Thursday, 2 May 2013

Opamp 3 channel splitter/mixer ON HOLD

Updated 6th February 2014

I have a TC Electronic G-Force and Eventide Eclipse in the parallel effects loop of my amp and use a rack mixer to keep them in parallel with each other, and so allow me to use them both 100% wet and so avoid having to digitize my dry signal.  I do have OCD about keeping my core dry signal out of the digital domain completely.

But I also have an Eventide Pitchfactor and would like to put this in too, and although the rack mixer can accomodate 4 in and 4 out, adding the Pitchfactor to an already limited rack space would be a pain, so I wanted to make a little 3 channel utility splitter/mixer to use instead of the rack mixer to save 1U and allow me to put two pedal trays in my 6U rack.

I used the splitter scheme you can see at Paul in the Lab, and a generic mixer circuit I found so that everything can be done with one quad opamp.  I may even make this with 3.5mm jacks for the send and returns, make up my own high quality cables and then this will all fit comfortably in a 1590B.  It would pretty much be a set and forget utility for me so I suppose I could even use trimmers instead of pots, set the levels and tuck it away somewhere.  So I thought I'd add it here and it may be useful for others to adapt for their own uses.



42 comments:

  1. I have a feeling this will be extremely useful, and it's a heck of a lot simpler than a lot of the other mixers. As a bassist, you really want a dry signal underneath a lot of your effects, and this may be easier than incorporating a boatload of JFET blenders into all the things I build.

    ReplyDelete
  2. NICE! This is going to be my next build!

    ReplyDelete
  3. Incidentally, you can of course omit the channel level controls if you wish. If this is being used to put effects in parallel, and those effects each have their own level controls then an additional one may well be superfluous for your use. In which can just take the return jack straight to the board.

    ReplyDelete
  4. If I wanted to use this for a clean blend could I just link one of the send-returns for the clean part or does this circuit rely on there being some kind of buffer in between? I'm thinking of adapting this to add a ckean blend to a fuzz pedal.

    ReplyDelete
  5. Yes you could do that no problem

    ReplyDelete
  6. I was going to do the 4 channel mixer until I saw that people were having problems with it.

    I think I'll trust this one and just go for it!

    ReplyDelete
    Replies
    1. One person. And it is verified...
      +m

      Delete
  7. Could this be adapted to suit an a/b/y pedal?

    ReplyDelete
    Replies
    1. Yes but you wouldn't need any of the mixing components or the returns, and obviously only two sends. Probably the easiest way would be to connect two of the sends to their own stomp switch. Only a SPDT would be required, or DPDT if you want LED indication. Like this:

      1 - From Board
      2 - To amp
      3 - Ground

      So with two stomp switches you can turn each amp on and off independently, or both on of course. And amps bypassed with have the connection to the amp grounded to prevent noise.

      Delete
  8. Has |v|ark or anyone else successfully built this? I've given it a couple of trys and I can't figure out why there's no signal getting through any of the channels. I may get around to troubleshooting it on a breadboard and will update if I figure out what I'm doing wrong.

    I'm a long time listener, first time commenter, so I want to get in a thanks to |v|ark & Mirosol & the regulars for sharing & verifying.

    ReplyDelete
    Replies
    1. Can you post the IC pin voltages

      Delete
    2. I've tried a couple of different TL074's and they both show about the same voltages. I just realized pin 13's v changes with the master volume pot. This is with the master volume pot all the way turned up:

      1: 4.45v
      2: 4.45v
      3: 3.09v
      ---
      4: 8.86v
      ---
      5: 3.09v
      6: 4.45v
      7: 4.45v
      ---
      8: 4.45v
      9: 4.45v
      10: 3.09v
      ---
      11: 0.0v
      ---
      12: 0.0v
      13: 7.95v
      14: 7.95v

      Delete
    3. The voltages on the non inverting inputs of the splitter channels don't look right to me on pins 3, 5 and 10. The two 10M resistors create a voltage divider there which should be closer to 4.5V.

      Forgetting the mixing channel entirely for the time being, are you getting nothing at all from any of the Sends?

      Delete
  9. I finished it, but output volume is about the half of the bypassed volume.
    Also each channel pot works, but even with 0 position, I can still hear the effects I plugged in.

    Ivlark mentioned about the voltage on pin 3, 5, 10. I got 3.2, 3.1 3.1v.

    ReplyDelete
  10. i built one, but only hear hum sounds. Do I have to connect the gnds together?

    ReplyDelete
    Replies
    1. Yes, all grounds must connect to the negative pole of the power supply.
      Thanks
      Dave

      Delete
  11. After a discussion on FSB I've updated this layout to what I believe will now do the trick. I think the problem with the last version of the layout was that I grounded the non inverting input of the mixing opamp channel, this now connects that to a half reference voltage. Changes have also been made to the input and so the new board is slightly larger and you're probably going to have to start again with a new board. Sorry to those who have already built this.

    ReplyDelete
  12. Has this been verified by anyone??

    ReplyDelete
    Replies
    1. Did you build the first or second revision?

      Delete
  13. Ivlark,

    Please help me.
    I made the second revision, channel 1, and 3 works fine, but ch. 2 pot is not working right.
    Also when I turn the master pot, the sound goes off after 9 o'clock.
    I checked my build several times, but I can't seem to find any faults.

    ReplyDelete
  14. The second revision got me very close to where it should be. Even with the master pot problem, right before the volume goes off, the amount of output volume is the same as input volume.
    ch 1 works good, ch 2 gives click sound around 9 o'clock and the volume curve is a bit strange than ch 1 and 3.
    ch 3, I shorted send and return jack for dry signal.

    ReplyDelete
    Replies
    1. Are you sure the opamp isn't damaged? Post all your pin voltages to see if it gives any clues about a problem area (other than the obvious one around ch2).

      Delete
    2. 1- 4.64v
      2- 4.64
      3- 4.22
      4- 9.29
      5- 4.22
      6- 4.64
      7- 4.63
      8- 4.64
      9- 4.64
      10- 4.22
      11- 0
      12- 4.64
      13- 3.53
      14- 8.61

      Delete
    3. I don't have an extra opamp at the moment.
      Also all the pots made crackle sound when I turn them. :-(

      Delete
    4. The voltages don't look too bad actually, until you reach pin 14 which is far too high and I can't see anywhere obvious that may be coming from. Crackles may signify a faulty pot or probably a poor connection somewhere. Can you post a front and back pic of your board

      Delete
    5. Ivlark,

      Here is the link of the pictures.

      http://weplaymusic.co.kr/187


      I do love this small box.
      This is the best utility box since I made loop box.
      I just need to make it 100%.

      Delete
  15. Just built this and am having very similar problems as Jaemin. Wondering about a possible error... Mark, in your post from May 2013 you mention the "10m resistors" and in the Paulinthelab layout he specifies 10m resistors but the layout above calls for 10k resistors everywhere and I built off of that. What are the odds that swapping up for the higher values will fix these issues?

    ReplyDelete
    Replies
    1. Yes originally I did it based on Paul's splitter and a generic mixer circuit, but someone on FSB mentioned that it wouldn't work like that and posted this correction
      which the layout is based on. So instead of 2 x 10M resistors making a voltage divider directly at the non inverting input, two 10K resistors were used and then the non inverting inputs connected to it via the 1M resistor after the input cap.

      And yes you're right about the number of tracks, that happens sometimes when I make a layout I've already finished slightly more compact but forgot the change the note.

      Delete
  16. Also, the layout calls for a 16x12 vero, but my count has it at 16x11 (I caught it befor I started making cuts and ended up with an extra row. No biggie obviously...

    ReplyDelete
  17. Just quadruple checked my build, track cuts, off-board, etc. Can't find anything obvious. Here are my voltages:

    1: 4.66
    2: 4.66
    3: 4.23
    4: 9.33
    5: 4.23
    6: 4.65
    7: 4.65
    8: 4.66
    9: 4.66
    10: 4.23
    11: 0.0
    12: 4.66
    13: 4.66
    14: 4.66

    Please let me know if there is anything I can do to problem-solve this further.

    ReplyDelete
    Replies
    1. Your voltages look pretty much text book to me, exactly what you would hope to see. But there is obviously an issue which is why I've now marked this "on hold" and so put it to one side for a while and let me look into it a bit deeper to see what could be amiss.

      Delete
  18. Hi all! Long time lurker, occasional builder here.
    I've built this with the original layout last summer, but only the splitter part worked, so I've used it as a splitter since. I really needed a small mixer, so I managed to make it work, though I have no idea why it works.

    SOLUTION: Put a 33k resistor from pin13 of the opamp to ground. The 47uF cap is also necessary (it was missing on the first version).

    This gives you the mixed output of the 3 returns. It seems to be a bit higher than unity, on full volume I get some nice overdrive on the low B of my bass, but it is fine on lower volumes. I've soldered my resistor from the master pot to the ground lug of the output jack, as I was too lazy to pull the whole thing out of the box. As soon as I get my hand on a scope, I'll debug this properly.

    Anyone care to enlighten me why my solution works? I'm not real big on theory, and this makes no sense to me. Used my sick bending skills to stumble upon this solution :P

    ReplyDelete
    Replies
    1. Thanks for the info. That would make the schematic something like this:
      http://oi62.tinypic.com/10x57bc.jpg
      and I still can't see why it would be required. It really does my head in that I'm fine with much larger effect circuits, but something as simple as this has caused problems and I don't know why! :o)

      Delete
  19. I solved Jaemin Choi's problem by putting a B25K for Master instead of B100K.
    Really helpful tool. Thanks a lot, IvIark.

    ReplyDelete
  20. Hello, I´m a beginner here, I just finished this building, this my 3rd one but wanted to give it a try because it seemed to be perfect for some sound I was looking for. After finishing it had the same problem than Jaemin Choi so did as nicolas.lesage suggested, using a B25K instead of the B100K. After that it just works perfect, keeps the sound of volume and every channel works perfectly well, no noises, simply fine. By the way I switched channel one straigth to keep always one dry signal, and use channel 2 and 3 to connect effects. Also added a DBDT switch to cut the dry signal and leave only the sound of channel 2 and 3, and a 3DBDT switch to bypass the entire thing. Plus one led on each switch to know its position. So from my point of view you can remove the ON HOLD tag to VERIFIED ;) Thank Ivlark for your job and dedication here and everibody else for their contributions, this is just great!

    ReplyDelete
  21. Could some one explain what you would use this for?lol.

    ReplyDelete
  22. Hello, I hope this thread is not dead because I have been looking over a schematic I drew up of this in order to understand what the problem is. I have a few questions and maybe some insight.

    I am curious why the mixing stage is an inverting amplifier. You could have done another voltage follower (like the splitter stages), only with higher gain. It might look like: 100nf capacitor coupled input straight to pin 12, 22k from pin 13 to GND, 100k pot from pin 13 to pin 14.

    I am also curious as to why 10k resistors are used on the return. They make a voltage divider with the 100k. The 100k master pot, when divided by the 10k resistors, will give the stage a gain of 10, whereas if you replace the 10k's with 1k's, you would have a gain of 100 when the pot is cranked all the way.

    This might explain why someone earlier stated that a 33k to ground worked out for them, only that is not a good way to execute that. I would think it creates motorboating.

    ReplyDelete
  23. I've thought about it, and the input impedance of the inverting op amp is divided between three channels, so perhaps lower the 10k input resistors --I think you could get away with something like 470R. Or you could increase the gain as I said before... you might do better with a 1k or 10k from pin 13 to ground. I have not bread boarded this. I bread board circuits like this all the time and its not difficult to find which values to swap out.

    ReplyDelete