Experience with this DIY DAC ?

[GerardA]

Because the topend sounded not good I did some experimenting with listening and measuring with audiotester.
It looks like there is a 6 dB peak at about 33 kHz after which it goes down rapidly. With the RC the peak gets lower but a small peak at 20 kHz.
With the primary C there is a small influence too.
Then measuring the phase it looks horrible with the C's in!
Out the C's and the phase is flat between 200 and 20KHz.
Listening I did with Hallelujah by Bon Jovi. With the C's the voice gets really cracky, raspy, without the C's it sounds natural.
Next step try the listen to the difference of the shunt-R on the primary.
Maybe with it it sounds a little bit sat on.
So far..

[Bill Fuss]

If the resonance peak of the trafos is truly at 33khz I would deem them unusable. That is much lower than any I have tested. Please check your testing setup, does your gen have enough bandwidth for testing.

[brianco]

Having lived with this DAC for some months set up with serial resistors and a small teflon across the inputs along with a small teflon/resistor, in series, across the secondaries to my A-20 transformers, I found that the treble was becoming very 'metallic' and shrill. It also seemed that there were some phase anomalies. I have also taken the whole dac off breadboard and fitted it in a 12"x6" 1/4"thick 'U' channel of alloy. This emphasized the above problems!

I have now put a 1k resistor across the input and lo behold it has not only solved the above problems but has given an enormous up in the overall sound quality. The most noticeable to me are that the sense of recording location is now very much improved and that solo piano has taken such a vast improvement in terms of control of attack and delay that the sound is now far far more credible than it was before this change.

Thank you!

[Ryssen]

This is the setup with my DAC:



There is a pair of LL1517(thanks Havoc08)with 2x560ohm in series with the primary.A RC at the out put with 1k and 2,2n.
The first thought was that it sounds more relaxed than the onboard opamps,
will listen more...
Is there any dissadvantage to leave the opams on the board,while taking the signal trough the trafos?

[McGyver]

Get rid with that four green Nichicon Muse caps. And You don't need 15-0-15V supply.

[piero7]

Originally Posted by Ryssen

This is the setup with my DAC:



There is a pair of LL1517(thanks Havoc08)with 2x560ohm in series with the primary.A RC at the out put with 1k and 2,2n.
The first thought was that it sounds more relaxed than the onboard opamps,
will listen more...
Is there any dissadvantage to leave the opams on the board,while taking the signal trough the trafos?

just disconnect the 12+12 supply

[Havoc08]

Originally Posted by Ryssen

This is the setup with my DAC:



There is a pair of LL1517(thanks Havoc08)with 2x560ohm in series with the primary.A RC at the out put with 1k and 2,2n.
The first thought was that it sounds more relaxed than the onboard opamps,
will listen more...
Is there any dissadvantage to leave the opams on the board,while taking the signal trough the trafos?

Np ryssen. I will follow you findings closely as I have not had time to experiment myself yet.
Have you tried different resistors?
What type of caps are you using?
What does your setup look like at home and which music type do you listen to?

[GerardA]

just disconnect the 12+12V supply

How about using a 12V battery for the 9V supply?

@brianco
For me the sound gets much better with a low sourceimpedance to the trafo.
I lowered the seriesresistance from 2 x 220 to 2X 47 ohm.
Everything improved, more natural and strong bass, cleaner and brilliant highs and more depth. A parallel resistor at the primary lowers the sourceimpedance too but for me it gave a loss of dynamics and sounded a bit strange.
Maybe I'll lower the seriesresistance to zero. Hope the dac doesn't mind.
It looks like the impedance is more important for the trafo to be right then for the DAC-chip.
The datasheet and measurements confirm a lower phasedifferense in the bass with lower sourceresistance. Distortionmeasurement gave at least 6 dB less distortion and 20 degree less phasechange.

@ Ryssen
Try the above and I'm sure you will hear a clear difference!

[jkeny]

ryssen, I will too - I haven't had a chance to experiment with my LL1517. I tried them with 1K across primary, I think & i didn't like the sound - it was lack top & bottom end but this was just an quickie lash-up to get a flavour of their sound - I knew it was far from ideal.

I got my Sescoms MI-97 & connected them again quickly to get a flavour of the sound - much more open top & bottom but I need to tweak a bit before I comment further. I'm not using the DAC or board in this thread

[Ryssen]

Have you tried different resistors?

No not yet

What type of caps are you using?
What does your setup look like at home and which music type do you listen to?

The caps where some orange ones I had in the
junkbox.
The setup is Squeeze box Duett,Roland M-1000 digitalmixer-DAC-F3-and homemade speakers with 12" sub.
I listen mostly on hard rock ex.Joe satriani and similar.and some jazz to.

Maybe I'll lower the seriesresistance to zero. Hope the dac doesn't mind

No dont...I tried that by misstake,at first start I had no resistors at all,there where sound about 10 seconds then nothing until I pulled the mains plugg and inserted it again,then 10 seconds again....so the DAC doesn´t like that..

Yes,i will try some 47 ohms if I have some home.