Any ideas to improve IV/output of 1704 DAC and paraleling DAC's

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi there, looking to upgrade IV/amplification of this dac. dont like the sound( sintetics clear hights, dark midrange without texture,looking for more organic, open ,analogic sound)

as there is 4 chanel dac (actualy 5+1) and i use only two( thinking to simply paralel them (2+2) by paraleling data input and cuting wires from df1704) shoud

any ideas apreciated, thanks :)
 

Attachments

  • daccc.jpg
    daccc.jpg
    1,018 KB · Views: 214
Looks like a lot of work to make that better. Fix/improve one area and the remaining areas that need work may be more audibly exposed.

Looks like clock distribution design is far from ideal. Maybe running from a PLL instead of a crystal too, can't tell since it looks that signal comes in from some other board.

Easy things to play around with might include replacing the I/V opamps with OPA1612. Replace any caps in the audio path with at least 50v or 100v NPO, good film might be better (wound foil and polypropylene film are quite good but not cheap, good metalized film down a level but depends on exact cap design). Decoupling caps for opamps might be easy to swap, again good, low inductance, low ESR film can help a lot for decoupling analog circuitry (not ideal for digital though), but it depends. Have to try one thing at a time and see if you like it better, different, worse, or no change.

Clocking quality tends to affect sound quality significantly too. Change I/V so you like it better, and, again, you may just expose the next lower level problems more clearly.
 
Last edited:
Hi there, looking to upgrade IV/amplification of this dac. dont like the sound( sintetics clear hights, dark midrange without texture,looking for more organic, open ,analogic sound)

as there is 4 chanel dac (actualy 5+1) and i use only two( thinking to simply paralel them (2+2) by paraleling data input and cuting wires from df1704) shoud

any ideas apreciated, thanks :)

Is that a Denon AVR-5700?

If you don't like the DAC sound do you like the sound of the power amplifiers with an external source? [Some people like those power output devices and the power amp design.]

If you do like the sound of the power amplifiers with an external source and only use two of the DAC channels have you considered an external DAC? Have you heard one that you like that could be used with the AVR?



If you don't particularly like the sound of the power amplifier with an external source and you also don't like the sound of the DAC then you might consider that some people do like those AVR models and perhaps you would be better off selling it to someone who wants that DAC & Amplifier. (And getting a different one with a sound that you like.)

I say that since the PCM1704 chips are pretty expensive online and some people do like them. (They are in two pieces of my favorite gear.) The power amps have nice output power transistors too.

I don't think paralleling two PCM1704 will change the basic character of the sound. (If you don't like the character of the sound.)
 
its a Denon AVC-A1D

Well I like sound now much better, i did small upgrade to unit which seems change sound a bit

1) one of power suply modified and hided in front panel with custom heatsink. powersuply caps shunted wth variuos MKP (20-30psc) Need to fight with milimeters here.

2) dedicated ultrafast diodes and custom heatsink

3) omron DPDT relays for 5.6ohm load when speakers protection unload amp

4) video input/ output pcb boards removed and together with 1) enabled to fit second heatinks(total 4 fans and 4 heatinks) fitin is so tight it needed to machine aluminium – fight for milimeters not centimeters

5) 5 channel darlington type power amp section remooved(left only termo sensors) and dartzeel model 108 clone fitted. I needed DPDT realys to protect it (it would blow up and no feedback-goes to osicaltion) when amp security checks and enables outputs

6) removed/bypassed every single amplification from preamps,inputs left only toshiba analog volume regulators, around 26 operational amplifers left out

7) remooved ALL coupling caps in the signal path. there were 9 signal pssing caps per channel (45psc total) no DC ofset of any kind after removal in any secton, not sure why they used so many Elna Silmic caps in signal path..

8) bypassed cuted all signals from pcb to main boards and back, changed with silver and cooper wiring. , most not needed relays signal paht bypassed.


Even with those small changes I would say amp+ dac combo sounds completely diferent, if I would not know I would say its diferent amplifier!


I also thinkink for the last work – to paralel not used channels dac chips and replace OPA IV and OPA gain stage DAC IV with jfet analog stage.
 

Attachments

  • 147506814_1834621653352951_6305930783135092823_n.jpg
    147506814_1834621653352951_6305930783135092823_n.jpg
    661.9 KB · Views: 143
  • 147660178_440896163726320_3217368295885111821_n.jpg
    147660178_440896163726320_3217368295885111821_n.jpg
    661.5 KB · Views: 136
  • 147687609_1379515692397000_6839147114902537222_n.jpg
    147687609_1379515692397000_6839147114902537222_n.jpg
    530.9 KB · Views: 115
  • 143029614_489911918665498_5765448888906200242_n.jpg
    143029614_489911918665498_5765448888906200242_n.jpg
    418.2 KB · Views: 118
  • 144691247_329696401660665_5263846874766596393_n.jpg
    144691247_329696401660665_5263846874766596393_n.jpg
    183.9 KB · Views: 63
  • 145065109_237089984576530_7333809748248200471_n.jpg
    145065109_237089984576530_7333809748248200471_n.jpg
    313.3 KB · Views: 71
  • 146733552_420915862493889_1086709604541496156_n.jpg
    146733552_420915862493889_1086709604541496156_n.jpg
    726.4 KB · Views: 70
Last edited:
Looks like an interesting project! I like the approach suggested by Extreme_Boky.



The output current of the PCM1704 is quite low and it seems to benefit from CFB based I/V. I used the LME49713 for I/V with a PCM1702 in a Classe CDP.3 and it improved in a significant way.



When using a CFB, you may need to experiment with different feedback resistor values from 1.2K to 3K depending on the board layout and power supply decoupling. For this application carbon film resistors may have an edge due to the inductive nature of some metal film types. Capacitors should be removed from the I/V op amp feedback loop for stability when using a CFB.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.