1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
  2. Where we are now: The site is back up and running! The character set encoding issue has been fixed so old posts should now look ok. I am currently working on themes/skins etc and I will be deleting the temporary green skin very soon as it has many issues. I will do a very quick variation on the stock blue skin first just to get a light-grey background and san-serif font, and I will set this as default. Later this week I will hopefully add a professional third-party skin in pfm colours (life is way too short to do this crap myself, and I've found one I really like than needs next to no tweaking). We are getting there slowly...
    Dismiss Notice
  3. May I please request people check and update their email address if it is out if date. I ask as I’m getting countless bounce errors as the server fails to send notifications through etc. I’ll eventually figure out how to send these to a black hole somewhere, but it makes sense to address it at source as quite a few folk are obviously not getting the thread and conversations notifications they have set in their preferences.
    Dismiss Notice
  4. 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

Death and Life of a Nait3

Discussion in 'd.i.y.' started by nickcase, Oct 1, 2014.

  1. BobMaximus

    BobMaximus pfm Member

    It could just be a problem with the heatsinking of the output transistors? Are they screwed down properly or has the paste dried out? If the heatskink itself isn't getting warm, then it's certainly that. Thermal runaway is the expected behaviour if so.
     
  2. a.palfreyman

    a.palfreyman pfm Member

    I'm no expert but this posting (low bias current across the 0R22 but still running hot) suggests that despite low bias current, there is a large current "going somewhere" to heat the output trannies as MJS said in an earlier post. These two factors suggest that there is a connection passing current between the cases (collectors) of the output devices. Check the insulation of them to the heatsink because it looks to me that current is going somewhere to cause this heating. An ohm or so between the two cases would still allow current to pass through the 0R22s to allow you to set the bias but would be lower than normal.
    AP
     
  3. hacker

    hacker Delicious and nutritious

    Les adds a 68uF BC128 // 100nF film across that position on the NCC200.
     
    nickcase likes this.
  4. nickcase

    nickcase Enlightened Member

    Thanks for the input everyone. Tr17 is mounted underneath the board, as such, only by taking the board out revealed a BC182B in place of a ZTX384. Replacing it with a BC550C has sorted the overheating.

    Now have a sensible stable bias voltage of 7.5mV on both channels (across both 0R22 resistors.) DC offset of 15mV at the speaker posts too.

    One remaining difference is that powering up with a load on the left channel makes the offset at that channel 200mV. Plugging in a load after power on doesn't change the offset. Other channel doesn't care. Load is a crappy victim speaker.
     
  5. martin clark

    martin clark pinko bodger

    Sounds to me one of the coupling caps is a bit leaky, to hold such an offset - either feedback or input, and I think of the two - I'd try a swap for the input cap for anything else, regardless of value, first. Also check r27 and R28 are in spec, or at least close (these are non-critical values) and properly connected - no dry joints on R28 esp.
     
    nickcase likes this.
  6. nickcase

    nickcase Enlightened Member

    Thanks Martin, plenty of 10u Wimas available but I don't understand how the signal input section can be in play since the preamp is completely isolated, i.e. the NAP90/3 board has no connection to the Nac92r board.
     
  7. martin clark

    martin clark pinko bodger

    I'm thinking - and might be wrong! - that having a small load on the output at switch-on effectively 'bolts' the feedback line to 0v.

    What parts then might develop or hold an offset voltage, at switch-on? Has to be developed somewhere (front end is likely) and 'stored' which means one of the larger time-constant caps and only then if feeding an effectively-large impedance.

    I would expect it to bleed-down to ward the few-mV range fairly quickly though... something, probably quite simple, isn't right.
     
    nickcase likes this.
  8. nickcase

    nickcase Enlightened Member

    Thank you. I must say that your familiarity and flexibility of thought within a circuit is very impressive.
    I am grateful for helping me learn, reminds me how I've missed this place and just how many unstarted projects I still have.
     
  9. MJS

    MJS Trade: Witch Hat Audio

    If the output has been +ve DC for any length of time then the feedback capacitor gets reversed biased and can go leaky. That's usually the cause of these hundred mV offsets.
     
  10. martin clark

    martin clark pinko bodger

    Doh. of course, I keep forgetting its a tant...
     
    nickcase likes this.
  11. nickcase

    nickcase Enlightened Member

    What's worse is that I've found the pre-board has a fault and dumps 24V onto the ch1 input! Luckily the boards are currently separated.

    So, the feedback tants must go - I've some 47u MMK's and some lytics lying around.

    Despite the circuit diagram, on this poweramp board there are no input caps at C9, simply wire links. Wima's can go in here but do I need them? Probably not.
     
  12. MJS

    MJS Trade: Witch Hat Audio

    I'm not sure I've seen that on a production NAIT3 but seeing as the pre-amp and amp are always connected to each other then you only need the one tant between the preamp output and amp input, so no need to put one in.
     
  13. nickcase

    nickcase Enlightened Member

    This is no production pedigree unit. It's a hatchet job 'Nait3' comprising my old 92R, a randomly acquired 'sold for spares' 90.3 board and a CD3 trafo.

    The 90.3 board was a mess and is getting worse with every lifted track.
    The 92R has been round the block - at one time it was a remote control for my starfish :)
     
  14. MJS

    MJS Trade: Witch Hat Audio

    You're doing better than me then. I can't find the effort to stuff a chipamp inside a fried NAP90 case.
     
  15. grazie

    grazie pfm Member

    Yes, of course. 😁
    Thanks Martin and mr wha for giving up your own wealth of knowledgeable so freely.
     
  16. nickcase

    nickcase Enlightened Member

    Don't beat yourself up about it, must be 10 years ago I first conceived this project. There's a clever witicism in your last post involving 'chip'amp and a fried case but it eludes my whisky brain this morning ��
     
  17. nickcase

    nickcase Enlightened Member

    Nearly there, fixed the pre-amp (dry via on the gain section) and broken track on one channel of the power amp but the overheating left channel is still a problem.

    This channel dumps 5V DC across the output when powered up with the speaker connected. At start-up I measure 4.5VDC across the speaker, this increases rapidly to 5.5V, at which point I turn it off.

    Powering up with no load and connecting speaker afterwards is fine (12mV DC offset), but playing music at a very modest volume overheats the channel within 5 minutes. The right channel is fine. The board is hanging in the air and not screwed down to any heatsink.

    So far:
    C9 (input cap) is a wire link.
    R27, R28 and R29 all measure within 5% of their value.
    C11 feedback is now an MMK (measured at 47u)
    R36 R33 measure within 5% of their value
    Diodes D8 D9 check out.
    TR23 and TR24 check out when tested with the meter diode tester.

    Should I try replacing C10 (470pF) or C13 (100pF)? Not got any suitable alternatives in stock.

    Any other suggestions gratefully received.
     
  18. MJS

    MJS Trade: Witch Hat Audio

    Even the quiescent current in these will fry the output devices in no time - they have to be bolted to the chassis.
     
  19. nickcase

    nickcase Enlightened Member

    Sorry, perhaps I misled; I've also tested running with the board bolted into place with proper heatsinks.

    The left channel still gets excessively hot within 5 mins whereas the right is just warm. The left also cannot handle any load at power up, several volts appear on the output and a large current is drawn, only the limiter I have in place prevents the amp pretending it's a toaster.

    I can hook up the speaker after power is applied or, if power has been applied recently; i.e there is still some charge in the reservoir caps.

    Tempted to swap the reservoir caps to see if the problem migrates channels.
     
  20. MJS

    MJS Trade: Witch Hat Audio

    Reservoir caps are shared and won't be the issue here. It'd be worth changing the R22 emitter resistors as they're only metal film rather than wirewound IIRC and I've seen silly faults with these going open circuit or shorting to something.
     

Share This Page