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

    I'm trying to get my head around this. Most of the designs I've studied have had either LTP or CFP front-ends, neither of which has any voltage offset.

    In the 729 is it as simple as 1 Vbe drop per emitter-follower (2 Vbe drops total), or is it more complicated than that?

    The 321 is more of a head-scratcher for me. The output side looks familiar enough with an emitter-follower being driven by a double-BJT current sink. But I'm not sure what the function of Q1 and Q2 are. Is Q1 just an integrator for the feedback? But then I'm still unclear on what the function of Q2 is....
  2. Jeff Young

    Jeff Young pfm Member

    Confession time: I still haven't finished my NAP250. I can't seem to find anyone to make a transformer with the matt black tape coverings. (I'll probably end up un-covering and re-covering a Toriody, but I haven't gotten there yet.)

    I have compared the HPA-1 with a NAD3020, a Bottlehead S.E.X, a Torpedo III and a Rogue Audio Cronus Magnum. The other 4 all have strengths and weaknesses; the HPA-1 is better than any of the others on every measure.
  3. jpk

    jpk pfm Member

    NAD3020 (actually a 3225PE) was my first amp :) I liked the buttons, they looked like Braun HiFi buttons...

    The Pass clones had a nice natural mid range and sounded open, the music had a breath to it, and wooden instruments had the right amount of wooden texture...
  4. DaveMc64

    DaveMc64 pfm Member

    Yes, that it as simple as that.

    For the 321, if you look at the feedback loop, you'll see it refers back to the emitter of the input transistor. This would normally expect it to mean there was a 1 vbe drop from input to output. However, there is also a bias current flowing through the 12k feedback resistor (and on through the 330k resistor) adding an additional ~0.5v drop meaning the output sits at about 1.1v below the input.

    I'd need a bit more clarity which ones you mean by Q1 and Q2 but if you're refering to those in the acoustica 321 drawings then maybe the way to envisage it is this: Q1 as the equivalent of the LTP input without the second transistor and feeback going to the emitter. Q2 is then the eqivalent of the VAS in a standard power amplifier and then Q5 is the equivalent of the output transistors but only single ended and being loaded by the current sink as you say. As I say this is a way of trying to understand it not as an exact explanation of how it works.

    As it happens, I have a spare NAP250 Holden and Fisher power transformer if you're interested. A project that never got off the ground.
  5. Jeff Young

    Jeff Young pfm Member

    That helps lots. At least I'll have a starting point when I stick it in SPICE.

  6. jpk

    jpk pfm Member

    :):):):):) Thanks a lot for your help!!!

    Which Toriody model would suit your 250 amp?
    Last edited: Jul 20, 2020
  7. jpk

    jpk pfm Member

    OT: the transformer for the NAP250 would need 2 secondaries with 40-0-40 each, right?

    So a dual supply for the NAP250 (or for two NAP135s) would need 4x 40-0-40...?
  8. Jeff Young

    Jeff Young pfm Member

    x-0-x refers to a centre-tap secondary, with the "x" being the voltage on either side of the centre tap. So a 40-0-40 is actually an 80V secondary with a tap in the middle.

    You can use each side of that secondary as a 40-0, in which case you'd only need two secondaries for a dual mono supply.

    Or you could use non-centre tapped 40-0 secondaries in which case you'd need 4 for a dual mono supply.

    Or you could ignore the centre-taps of a 20-0-20 secondary for each rail (in which case you're back up to 4 for a dual-mono supply).

    Oh, and putting the common ground between the speakers was a mistake that Naim later rectified. It should be the bar between the two capacitors.

    Somewhere I found a diagram of how Naim wired theirs, but I can't remember where now....
  9. jpk

    jpk pfm Member

    I attached a diagram to my last post: is that different from Naim wiring?
  10. Jeff Young

    Jeff Young pfm Member

    Oh, right, I forgot that Naim used centre-tapped full-wave rectification instead of bridge rectification. (Bizarrely with bridge rectifiers, but that's another story.)

    So both your diagram and your voltages are correct (with the exception of moving the ground reference to between the capacitors).

    All of my previous comments are for the (more common) bridge rectification.

    On the other hand, if you're going dual-mono I'd be inclined to use non-centre-tapped secondaries and bridge rectification. It'll be cheaper and easier to source a transformer that way.

    Here's a good run-down of the three types of rectification (half-wave, full-wave, and bridge): https://www.physics-and-radio-elect...uits/rectifier/rectifier-whatisrectifier.html
    Pete MB&D likes this.
  11. Pete MB&D

    Pete MB&D Pete Maddex, the one and only!

    Single center tapped winding with the centre tap to between the smoothers and single bridge his the best sounding configuration.

    Jeff Young likes this.
  12. jpk

    jpk pfm Member

    OK, I made a NAP135 version of the diagram:


    From the transformers 40VAC I should get 56VDC after smoothing, right...?
  13. Pete MB&D

    Pete MB&D Pete Maddex, the one and only!

    Yep 56v ish depending on the transformer regulation and load.

    Are the green wires connected at the smoothers, they should be.

  14. jpk

    jpk pfm Member

    Yes, the diagram is earthed between the big caps. So for two 135 style monos I would need 4(!) center tapped 40-0-40 transformers with 300VA each and rectify with 8 diodes in total?
  15. Pete MB&D

    Pete MB&D Pete Maddex, the one and only!

  16. jpk

    jpk pfm Member

    If "one bridge" means 2 diodes I can see how this could work for one amp, but it would half the power, right? It seems to me that the above diagram is similar to your suggestion but only doubles it to have at least full wave for each rail...?
  17. Pete MB&D

    Pete MB&D Pete Maddex, the one and only!

    A bridge rectifier is 4 slides in a package.

    if you look at the last diagram in the link that’s how my amps are wired.

  18. DaveMc64

    DaveMc64 pfm Member

    Going back to your 135 diagram, implementation is key to getting this right. There are large charging current passing down the 0v lines from the transformers. These need to attach closest to their respective capacitor terminals so, for instance, the 0v line associated with the blue windings needs to be the closest connection to the positive terminal on the lower capacitor.

    The advantage of the way Naim use centre tapped transformers in this instance is to stop the charging pulses travelling along the star earth. If you go for the more standard full bridge with centre tapped 0v configuration, then the star earth should be implemented differently.
  19. Jeff Young

    Jeff Young pfm Member

    ... ideally switching to a CRC filter. The charging pulses then go to the first set of caps while the star ground goes to the second; the two being coupled by the very low (but non-zero) resistance of a wire or bus bar.
  20. jpk

    jpk pfm Member

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