1. Things you need to know about the new ‘Conversations’ PM system:

    a) DO NOT REPLY TO THE NOTIFICATION EMAIL! I get them, not the intended recipient. I get a lot of them and I do not want them! It is just a notification, log into the site and reply from there.

    b) To delete old conversations use the ‘Leave conversation’ option. This is just delete by another name.
    Dismiss Notice

Arcam and the TDA1541

Discussion in 'd.i.y.' started by fatmarley, Dec 18, 2006.

  1. fatmarley

    fatmarley Registered User

  2. martin clark

    martin clark pinko bodger

    Not really - the 10v reading will move slightly depending on temperature of either reg. You'd need to trim the 2K4 on the -15v reg to about 2430ohms to be spot-on, and that's simply not worth the effort.

    Also - given the rarity of good TDA1541s these days - I'd be happy to slightly under-run it on voltage in the interests of ensuring long life - these chips dissipate significant power to which the -15v supply is the biggest contributor.



    [size=-3]... though I have one that has survived 12v on the +5v pin... just dont ask...[/size]
     
  3. fatmarley

    fatmarley Registered User

  4. anubisgrau

    anubisgrau pfm Member

    Sorry for jumping into your tech talks, guys... just wondering if anyone with more experience with A5 mods can comment the following. I have just purchased Avondale modded A5 and I am awaiting for a delivery. On another fora I read several comments that AAA5 actually sounds grainy in comparison with Naim ACD3. As my system is hyper-revealing, obviously this doesn't sound promising so I am wondering if anyone noticed this issue and could it be somehow addressed with further mods. I remember Zanash mentioned once that he exchanged Les op-amps for something different and improved it (in his opinion).
    Any comments? Cheers.
     
  5. martin clark

    martin clark pinko bodger

    They both are modified to use non-oversampling, and - IME - a degree of graininess is definitely inherent in this choice, because it leads to aliasing within the audio band (surprisingly like vinyl surface noise, actually*)

    That's certainly not to say it can't be enjoyable though!


    [size=-1]*listening to an early ACD3 right now...[/size]
     
  6. fatmarley

    fatmarley Registered User

  7. anubisgrau

    anubisgrau pfm Member

    well, it is enough to mention ATC20 actives (a pro version).
    you hear every mistake upstream:)
     
  8. anubisgrau

    anubisgrau pfm Member

    hmmm, in a few days i will find out what actually some mean by "grainy sound". i have a problem imagining that.
     
  9. fatmarley

    fatmarley Registered User

  10. ron

    ron Tweaker

    My modded A5 sounds far from grainy;)
    Sort the supplies,output stage,clock out, bung in a double crown!
    I've not heard a AAA5 although I can't imagine it sounding grainy, all the reviews I've seen have been positive
     
  11. IDM

    IDM pfm Member

    Hi Ron,

    Thanks for the information on setting up the I/V conversion as per Pedja Rogic.

    I think I will go down the passive I/V route i nthe first instance and then build the AD844 circuit some time in the future. I think I will have a go at your idea of using a valve output stage coupled onto the AD844.

    However, before all that I need to etch another PCB for the DAC, so I need to confirm my basic understanding of grounding. I assume that the digital +4 and -5V supplies are decoupled to the digital ground, The I2S is relative to the digital ground and that the -15V is ddecoupled to the Analogue ground, is that correct? Secondly do you or at what point do you couple the digital and analogue grounds, a star earthing point?

    Cheers,

    Ian
     
  12. ron

    ron Tweaker

    Its quite a tricky dac to work with but the last one I etched I had a star earthing point on the bottom layer (not beneath any of the I/C's), all my grounds went to that including the msb/lsb grounds.
    I did a dual sided board, the I2S and output signals was on the top layer, the rest of the top layer was used as a screen which connected to the bottom star earth with a pin going through the middle of the board soldered both sides.
     
  13. IDM

    IDM pfm Member

    Wow that was a fast response!

    So from what you are saying, by using a star approach their is no requirement to decouple to digital or analogue grounds seperatrely as each point joins at the star earth?

    Cheers,

    Ian
     
  14. martin clark

    martin clark pinko bodger

    Sounds about right. 'Digital / Analogue Ground' usually refers to the parts connected to inside the chip, but externally they belong 'bolted together'.

    There is almost certainly some advantage in keeping the resulting decoupling paths/loops separate as far as possible though ie not crossing or sharing traces on the route to the ground plane.
     
  15. ron

    ron Tweaker

    I try my best:D
    Yes, for the actual dac board I had all the grounds going to a star earth although this was with passive I/V, the I/V resistors was also on the dac board, the 6c45pi output stage was built on a separate board, I had the I2S from a modified cdp (Marantz cd72)
    I managed to fit the dac board and 6c45pi board in the cdp's case, the transformers and psu was in a different box
     
  16. ron

    ron Tweaker

    Yep and this dac certainly needs plenty of grounds! I'd certainly use double sided if you can
    Mine looked like a spider underneath with lots of legs:D
     
  17. martin clark

    martin clark pinko bodger

    The right place, just musing about active I/V and the current loops involved.

    Output from the 1541 is a current going into the dac (that is an output of -2mA at idle. Using an active I/V stage, this is the signal current which must be sourced from the [active stage's] positive supply. Which I suppose should be decoupled to the same 0V point as DACs analogue output to close the loop as tightly as possible.

    Or another way of looking at it - separate the DAC too far from an active I/V stage and you risk signal-correlated errors due solely to the ground impedance between the two. Oh isnt this fun...


    edit for aftherthought: for 2v CD standard output, the LSB is 30uV. Peak current out of the TDA1541 is 4mA. So I think you need less than 30uV/0.004, about 4mOhm impedance between the I/V stage and DAC grounds for errors less than 1/2LSB. That's actually quite awkward at HF!
     
  18. ron

    ron Tweaker


    Isn't it just:p not exactly an ideal dac to work with especially for a beginner, still the best sounding chip out there to my ears!
    With the passive I/V I had the resistor grounds going to the star earth too.
    There was the fun with the oscillation from the 6c45pi's but thats another story:D
    I'm hoping to have a play about with a PCM1794 next, no idea what results I'll get with that
     
  19. ron

    ron Tweaker

    Reason I just got the pcb from Pedja to build the latest one with the common base I/V:)
     
  20. D.A.R.R.Y.L.

    D.A.R.R.Y.L. New Member

    Hi All,
    I am new here; first post.
    I own a Marantz CD-7 and have service manual in pdf.
    Maybe posting the diff-mode TDA1541A S2 config would be helpfull for al your
    design pleasures?
    Regards,
    Ramon
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice