That, and why would you want to butcher the TDA1541A’s linearity by loading the output into a >0R resistor and then use a transformer. There’s perfectly fine active I/V conversion schemes to be found that are far superior in every way.
There is no resistor across the dac outputs and I prefer a minimalist approach provided I like the sound. Don't want to add more active devices and no preamp is required, just a pot between the dac and my tube amp. (when I've finished it, and you cannot get more simple than 3 tubes and a opt for 25w pure class A p-p )
As regards the inductance of the dem cap traces, I'm using low frequency dem clocking so the tiny inductance is not an issue.
What I/V conversion did you use?
The benefits of differential seems to me to be best suited to bifilar wound transformers as these will both cancel out the 2ma output offset and greatly reduce quantisation noise so no further filtering or amplification is required.
Precisely. I have listened to the same cds on a Teac slot in player, Arcam CD73, and a Technics high end Mash Dac player but still prefer my dac, despite whatever shortcomings it may have.
I have a pair of custom Sowter trafo which was made for this purpose but have never got around to understand & use it. Lol
To be honest it was my early days into diy dac, told sowter that I wanted to build a parallel dac with trafo out but somehow they misunderstood & that's how I ended up with these trafo. lol. I've now matured from endless mucking around & have taken a slightly different route but still using trafo output but in 1 to 1 line opt instead.
Cheers & Merry Xmas
Cheers & Merry Xmas
Well bateryman is using 100 uF lytics as DEM caps...I believe he is needing to try the best practice around this chip instead. Sometimes ones need an extra 10 years of experiments before to try something elqe. As the Ldovr regs, I tried it...Sterile sound...imho because the layout and output caps of the tiny pcb...not good for these tda1541 powerline.
Do you a favor, use only one chip, S1, S2, or late SouthAsia fabricated, active IV, diamond stage style with zero V trimming between the two channels to avoid dc coupling caps, no feedback, no crappy lytics for the 14 DEM decoupling but if you 14 BG N caps perfectly matched with less than 2.5 mm leads pitch. Choose sota IV resistor and select them with ears, do spend hundred hours on working of the ground decoupling and choices of regs typologys with good caps where it is needef.
...then only use outputt traffo if your pre needs symetric inputs
But who am I, it is not if we never talked of that before on numerous thread...again best documented dac chip ever.
At the end we will all listen to Maria Carey Christhmas song...so why bother, hey.
Do you a favor, use only one chip, S1, S2, or late SouthAsia fabricated, active IV, diamond stage style with zero V trimming between the two channels to avoid dc coupling caps, no feedback, no crappy lytics for the 14 DEM decoupling but if you 14 BG N caps perfectly matched with less than 2.5 mm leads pitch. Choose sota IV resistor and select them with ears, do spend hundred hours on working of the ground decoupling and choices of regs typologys with good caps where it is needef.
...then only use outputt traffo if your pre needs symetric inputs
But who am I, it is not if we never talked of that before on numerous thread...again best documented dac chip ever.
At the end we will all listen to Maria Carey Christhmas song...so why bother, hey.
Right now 3x stacked AD844/OPA627, I'll maybe try OPA681 one day.
I'm flirting with a passive I/V followed by Aikido tube preamp, definitely not going the discrete JFET/BJT route.
Headphone amp is a INA1620 w/ differential input.
diyiggy,
you have stated some important principles. Sterile sound is not caused by layout, grounding scheme or output caps, but rather by circuit topology.
you have stated some important principles. Sterile sound is not caused by layout, grounding scheme or output caps, but rather by circuit topology.
So are you saying that when circuit topology has been optimized the rest won't matter N101N ?
MERRY XMAS & HAPPY NEW YEAR TO ALL
Cheers
MERRY XMAS & HAPPY NEW YEAR TO ALL
Cheers
That's a generalisation which anyone can make. What balanced topologies have you listened to? Which I/V conversion strategy did you use?
Was it a NOS or OS design?
Was it a NOS or OS design?
I make generalisations where appropriate. Balanced topologies do not produce even order harmonics. There is some talk about single-ended sound...
Thus spoke Zarathustra, some toothless people in their ignorance think that they stand above physical laws.