Electronics [Updated] CMoy headphone amp - Complete, but naked!

Discussion in 'Modding' started by Byron C, 6 Aug 2010.

  1. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    [Update 29/08]: I have now completed building the amp, but it's naked at the moment! See latest post.

    Hey all

    I'm planning to build a CMoy headphone amp using this schematic, but have a few queries when it comes to powering the unit. I plan to build this into a sort of desktop breakout-box (possibly one of these - if it'll all fit inside), along with 2 LED VU Meters and the HD44780-based parallel display I've just built. Got no problems with actually building the unit at all, however I do have a few queries with regards to the power supply.

    I plan to use a standard molex power connector to power the whole desktop box; I want to mount DC jacks on an expansion card blanking plate (separate jacks for 12v and 5v), hook them up to a molex power connector, then run power cables to the breakout box. The LCD display runs on 5v, hence the separate power jacks; keeps things simpler. I am assuming that my 400w Corsair PSU will be enough to run all this - none of these circuits look like they would make a PC PSU sweat.

    More importantly however, how do I run the CMoy amp on 12v? The schematics show two 9v battery sources in series. How can I replace these with the 12v line from the PC's PSU? I have read the tweaks section on adding a DC power jack, but it doesn't show me how to integrate that into the above circuit. It only shows the positive output from the DC jack going to the amp circuit, but how do I get the negative supply? I don't know enough about how that power circuit works to try and bodge it myself. Is it simply a case of replacing the batteries with the 12v power line? If so, what do I do with the ground line of the PSU?

    This is, of course, all assuming that the CMoy amp can actually run from 12v as opposed to the stated 18v - can it? It seems daft to run it from batteries when it's going to be in a breakout box on my desk :), and I don't particularly want to run a wall wart to the box if I don't have to.

    EDIT: Any help is most appreciated! :thumb:
     
    Last edited: 29 Aug 2010
  2. capnPedro

    capnPedro Hacker. Maker. Engineer.

    Joined:
    11 Apr 2007
    Posts:
    4,381
    Likes Received:
    241
    What op-amp are you using? The OPA132 only needs a +2.5v supply, so it'll be OK because what you'll be doing is splitting a 12v supply into a +6v supply.

    Replace the 18v battery supply with your 12v supply and you're done. You don't need the diode OR gate because you wouldn't have the batteries at all. You can miss out RLED and D1, but you still need R1 and C1 in the supply section, because that's what splits the supply into a positive and negative, with the virtual ground. You'll need to use this virtual ground in the amp section of the circuit rather than tying into a black molex wire.
     
  3. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    Wasn't expecting such a quick reply! :thumb:

    According to this parts list, it's the OPA2132PA.

    So if I'm reading you correctly, the power supply circuit would now look like this?

    [​IMG]

    And the molex ground wire is not connected to this or the amp circuit at all - or have I completely misunderstood?

    Many thanks! :D
     
  4. capnPedro

    capnPedro Hacker. Maker. Engineer.

    Joined:
    11 Apr 2007
    Posts:
    4,381
    Likes Received:
    241
    Maybe I didn't explain it very well, we'll say. When I meant the ground is unconnected, I meant all references to ground in the amp circuit refer to the virtual ground.

    [​IMG]
     
    Byron C likes this.
  5. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    I'm way out of my comfort zone with this "virtual ground" and splitting one supply into positive and negative stuff; give me just plain old supply/ground and I'm fine!

    That makes much more sense than my attempt!

    Many thanks for all the help :). I don't actually have all the parts to build everything yet, but none of this looks particularly expensive. Once I manage to get something up and running, I'll post up a thread.
     
  6. capnPedro

    capnPedro Hacker. Maker. Engineer.

    Joined:
    11 Apr 2007
    Posts:
    4,381
    Likes Received:
    241
    It's neither particularly expensive or overly complex at the moment. It should work out OK as is, but if you find the lack of volume* from using a 12v supply rather than an 18v supply be be an issue, that's when things start to get complex. I was thinking about this a little earlier and I reckon you'd probably have to end up using an LM7809 to get a +9v supply, and an LM7909 to get -9v. However, the LM7909 needs -11v or more negative so you'd basically take the yellow wire as Vs- and the black wire as Vs+. I'm pretty sure the output side has to have a common ground (the PSU's +12v) with the input, which in itself wouldn't be a problem. But you have to consider that you need a shared ground between the +9v and -9v supplies there is now a problem. The positive side's ground is the normal ground wire, but the negative's ground is actually the 12v line - and you can't just tie the two together.

    The problem needs some work... But it's not really a problem anyway, because the amp should run OK off 12v. In fact, a "dead" PP3 cell is ~7v so anything designed to run off (2) 9v PP3 batteries should function as normal from a 14v supply.

    * Theoretically, the output will have a voltage 30% lower than if there was an 18v supply, but due to your gain settings, you wouldn't have really been using the full (potential) range of the output anyway. Also, dB is a logarithmic scale, so you're only losing like a decibel or two of output or something anyway.
     
  7. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    The original design actually only calls for one PP3 split into +4.5v; as I understand, the addition of a second PP3 in series was in order to drive high-impedance headphones and/or get more volume from the amp. From what I've read, this amp design can deliver some pretty ear-splitting volumes so 12v sounds like it should be fine.

    The alternative is that I can run the whole breakout box from an 18v wall-wart and use separate voltage regulators for the circuits that need less than 18v (such as the VU meters and LCD).
     
  8. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    Well, I finally got round to buying the parts at the weekend, and I requested samples of the op-amp from Texas Instruments (which have now been shipped). Hopefully I should be able to get this running by the weekend :). I've already started mounting power jacks on a spare PCI blanking plate.

    Will keep the thread updated with progress. (Posting from my Desire, so apologies about any spelling errors!)
     
  9. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    Well, I've managed to put this together and it works brilliantly. Everything I've seen about this amp design has been confirmed - the sound quality is very good for such a small circuit, good bass response (but not overpowering) and clear treble.

    There's just one problem. If I plug my iPod or phone into the amp, it sounds great. If however I connect it to the speaker-out connection on the PC (on-board Realtek HD Audio), it's horribly distorted. At normal volumes, all I get is crackly noise - but noise in time with the sound that the PC is playing. If I turn the amp right up (and nearly deafen myself), I can hear the music but it's still heavily distorted. Adjusting the output volume on the PC makes no difference. I've also tried connecting the amp to my Behringer UCA-202 USB audio interface, but it sounds exactly the same.

    This is bound to be a question of impedance or volume levels, but I have no idea where to start debugging! Anyone have any ideas?

    PS., the basic design is as per the schematics I linked to in my first post. I found a veroboard layout here which I've follwed and I'm powering it from the 12v Molex on my PC; as per the initial plan, I now have a PCI blanking place with 5v and 12v jacks powered by the PSU.

    EDIT: Please note that I'm using different capacitor values than the veroboard schematic. The veroboard schematic on head-fi shows 470uF radial caps, but I'm using 220uF - as per the tangentsoft schematics. All other component values, including the caps on the amp input, are the same however.
     
  10. capnPedro

    capnPedro Hacker. Maker. Engineer.

    Joined:
    11 Apr 2007
    Posts:
    4,381
    Likes Received:
    241
    Can you change your soundcard's output from normal levels to line levels in its control panel?

    What happens if you power the circuit from a battery supply rather than the com0puter's PSU, does the distortion persist?
     
  11. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    Tell you what, I'll PM you next time :p ;). Seriously though, I do appreciate all the help - it's been invaluable.

    It doesn't appear so; at least not for the on-board RealTek. The Behringer only outputs at line-level, AFAIK.

    :duh: That's perhaps something I should have tried the first time, as that works... I've got it running off 8x AA rechargeables (@1.2v each), and the sound from the PC is fine (in fact it sounds very good if I plug the amp into the PC's Behringer audio interface) - not a hint of distortion. Portable devices also sound fine.

    So why would it sound fine when running from the PC's 12v and a portable device, but not running from the PC's 12v and the PC's sound output? Clearly it's something to do with power, but I don't know what.
     
  12. capnPedro

    capnPedro Hacker. Maker. Engineer.

    Joined:
    11 Apr 2007
    Posts:
    4,381
    Likes Received:
    241
    I didn't think of this before, but it's because you have a shared supply with the audio source, so the signal is biased against different voltage levels. Oh boy that's a major annoyance. I'm trying to think of a solution, but I've got a splitting headache and I just keep coming up with half an LM7909 based solution...

    There's probably a simpler solution like charging a pair of large capacitors (isolated by diodes) or a small battery and deriving the virtual ground from that.

    Or maybe isolate the input by using some sort of buffer.
     
  13. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    Don't lose any sleep over the power supply problem - to be honest, it sounds like using the PC's PSU is just overcomplicating matters. Using the PC's PSU is not exactly 100% essential, it was just a more convenient power source. I can use either PP3 batteries or a spare 12v wallwart. I am planning to re-design my current layout anyway, as it's a little inelegant. In fact, it's downright ugly - apologies about the evil flash:

    [​IMG]

    I do appreciate all the help, though.
     
  14. capnPedro

    capnPedro Hacker. Maker. Engineer.

    Joined:
    11 Apr 2007
    Posts:
    4,381
    Likes Received:
    241
    It's funny you should mention losing sleep - I've always found it hard to make my brain switch off at night to get to sleep so I tend to lose sleep by working on problems. :lol:

    If you don't mind using a wall-wart that would probably be the easiest solution which should work all right.
     
  15. Byron C

    Byron C Official Necromancer

    Joined:
    12 Apr 2002
    Posts:
    7,401
    Likes Received:
    2,028
    Heh, I know the feeling. Yeah, using a wall-wart is not a problem. In fact, even just using a 9v battery shouldn't be a problem - having used it on battery power for the last day or so, it sounds just fine.

    I've managed to come up with a stripboard layout that I'm finally happy with:

    [​IMG]

    I've also made a few modifications to my original design: I've replaced the 220uF power supply caps with 1000uF caps, to give the bass a bit of a boost, and I've also replaced the 0.1uF caps on the opamp input with 0.47uF metallised polyester caps. The bass has definitely had a massive kick up the backside and the treble still stays nice and clear. The volume is far more than I will ever need - I nearly deafened myself when testing it, as I'd left the volume pot at full :duh:. I've also replaced the RCA inputs and 6.35mm output with two 3.5mm jacks - far more convenient and compact.

    It's still naked at the moment, I need to find a nice small enclosure for it. Although I'm not sure I'll be squeezing it into an Altoids tin like everyone else!
     

Share This Page