DAC AD1862: Almost THT, I2S input, NOS, R-2R

The 3a5 is not for the faint hearted. It is a directly heated triode, and is microphonic. You will need discrete filament heater supplies (Rod Coleman), and something like Ale Moglia's high voltage gyrator supplies, with the various transformers to do all that. Its certainly not a cheap version, and of course it will measure poorly compared to any solid state option.

The advice to build and use op-amps first is a really good one. Socket the op-amps and then you can always tap from the output pins afterwards if you want to experiment later.
 
I would appreciate it if anyone could share their opinion on this stage + ad1862.

Be able to tell you soon, one arrived today.

IMG_0797.JPG

Also tried to use the Sjostrom Diamond buffers but there was too much DC, need to work it out.

IMG_0793.JPGIMG_0794.JPG
 
  • Like
Reactions: voxxonline
Imo that opa could rather makes a good buffer according the manufacter that produced it.... YMMV. That's cool someone posted a two opa pcb for such a good chips as buffer... AD8597 us very good dor the tasl among some other...lm6371, opa1641 (in quad if one want more current for very difficult loads...the opa1641 should be enough though for all the 47k input pre or amp...and surely even less Z...)
 
@abraxalito your Dark matter generator works a treat mate. Got it running at 15V rails because i'm lazy and need to match resistors. Is there any reason higher rails are better - less THD?

IMG_0798.JPG

So far I can't say it's miles better than the CEN I/V, maybe clearer, more front of stage. Great bass, sharp Vocals. superb engineering once again, straight out the box, plug it in and go. I guess the CEN boards have more sentimental value as successfully building any of Patricks designs is a great achievement, they don't build themselves.

@jimk04 you're a legend boy. Cheers pal.
 
The CEN IV also has an option for a third order Sallen Key filter.
It is not often used because Fran said he can't hear any difference.
If you prefer NOS, as in the Miro, then adding a high-Order low-pass filter seems contradictory.

I only use first order low pass way above 20 kHz, as I prefer to use oversampling myself.
And only with R-C. No inductors.


Patrick
 
Last edited:
  • Like
Reactions: passive420