Experience with this DIY DAC ?

Hi Franz!

Is it possible to show a photo of the underneath of this extra circuit board containing the pair of Studer transformers please? I'd like to copy this wiring for when my pair arrive :)

Also whats the ohms on these transformers - I remember you saying it needs to be 1K or more for this DAC, and the only reference I can see to ohms on the datasheet is 146 and 173ohms :xeye:

Many thanks,

- John
 
For those using op-amps, I had a vast improvement in sound by using a pair of OPA627s on a Brown Dog converter board. Those who have been reading since the beginning will know I initially favoured the LM4562.

However since performing the 'official' Cirrus Logic mods to the output stage that Kevin kindly outlined a few pages back I felt the LM4562 sounded too laid back, with not enough passion in the performance.

The OPA627s really seem to fit in well now - I love the deep bass in particular, but never at the expense of the transparent midrange and effortlessly clear treble I'm also getting. Like Kevin I also added two 0.1uF ceramic caps to the +V / -V legs of the op-amps to avoid oscillation.

Looking forward to trying out the Studer transformers and seeing just what can still be squeezed out of this DACs performance.

- John

P.S. I'm not sure whether to implement the mods to the PLL section or not? There's one Cirrus Logic update which claims one should change the values to: 3K (Rft), 22nF (Cflt) and 1nF (Crip). However another datsheet seems to suggest that these values are only for optimising sample rates of 96-192... anyone care to comment? I have attached them both again for reference.
 

Attachments

  • tweaks.jpg
    tweaks.jpg
    50.2 KB · Views: 1,526
Some people asked me for measurements of the Studer trannies.

O.K.

I attached a Yellotec PUC professional USB sounddevice (I really dont like the sound of this device...) to my little Asus EEE.

Here the result:

An externally hosted image should be here but it was not working when we last tested it.


The red line is the characteristic of the sounddevice, output connected to input.

The blue line is the Studer 1:1 trannie.

You can recognize the soft (-3Db at 30Hz) highpass and exactly following the characteristic of the soundcard over 15kHz.

So, it will probably be linear up to higher frequencies.

BTW: people, buying one of this trannies from me:

Be aware, that it is nearby impossible, to unsolder them from a board.

The plastic will get weak and you will probably pull out one or several of the pins.

Within Studer service departement it is not allowed to use one of this unsoldered trannies any longer.

Franz

/Edit
Here a foto from my test setup on the kitchens desk:

An externally hosted image should be here but it was not working when we last tested it.
 
Hello

As I dont want to bust this thread: my intention to offer this Studer trannies was to help DIYselfer to get at a reasonable price trannies.

The price I ask is slightly below the official Studer pricelist and I also include the shipping costs for forum readers.

And it helps me, to finance my own two pairs of this trannies :cool:

Some new information for me, I learned today:

Neutrik is offering a possibly similar trannie:

Look here, the NTL 1:1:

http://www.neutrik.com/ch/de/audio/slch_948589320/Übertrager_productlist.aspx

Sorry, in German language. Maybe you find the adequate english Neutrik page.

And I dont know the price.

Kind regards
Franz
 
Re: Usable Transformers?

Mush said:
I am no expert in this realm. I found these from Edcor and am wondering if any of them look like good candidates. Your expert advice is welcome.
Ciao

http://www.edcorusa.com/Products/BrowseSeries.aspx?SeriesID=29

At these prices why not give them a try ?

I would be surprized that they are in the big leagues but sometimes it can be interesting.

I would try the 1:1
600 ohms to 600 ohms
 
In my experience, setting the CS8416 pin 20 high (PDUR=1) causes occasional loss of lock on 96K and 88.1

Yes, even though the data sheet says it's OK to 108K. It is not.

Anyone else see this?

I also use components for the CS8416 phase lock loop C-FILT, R-FILT and C-RIP that make it take longer to lock and have lower corner frequency for jitter.

I suppose you will be OK with (PDUR=1) if you only listen to 44.1 and 48K
 
I got questions, how to connect this trannies:

An externally hosted image should be here but it was not working when we last tested it.


Gnd can be connected to the output ground, but must not.

My implementation:

An externally hosted image should be here but it was not working when we last tested it.


Would like to see some original circuits, how they are used? Maybe we find some hints, to refine the usage for the DAC?

Just download this service manual:

ftp://ftp.studer.ch/Public/Products/Mixing_Analog/961-962/Manuals/961-962_Op_Serv.pdf (100MB)

Look for line input circuits in the schematics chapter.

Or the 980 mixer (I did not look into up to now). Some engineers told me, this was the peak of this technology: two input trannies per channel. So it must be the circuit from Paul Zwicky, I showed some postings earlier.

ftp://ftp.studer.ch/Public/Products/Mixing_Analog/980/Manuals/980_Op_Serv.pdf (147MB)

Kind regards
Franz
 
During coffee break today, I was talking with a Studer development engineer (I am not, I am working with IT, Radioautomation and VoIP for Radios).

We where discussing the 30Hz roll off from the small Studer trannies and I told him, that I ordered some Lundahls (I could not resist :D ).

He asked me, if they are bigger than the Studer trannies. This must be the reason, they are more linear down to lower frequencies, not just the amorphous material.

In this moment, I remembered that I saved one day a pcb out of the recycling box, with 16 big Digitec (a former partner from Studer) transformers. Probably also amorphous material.

Unknown spec.

An externally hosted image should be here but it was not working when we last tested it.


I just unsoldered one and measured the linearity: absolutely linear from 20Hz (cannot measure lower) up to 20kHz (or more).

And about the same output level as the input level. I guess, this are 1:1 transformers with very low loss, as I get the same level in both "directions".

Soon, I will try them. Just one quest to solve: wich side is the primary side?

Happy to have easter holiday starting now and no fix plans :cool:

Franz
 
I just attached two of the Digitec trannies to the chinese board:

An externally hosted image should be here but it was not working when we last tested it.


Absolutely great: nothing missing compared to the Studer's but more thight bass.

To not dissapoint people who already bought Studer's from me: I offer everyone who purchased up to now from me trannies, to get two of this Digitec's for only the shipping cost (15$).

Franz