another way?
I get dc offsets of 1.1mv and 3.2mv when I adjusted the CCS to get
exactly 2.7V. This isnt a problem for my transformers but it would be nice to have
a way to get to zero dc offset. Are you suggesting that zero offset can be achieved
by adjusting the CCS? If so how ? Or is there another way?
Zeta4,
You may find this is the result of iv resistor tolerances etc, probably won't be able to change much else to get to zero unless you get a Z foil trim pot and parallel the IV resistor to get them exactly 0v. This will add some of the z foil flavor which you may not like.
Cheers,
Drew.
You may find this is the result of iv resistor tolerances etc, probably won't be able to change much else to get to zero unless you get a Z foil trim pot and parallel the IV resistor to get them exactly 0v. This will add some of the z foil flavor which you may not like.
Cheers,
Drew.
Differential voltage offset
The Rload resistors should be matched within 0.1%. Note that a 1% mismatch in Rload resistors calculates to a 20mv offset for 2.7 volts into 133 ohms. A 0.1% mismatch results in a 2mv which is acceptable.
My DAC output buffer provides an adjustment which can further lower the offset down near 0.0mv. I will explain the buffer in my new single board upgrade Thread.
The Rload resistors should be matched within 0.1%. Note that a 1% mismatch in Rload resistors calculates to a 20mv offset for 2.7 volts into 133 ohms. A 0.1% mismatch results in a 2mv which is acceptable.
My DAC output buffer provides an adjustment which can further lower the offset down near 0.0mv. I will explain the buffer in my new single board upgrade Thread.
As promised, below is my implementation of DDDAC.
Some description on the implementation :
1) Rubycons ZLGs are used throughout the DAC and main board (I personally prefer Rubycons over Silmics and Muse)
2) I perfer using tantalums over the MKP films.
3) Belleson SPJ12V individually powering each of the channel while another external 5V psu (not shown in picture) is powering the main board.
4) The cable in blue is a I2S cable going into my lightly modified streamer (not shown in picture) which is this one below which has I2S out capability.
D100 PRO?Ultimate 32bit,192KHz,HDD,Network Transport System - Products Show - ??????
5) The 2 boards with the big caps (Mundorf Mlytic caps) are the left and right channel psu boards with ultra fast recovery diodes as rectifiers and they are powered by an external 160va transformer.
6) The output caps are Mundorf Supreme which I much prefer over Mundorf Silver Gold Oil (Silver Gold Oil a little slow to my taste).
7) The resistors are still the Vishay Dale but I am looking at changing to Caddocks.
Next steps :
1) Build a acrylic top cover
2) Belleson SPM8v and SPM3.3v for the individual dac regulators (a lot of money here so still thinking hard)
3) Add a raspi or BBB so that it can function stand alone.
Comments and questions welcomed.
An externally hosted image should be here but it was not working when we last tested it.
Some description on the implementation :
1) Rubycons ZLGs are used throughout the DAC and main board (I personally prefer Rubycons over Silmics and Muse)
2) I perfer using tantalums over the MKP films.
3) Belleson SPJ12V individually powering each of the channel while another external 5V psu (not shown in picture) is powering the main board.
4) The cable in blue is a I2S cable going into my lightly modified streamer (not shown in picture) which is this one below which has I2S out capability.
D100 PRO?Ultimate 32bit,192KHz,HDD,Network Transport System - Products Show - ??????
5) The 2 boards with the big caps (Mundorf Mlytic caps) are the left and right channel psu boards with ultra fast recovery diodes as rectifiers and they are powered by an external 160va transformer.
6) The output caps are Mundorf Supreme which I much prefer over Mundorf Silver Gold Oil (Silver Gold Oil a little slow to my taste).
7) The resistors are still the Vishay Dale but I am looking at changing to Caddocks.
Next steps :
1) Build a acrylic top cover
2) Belleson SPM8v and SPM3.3v for the individual dac regulators (a lot of money here so still thinking hard)
3) Add a raspi or BBB so that it can function stand alone.
Comments and questions welcomed.
Last edited:
Interestingly the Acko Labs site states this in relation to their ESS Dac:
-Simple setup. Can be powered by a single 5V master supply or better still separate supplies for Analog and Digital sections
David
Thanks for the tips Dave. I have discussed this with PET-240 about using seperate unregulated raw ps each for analog and digital sides of DAC. My thoughts would be CRCR for digital and LCLC for analogue. Appreciate any inputs here!
Thought I share with you guys my setup with the Pi i2s direct to DDDAC since my BBB went kaput few days ago. Note it wasn't notthing to do with i2s connection, it was just because I was experimenting the input jack DC with my linear powersupply. Anyway, a replacement is on its way.
Here are the video:
Raspberry Pi i2s direct to DDDAC - YouTube
Raspberry Pi i2s direct to DDDAC - YouTube
good stuff. If you're using i2s direct from the pi, you get a nice upgrade from adding a reclocker like the Acko S03 inbetween pi and dddacThanks for the tips Dave. I have discussed this with PET-240 about using seperate unregulated raw ps each for analog and digital sides of DAC. My thoughts would be CRCR for digital and LCLC for analogue. Appreciate any inputs here!
Thought I share with you guys my setup with the Pi i2s direct to DDDAC since my BBB went kaput few days ago. Note it wasn't notthing to do with i2s connection, it was just because I was experimenting the input jack DC with my linear powersupply. Anyway, a replacement is on its way.
Here are the video:
Raspberry Pi i2s direct to DDDAC - YouTube
Raspberry Pi i2s direct to DDDAC - YouTube
The latest ESS dac from Acko can be powered by seven independent analog/digital power supplies, eight if you put the clock on the board and power it indpendently. Sorry for the OT post Doede...
Cheers James!
I supposed to have my UFL adaptors here since last Thursday, somehow got lost somewhere with our Australia Post. Not sure when I will be receiving them, otherwise, I would have BBB i2s sync reclocking to SO3 and S03 sync to DDDAC at 45.xx/49.xx. Will post some photos when they become available if interest?
Interesting choice of caps.
Looks like you left L1/2 L3/4 out as well as R1/2 R3/4 while still running with the LF80 and LF33 for Vcc and Vdd. I think most people came to a different conclusion, to leave those filter elements in. Even after upgrading the regulators to shunts. I have anyway.
Hi Chanh,
I tried clc and crc on both analogue and digital so far, with marginal preference for clc. I will try clc on analogue and crc on digital following your note. What is the logic for the crc on digital?
David
I had great results using Rubycons in both analog and digital in my past mods, which is why I sticked to these caps. I like the liveliness and neutrality of the Rubycons for analog but minimally, I think they are great for digital given their low ESR and ripple numbers.
As for the L1-L4 and R1-R4 removal, I am kinda still experimenting. May revisit it again when I pursue the Belleson regulators. But having used Belleson for years, I always find the best sound is obtained when I have the regulator work harder than having additional components to aid in the filtering/regulation.
Hi David,
There is no logic in CRCR, it is just an affordable alternative. My experience so far, hope those in the "know" can correct me here, digital ps has the least beneficial from exotic caps and chokes, unlike the picky analog ps.
Here is what PET-240 has came up for me based on a single audio grade EI Tranny.
