CDP / DAC output stage

Dedicated to the silver disk spinner
karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Mon Oct 22, 2018 2:40 pm

Thanks Phil,

Oooo, reading that I am not so sure about adding more of them in the chain. I don't know. Suck it and see I suppose. Not cheap to try though. I do have a pair of the 3.3uf spare that I could rob for a bit just to try.

Stu
:bigsmurf:

karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Mon Oct 22, 2018 2:49 pm

Hi Wolfgang,

I am pretty sure it has 2.5V DC on the dac output but then I might have measured wrong with my DMM. I agree that no cap is best if possible to do so. I will measure again.

NVA amps are like no other solid state amps I have come across. Richard Dunn the designer states that he intentionally designed them for use with his passive preamps. There is no input coupling cap, input impedance may be at a level to give a 10K pot an easy load, sensitivity is increased to make up for lost gain with passive pre. Plus on the output side there is no Zobel network, output inductor or resistor.

Sound mighty fine to me.

Stu
:bigsmurf:

User avatar
pre65
Amstrad Tower of Power
Posts: 17319
Joined: Wed Aug 22, 2007 11:13 pm
Location: North Essex

Re: CDP / DAC output stage

Post by pre65 » Mon Oct 22, 2018 2:51 pm

Why not get a SPDIF output on the Naim and try that cheapo DAC you have ?

Richard Dunn seems to like them. :wink:
Life is what happens to you while you're busy making other plans. John Lennon

G-Popz THE easy listening connoisseur. (Philip)

karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Mon Oct 22, 2018 3:02 pm

Well I could do that I suppose. I have found the i2s lines (Bitclock, wordclock and data) between SAA7376 decoder chip and TDA1305 dac chip.

Its a lot of buggering about though. :roll: What do i need ? Some sort of SPDIF chip ?

Stu
:bigsmurf:

User avatar
pre65
Amstrad Tower of Power
Posts: 17319
Joined: Wed Aug 22, 2007 11:13 pm
Location: North Essex

Re: CDP / DAC output stage

Post by pre65 » Mon Oct 22, 2018 3:07 pm

karatestu wrote:
Mon Oct 22, 2018 3:02 pm
What do i need ? Some sort of SPDIF chip ?

Stu
Not that I know of.
Life is what happens to you while you're busy making other plans. John Lennon

G-Popz THE easy listening connoisseur. (Philip)

User avatar
Nick
Site Admin
Posts: 11364
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

Re: CDP / DAC output stage

Post by Nick » Mon Oct 22, 2018 5:21 pm

karatestu wrote:
Mon Oct 22, 2018 3:02 pm
Well I could do that I suppose. I have found the i2s lines (Bitclock, wordclock and data) between SAA7376 decoder chip and TDA1305 dac chip.

Its a lot of buggering about though. :roll: What do i need ? Some sort of SPDIF chip ?

Stu
Yes, you would,

Phil:
the external DAC needs SPDIF, but that not the internal data transport in most dacs, internally they often use i2s which is 3 or 4 bits of data on separate lines. To convert that to SPDIF as said, it would need a converter of some sort. Can be done, but unlikely to be worth the effort (IMHO).

Something like this

https://www.cirrus.com/products/wm8804/
Resistance isn't futile it's V / I.

karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Mon Oct 22, 2018 7:19 pm

I thought so. Thanks for clearing that up Nick.

I will keep the dac as is. I already pimped up the decoupling and vref cap with oscons years ago. Gave each power supply pin (3 of) its own psu and pfm flea reg on the most important one and a sprinkling of capacitance multipliers.

Next thing i will do is fart about with the coupling. Then see what i think.

Stu
:bigsmurf:

karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Tue Oct 23, 2018 2:10 pm

TDA1305 Dac chip definitely has 2.5V DC on its output. I just checked again.

So I cant get rid of the coupling cap :( Can still lower it to 1uf and not have audible bass roll off ?
:bigsmurf:

karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Tue Oct 23, 2018 2:51 pm

Just read https://www.soundstageultra.com/index.p ... e-controls

Puts me off digital volume control but then I know naff all about it. Preamps are to be avoided if possible - when I removed my Avondale preamp board it was better with just a pot but that was when I still had the cdp's output stage in circuit.

The last thing I want to do is introduce another active preamp. I kind of like getting rid of bits I don't need :D

Currently my potential divider (10K in series & 768R to 0V) is about 15cm of solid silver wire away from the dac and the wires are not twisted or laid next to each other so capacitance of the wiring should be low (but ruined by the 10uf coupling cap).Effectively it is just some resistors and rca sockets tagged on the end of my cdp. I have low capacitance cable between potential divider and 4 monobloc power amps but they are 75cm and 100cm in length. I thought of building my system so the volume control is as close to the dac as possible and power amp pcb's as close after that as possible.

Reducing the capacitance by substituting 1uf for the 10uf coupling cap and keeping all cable runs really small should be beneficial as far as the impedance and capacitance forming a low pass filter.

Here is how it looks at the moment. 10uf leg lifted, 100k to signal 0V added, wire taken off to potential divider and phono sockets

Image

Pic of my preamp inside a cdp (kind of) :D


Image
:bigsmurf:

karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Tue Oct 23, 2018 3:45 pm

Anyone fancy casting an eye over this. Shows one channel of dac, dc blocking, volume control and input stage of power amp. I have not included the capacitance, resistance and inductance of cables etc cos I don't know what they are. Anything I can improve on (except reducing coupling cap value) ?????

Image
:bigsmurf:

Wolfgang
User
Posts: 37
Joined: Mon May 07, 2018 3:08 pm

Re: CDP / DAC output stage

Post by Wolfgang » Tue Oct 23, 2018 8:15 pm

You don't need the 100k. If you vary the 768R to ground you dump a lot of the signal coming from the DAC and lose transients and fine details depending on the volume settings (less than 768R at lower volumes).
A buffer which would present a stable load/impedance for the amp independent of volume settings would probably sound better. A 100k ladder type attenuator at the grid of a single tube buffer would be much easier on the DAC with tons of details and transients at the output of the buffer. Just saying.

On the article:

I am struggling with the information/opinion expressed int this article. Maybe somebody can add more clarity? Here is my take on it.
If we don’t talk about numbers but about what’s audible in combination with amps/speakers/listening room then the discussion about lowest noise floor below -80dB is kind of academic even with sensitive speakers. So if we cannot really use this information for something practical or audible we shouldn’t mention it just in order to say later that it probably doesn’t affect the whole thing practically very much as dynamic ranges and amps/speakers cannot reproduce anyway what a perfect DAC could deliver. What’s the point of all this?
But then again “noise levels” are the only handle for a comparison between analog and digital volume control. But if we move in this direction I would also like to see a chart that shows added noise levels from a pot/attenuator and different resistor qualities and would compare this with the digital noise. And a chart that would show how cable qualities in general in combination with changing impedances of an analog volume control (exception: ladder type) would impact the original signal and which effect is worse or more audible.
Of course 24 or higher bit resolution is a must in the case of digital VC. Also upsampling of the signal with a good algorithm comes to mind in order to get less noise from the digital VC. And if we look at the chart then we see that only below -25dB we would get into trouble regarding noise. If we simply set the gain of the buffer low enough so that the digital volume control doesn’t have to operate most of the time lower than -25dB one would still get the benefit of an analog signal without the losses coming from a passive resistor network. There shouldn’t be a problem from increasing the noise floor of the signal with the digital volume control.
The other argument used in the article which makes no sense in this discussion is the dynamic range/SNR: “The signal/noise ratio is not the same as dynamic range, but it gives some indication of the amplifier’s noise floor.” So what to do with this? Most good recordings (digital & analog) have an average dynamic range of 10-15dB. The biggest dynamic jumps I could measure were about 25-30dB. Even vinyl is fully sufficient and the “dynamic range” which is sometimes used to prove superiority of the digital signal over the analog has absolutely no practical meaning. To consider signal to noise ratio in this discussion without looking at the same time at the dynamic range of the actual recording makes absolutely no sense to me. We only experience “dynamic range” which is always worse than SNR and a lot “worse” than what any serious DAC will be capable of.

Wolfgang
User
Posts: 37
Joined: Mon May 07, 2018 3:08 pm

Re: CDP / DAC output stage

Post by Wolfgang » Tue Oct 23, 2018 8:21 pm

Sorry: you don't need the 100K if you give up the cap and rather correct the DC offset at the output if there is one. It simply shouldn't be there IMO. My mind played a trick on me ..because that's what I was thinking all along.

karatestu
User
Posts: 76
Joined: Fri Mar 03, 2017 1:34 pm
Location: Leeds/York

Re: CDP / DAC output stage

Post by karatestu » Tue Oct 23, 2018 8:27 pm

Thanks for your input Wolfgang. I am not adverse to putting in a buffer and will eventually try one i think.

As far as the article goes, i dont know enough to say if its right or wrong. Just thought it may prompt some discussion from you more learned members.

Cheers, Stu.
:bigsmurf:

Post Reply