Advertisement



  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

Starfish revival

Discussion in 'd.i.y.' started by jpk, Jul 10, 2020.

  1. Jeff Young

    Jeff Young pfm Member

    Jun sent me some more stuff on the 552, and I think Stereophile's description is a bit misleading. A bunch of those regulated supplies are eaten up by other stuff: a couple for the DAC TTL level stuff, a couple for the clock, one for the volume/balance control motors, one for the front panel illumination, one for the source relays, etc.

    Near as I can tell on the audio side there's just one set for the TA filter stage and one set for the gain stage. Interestingly, it does *not* appear to separate the left and right channels.

    This would suggest a single pair (+/-) of rails for 1 & 2 in the schematic in the first post, and a second set for 3 & 4, for a total of 4 rails.

    If one wanted to improve on that, I'd separate left and right (for a total of 8 rails) before separating 1/2 and 3/4. But my guess is that Naim did their homework and 4 hits the sweet spot.

    On another note, we *have* to use the TO-3 cased regulators. It's just so much of the Naim "look".
     
  2. Jeff Young

    Jeff Young pfm Member

    Oops, I was looking at the 555PS (for the CD player). Back to the drawing board...
     
  3. jpk

    jpk pfm Member

    That's my fault, I did not find info on the 552, but some pics of the NAC S1 instead:

    [​IMG]

    [​IMG]

    So the same circuits again! But no TO3 regulators this time: instead local TO220s near the golden caps...?
     
    Last edited: Jul 18, 2020
  4. Jeff Young

    Jeff Young pfm Member

    I spent some more time with that picture of the insides of the 552 posted earlier.

    It *does* have separate left & right supplies, so that's good.

    Focusing on the bottom channel (I haven't the foggiest if that's right or left), we have orange + orange/black feeding the gain stage on the left, then two pairs feeding the TA buffer: blue + grey/blue and pink + pink/black.

    So I think Naim is separating the two stages of the TA buffer, but not the two stages of the gain stage. That gives 6 rails per channel for 12 total. (The 552PS spec says it' has 7 pairs; I suspect the final one goes to the phono stage.)
     
  5. Jeff Young

    Jeff Young pfm Member

    Those black boards are nice, but I want my TO-3 cans back! (Yeah, I'd guess those TO-220s are the regulators.)
     
  6. Jeff Young

    Jeff Young pfm Member

    Jun, in your first PSU are the gyrators a Starfish addition, or did Naim use those in the HiCap or SuperCap?

    My first inclination is that if you want better than a 317 you need to go discrete.
     
  7. jpk

    jpk pfm Member

    The gyrators are a PFM mod (by Mr. Tibbs, see here). I don't think the 552 would takes phono boards - rather I thought the right most section could be the tape out buffer circuit fed from the remaining neg/pos regs?
     
    Last edited: Jul 18, 2020
  8. Yank

    Yank Bulbous Also Tapered

    Does the S1 have a phono stage?
     
  9. jpk

    jpk pfm Member

    It's nowhere mentioned that the statement has phono. The 552 has no phono, I just checked, but can power the external Naim phono stage "Stageline" which is standard Naim 24V powered, see here. The Stageline plugged into the 552 will then be powered from the 552PS.

    Jeff, are you sure you want to re invent the S1...??? :D:D:D
     
    Last edited: Jul 18, 2020
  10. Yank

    Yank Bulbous Also Tapered

    Why does this preamp need so many stages of low-pass filtering? I count five if you include the feedback around the second stage of buffering.

    Would this work that much worse?

    [​IMG]
     
  11. jpk

    jpk pfm Member

    The earlier Naim preamps used the 324 circuit as input buffer:

    [​IMG]

    Later the 324 was used for tape out buffers, and the input buffers were upgraded to the 729 TA buffer.
     
    Last edited: Jul 18, 2020
  12. Yank

    Yank Bulbous Also Tapered

    What are the two inputs for?
     
  13. jpk

    jpk pfm Member

    I don't know, maybe for phono or if a second pair was used as tape buffer. It seems 324 boards without the second input populated existed:
    [​IMG]

    Also 729s with only one input populated existed:
    [​IMG]
     
    Last edited: Jul 19, 2020
  14. Jeff Young

    Jeff Young pfm Member

    Well, yeah, that's the $1000 question. A single buffer (324), a double time-aligned buffer (729), or none at all?

    I think part of the need for a buffer is to isolate the input impedance from the tape out. So what if you have no tape out? That's complicated by the fact that the buffers in question also include low-pass filtering (as mentioned).

    I've heard that the NAP250 sounds most "Naim" when used with their pre, and the supposition that this had to do with the filtering.

    The NAP250 isn't terribly wide-bandwidth to start with (SPICE tells me about 70KHz). The power amp's input filter has a 102KHz knee, so while it might help stability it's probably neither here nor there in terms of sound.

    The 729 TA's input filter is even higher (129KHz), so again I can't imagine that having a material effect. The filter in the middle of the TA looks to be about 66KHz, so now we're starting to affect the bandwidth, but not hugely.

    There's also 2 more tants in the signal path with the filter in. I wonder if those don't have more to do with it? One data point which points that way is the idea that the more modern Naim preamps are more generic and less "Naim" than the old ones were. The modern ones are +/- 12V, which is going to leave the voltage differential across the output coupling tant at around zero, as compared with around 12V for the +24V single rail versions. This is going to take a lot of the tant's distortion out of the picture. (And perhaps it's that distortion that contributes to the "Naim" sound.)
     
  15. jpk

    jpk pfm Member

    As far as I know only the 552 and the S1 are not powered from 24V PSUs such as Hicap, Supercap, Flatcap etc....
     
  16. Jeff Young

    Jeff Young pfm Member

    Yeah, that's what I meant by the modern ones being +/-12V. That leaves the signal centred around 0V, while the older +24V ones will have the signal centred on 12V (before the coupling cap, which removes the DC component).
     
  17. jpk

    jpk pfm Member

    So why not adjust the signal to zero and forget the signal path caps...? Or are they a good thing?
     
  18. Jeff Young

    Jeff Young pfm Member

    A hotly-debated subject, that. Some folks hate coupling caps. If you're using the differential supply (+/-12V), you could indeed dispense with them. But you'd be relying on your source to have no DC offset (likely, but not guaranteed), and any fault which produced a DC offset within the source or preamp might get amplified by your power amp (if it also had no coupling caps) and destroy your speakers.

    Personally I don't have anything against coupling caps. There are certainly good and bad implementations of them, but in this case, I actually think they're partly responsible for the "Naim" sound that I remember being quite fond of.

    My main pre-amp is considerably more sophisticated than the Naim designs, with a discrete regulated differential supply, CFP JFET input, Wilson current mirrors, complementary MOSFET output and DC servo. Even with the DC servo, it still has coupling caps. https://www.diyaudio.com/forums/pass-labs/300060-pass-hpa-1-a-13.html#post5937416
     
  19. DaveMc64

    DaveMc64 pfm Member

    I think the second input was connected to the mono switch on the 32.5 - the boards with that section not populated were probably tape buffer boards (maybe from a 72) where they would have been redundant.

    The 729 boards also had routing for the second input but unpopulated on my boards. Probably only populated for 32.5 owners who wanted to upgrade to 729s without losing mono functionality.


    Unfortunately not with the circuit as is. Output DC voltage sits below the input dc voltage for both the buffer and the gain stage so you're going to need output coupling capacitors at least even with a split supply.


    The 729 is, I believe, configured as a bessel filter hence the Time Aligned marketing. I've simulated it in the past and it has a -3dB point around 45kHz. Would help with any ultrasonic grunge from CD players of the time when they were introduced.

    In the absence of a buffer, the input impedance from any input except the phono would have been 20k from the pot. Probably a bit low for 70s tape machines to cope with. Probably not so much of an issue these days.
    That said, whenever I've tried bypassing them, I've always ended up putting them back in.
    cheers
     
  20. jpk

    jpk pfm Member

    I only know Aleph 5 and U.P clones, they sounded considerabely less involving than the Naim amps I compared them to. Also with the Pass clones it was difficult to follow the individual vioces in the middle-lower range when music got crowded... How does the HPA1 with your NAP250 sound?
     

Share This Page





Advertisement


  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