DAC IV and buffer, need your contribution

[andrea_mori]

I'm working on some DAC board using old style DAC chip like PCM56, PCM63, PCM1702, PCM1704, AD1862, AD1865, TDA1541 and TDA1543 in NOS mode.
I'm building these boards to be used with Ian's Fifo buffer (Asynchronous I2S FIFO project, an ultimate weapon to fight the jitter) and Ian's I2S to PCM daughter board (Drive NOS AD1865/62,PCM1704/02/63,TDA1541 from FIFO: Universal I2S-PCM driver board)

These boards can be configured as single or balanced mode and use a single resistor as IV convertor, preferably an high quality resistor such as Vishay bulk metal foil.

Then I'm planning to implement a Zanden filter (Zanden Parallel DAC Questions) and finally a tubes line stage.
I would like to avoid any capacitor in the signal path, so I'm thinking to work around a totally dc coupled tubes line buffer.
I'm seriously thinking to a Circlotron circuit, since in balanced configuration I can get opposite phase signal to drive directly the Circlotron.
But any other dc coupled tubes design could work as well.

Any suggestion will be appreciated.

Andrea

[regal]

Sounds nice but you would have to decide on a DAC chip first. I think its too late to do a Zanden stack though, as even the lowly PCM56K is not available in quantities. Only the PCM1702/4 can be bought of the shelf, go with 2 per phase.

Since most all of been done to death anyway maybe (with Ian's I2S-PCM) its time for a deep dive into an NOS PCM1704 board. Its the only one that can take hirez audio.

HD-Tracks to an NOS PCM1704 may be very interesting. I can show you how to set up the PCM1704 to work with passive I/V. Ideally it would be laid out to plug right up to Salas regs.

As far as the tube stage if you really want dc coupled Circlotron output you will need a timer relay in the signal path, don't know if thats any better than a cap, but I'm interested to see what you come up with. Personally I'd stick with the proven WE417 anode follower, or possibly a spot on the board for a nice stepup transformer then a DHT like the 4P1L (most linear tube available), or skip the transformer and two DHT's loftin white.

Lot of possibilities. But I think your best bet is the PCM1704, lot of people listening to 24 bit music nowdays and yours would be one of the first NOS dac's that could play hirez without nasty truncation and no dithering required.

[andrea_mori]

Hi regal,

Sounds nice but you would have to decide on a DAC chip first. .

I'm developing all these 8 boards (7 if I cut 1702) to compare them in the same configuration. Consider each board will contain the chip DAC only (2 or 4 for balanced output depending on the chip) and the passive IV (2 to 4 resistors) plus a few other components.

I think its too late to do a Zanden stack though, as even the lowly PCM56K is not available in quantities. Only the PCM1702/4 can be bought of the shelf, go with 2 per phase.

Still a few hundreds PCM56P-K in stock from Mouser, Digi-Key, Verical and Arrow. But I think the Zanden filter could be used for all the DAC chip, only for 1704 I need to change the notch center frequency due to 96kHz fs.

Since most all of been done to death anyway maybe (with Ian's I2S-PCM) its time for a deep dive into an NOS PCM1704 board. Its the only one that can take hirez audio.

I usually play 44.1/16 materials only, but I'm very curious to try 1704 in this configuration.

HD-Tracks to an NOS PCM1704 may be very interesting. I can show you how to set up the PCM1704 to work with passive I/V.

Your suggestions are wellcome!

Ideally it would be laid out to plug right up to Salas regs.

Personally I favor series regulation (I'm just testing my own series regulator and seems to be very quite - The Well Regulated Power Supply), anyway these boards do not include power supply, so any option is possible to supply the DACs. I'll provide Kelvin connection for all analog and digital supply.

As far as the tube stage if you really want dc coupled Circlotron output you will need a timer relay in the signal path, don't know if thats any better than a cap, but I'm interested to see what you come up with.

I'm not sure I'll need a relay in the signal path, since the Circlotron is naturally balanced. I would like to know how Atmasphere had approached the problem in his preamps, maybe Ralph Karsten could satisfy my curiousity, if he will read this thread.

Personally I'd stick with the proven WE417 anode follower, or possibly a spot on the board for a nice stepup transformer then a DHT like the 4P1L (most linear tube available), or skip the transformer and two DHT's loftin white.

My first option is a dc coupled tubes line stage like a Circlotron, I'm thinking to 5687 triode for that.
The second option is a single tube with output transformer like Lundahl LL1674 or LL1930. This could save complexity, probably I could eliminate anti-aliasing filter, since the transformer will cut the upper frequencies.

Lot of possibilities. But I think your best bet is the PCM1704, lot of people listening to 24 bit music nowdays and yours would be one of the first NOS dac's that could play hirez without nasty truncation and no dithering required.

Surely the 1704 is one of the chips I'll investigate. I look forward for your hint.

Andrea

[andrea_mori]

Just got some 5842, some C3m and some 5687.
The idea is a differential pair based on 5842 or C3m in triode mode with a CCS, direct coupled to a 5687 circlotron output stage and direct biased from the differential, that also act as phase shifter to drive 5687 bridge. No transformer, no caps.
DAC filter like Zanden filter, to minimize the phase shift in audio band.
Anyone knows where could I get some custom values air core inductor?