Thursday, 22 November 2012

Engineer's Thumb

Request, with CA3080 instead of LM13700 (but you can use the conversion daughterboard to allow you to use the LM13700 if you prefer).

Designed my Merlin Blencowe and posted on the various DIY sites, this is how he described it:

For a long time I've felt guilty that my pedal board didn't include a compressor. Finally, after about a year of breadboarding, I have produced a circuit that can take on a Dynacomp and win, boasting:

Up to five times more headroom than a Dynacomp
Less noise than a Dynacomp
Same available sustain as a Dynacomp
Fewer parts than a Dynacomp
Cheaper than a Dynacomp
Feedforward side chain
Possibility for all five controls
Fits in 1590B
Millennium bypass

Despite its final simplicity, I went though just about every possible configuration of OTA while chasing acceptable performance. As you can see from the schem, the OTA is working as a current-controlled resistor in the feedback loop of an opamp. This reduces the contribution of OTA noise.

The input signal is also coupled directly to the side chain, which consists of a precision rectifier (U1b) and precision current source (U2a) that dumps more current into the OTA control pin as the audio signal gets larger.

The ratio control blends compressed and uncompressed signals, so its variable from zero to infinite ratio (limiting)!






76 comments:

  1. awesome! thanks for posting this. when i get back from our thanksgiving dinner i'll be building this one.

    ReplyDelete
  2. Thanks for posting, sounds like this is a must build thing!
    Too bad I don´t have all the parts around. Next time I order I guess...

    Will be fun to compare my MXR supercomp with this and the Rothwell Love Squeze I´ll do over the weekend. I somehow have the feeling that the MXR will get its ass handed to it.

    ReplyDelete
  3. i was actually hoping that Mark was going to post the 5 knob version instead of the two (or three) knob one.

    ReplyDelete
  4. Hmmm, it looks like the +9V and ground are shorted to me, so i think that you meant to move the ground lead up one row.

    ReplyDelete
    Replies
    1. Second that. But to me it seems like the ground lead should be two rows down. Taking it one row up would ground the non-inverting input of IC2, which isn't right...
      +m

      Delete
    2. yep. you're right. it should go two rows down from where it is now.

      Delete
    3. Ha ha. I moved the ground connection before exporting so the wires were spread out a bit more from where it was, and have gone for the wrong supply row. Layout updated.

      And incidentally John, you have everything you need to make the 5 knob version. Replace the 1M resistor at IC3 pin 3 with a 1M pot for a Threshold pot, replace the 100R with a 50K or 100K pot (or whichever higher value works best) for an Attack pot, replace the 1M next to it with a 1M pot for a Release pot. All simple 2 lug variable resistors.

      It's all there, but it also offers the builder the option of building the smaller 1590B friendly 2 knob option if they want to.

      Delete
  5. This looks very interesting. bc327 is a pnp? So im guessing 2n5087 would be an ideal replacement?

    ReplyDelete
  6. would the lm13700 option sound better then the ca3080 or?
    can't remember if the lm13700 actually is a double ca3080 or not..

    ReplyDelete
    Replies
    1. Should sound the same. The LM13700 is just like 2 x CA3080s with additional independent buffer circuits too, and none of those extras would be used anyway.

      Delete
  7. Build it. Makes a decent overdrive, but it's not a compressor..

    There seems to be something wrong with the layout. For starters, the 220R should connect CA3080 pin 3 vref, not +9V.. I'm not sure about other stuff...
    +m

    ReplyDelete
    Replies
    1. I'll check it over mate

      Delete
    2. Yes the 220R needs to go one row lower, layout updated. Nothing else stood out

      Delete
    3. having a look at the layout, there`s a shared link in the row below ic2 but nothing else is connecting in that row, why the link?

      Delete
    4. This comment has been removed by the author.

      Delete
    5. Yup. That makes it a compressor :) But i refuse to tag it yet.

      For some reason my attack time is close to 10 seconds, which means that playing chords with this is impossible..

      I'll eye it out...
      +m

      Delete
    6. Dexxyy: Because the 220R was connected to it.
      +m

      Delete
    7. Yes it's unnecessary mate. I do that before I start the layout properly so all the power connections for the IC are made and with one in between row also getting the power connections which is useful sometimes to allow filter caps and other components to connect to 9V or ground. In this instance the row wasn't required for anything else and so the double link could be removed. You could even lose that row completely if you wanted to to make the layout smaller.

      Delete
  8. Where does everyone get these strip boards that are 20 deep. Looks like tayda only goes 19 nowhere near 20+

    ReplyDelete
    Replies
    1. I get mine from musikding and banzai (depending on how much other stuff i need for the same order) Both have the same boards and i buy them as 10x16cm pieces.
      +m

      Delete
  9. Replies
    1. It says in the notes, increase the 100R to increase the attack. Or swap it for a pot if you want that control.

      Delete
    2. I didn't have time to check my build yet...
      +m

      Delete
  10. any further developments on this one?

    ReplyDelete
  11. Yup. I went through it today. My issue was with release more than attack.

    I took wires outboard from 1M and 100R, breadboarded the wires with 1M and 20K trimmers. For me, 20K is good value for attack, and 430K is good for release. I'm still going to decrease the ratio pot for 500K log, as the 1M seems to be quite unusable.

    Layout works, but those three values should be played with...
    +m

    ReplyDelete
    Replies
    1. Do you think it can be verified now then mate?

      Delete
    2. drumroll...............

      Delete
    3. Well :) I did tag it after i got it working the way i wanted. So i can say it's safe to build. Everyone building it should just know that those two (or three) values may need tweaking. This behaviour may have something to do with my CA3080, but only thing wrong with it was just the attack/release times to begin with.. (after that 220 found its home)

      I'm confident that the layout is ok.

      Mark, you could take the unnecessary strip away..
      +m

      Delete
  12. i built one of these (the 5 knob version) using another vero layout with an LM13700 and it works great, so next i'm going to build Mark's layout with the CA3080 just to see if there's any difference between them.

    ReplyDelete
    Replies
    1. Oh Please do! I'd love to get build confirmation from someone else too.
      +m

      Delete
    2. Hi Miro,
      i'll see if i can get one built up by tomorrow. i think that i see a couple of small differences in Mark's from the one that i built, but it looks basically the same.

      i also drew up my own LM3080 version before Mark posted his, but it has four standup resistors to keep the size of the board small (mine's 21x15), and i know how he hates standup resistors. ;-)

      Delete
    3. i just finished it and have been testing it (i built it with the 5 control knobs). it passes signal fine but it doesn't have any compression whatsoever. i'm going to disconnect all of the pots with the exception of the volume and ratio to see if it works that way.

      Delete
    4. Curious.. Mine had massive compression that did release in about 10 seconds. Tweaking the 100R and 1M helped... Could it be that those CA3080s are sensitive to something we don't know about?
      +m

      Delete
    5. i don't think that it's the 3080. i built another comp that uses it and it worked perfect the very first time. i'll have to check and make sure that i didn't make any errors on the build, but i'm pretty sure that i didn't.

      Delete
    6. okay, i removed the three extra pots and now it's just like Mark's layout above (i.e., it just has ratio and level controls). it works perfectly (and sounds very good!). either i screwed up the wiring with the additional pots, or i had a build issue somewhere else that got solved with removal of the pots. i'm going to start adding one control at a time now to see if i can make it work with all 5 of them again

      Delete
    7. i re-added the 'release' control, and it works exacctly as it should, however, i recommend adding a 150K-200K resistor in series with lug 1 so it doesn't create a short for the 1M resistor that the 1M pot replaces when the pot's fully CCW (i.e the fastest release). next, i re-added the attack control and it works as it should too. i remove the 100r reisot from the board and soldered it in series to lug one on the attack pot since you want 100r to be the minimum anyway (fastest attack).

      next up, the threshold control.

      Delete
  13. i tried to get the threshold control to work, but no matter how i wired it, i wasn't able to. it works perfect on my other vero layout build, and it seems super simple to install it on this one (and on the schematic here: http://www.freewebs.com/valvewizard2/engineersthumb.html , but for some reason i just can't figure out why it won't work, so i'll probably just leave it out on this one and use it as a 4 knobber.

    ReplyDelete
    Replies
    1. Still. Great to hear that i'm no the only one who got this layout to work correctly :) Thanks John!

      BTW.. I thought i'd build a(nother) compressor for our bassist - Any suggestions what design would be perfect for bass? This or some optical, like Forest Green?
      +m

      Delete
    2. did you install a threshold control on your's? and if so, how did you wire it?

      out of all of them that i've built, the Keeley 4 knob and the EMMA transmorgrifier are both awesome on bass. i raised the values of the input and output coupling caps on them for use with bass. besides compression, the EMMA can actually get some nice overdriven tones out of it too. i also built a pigtronix philosoher's tone that also sounds great with bass.

      Delete
    3. I didn't try on any extra controls... Maybe i should, as i haven't boxed it - it is laying around on my desk :)

      Currently our bassist is using my modded build of Mr. Squishy. That's ok - not bad at all, but maybe something with more control would be even better. I have Transmorgrifier board unboxes, so i could try that...
      +m

      Delete
    4. i like my other ET that has a threshold control 10X better since it can also be used as a clean boost, or even dial in some grit, so i ended up finding a way to add a threshold to Mark's layout. i'll post a pic of it, since adding a threshold is NOT just a matter of replacing the 1M resistor with a 1M pot. the other two controls (attack and release) ARE a matter of replacing two resistors and i prefer to have control over those parameters too.

      funny you mention it, because the Mr. Squishy is already going to be my next comp build.

      Delete
  14. and here's how it needs to be if you want build the 5 knob version.

    http://johnkvintageguitars.homestead.com/Effects/Compressors/Modded_Marks_5knob_engineers_thumb.png

    ReplyDelete
    Replies
    1. here's another pic to show the differences:

      http://johnkvintageguitars.homestead.com/Effects/Compressors/ET_mods_for_5_knob_version.png

      Delete
    2. :D Talk about mods! Great contribution! Thanks John!
      +m

      Delete
    3. thanks. there were a couple of errors in the first post pf the mods but the link has now been corrected.

      Delete
    4. Is there a layout for a 5knob LM13700 version?

      Delete
    5. scrap that, i found it.

      Delete
  15. here's a pic of both my CA3080 version, and my LM13700 version. These two comps sound identical (and sound really nice!).

    http://johnkvintageguitars.homestead.com/Effects/Compressors/2-ETs-01.jpg

    ReplyDelete
    Replies
    1. Wow. The LM13700 version is smaller :)

      I just have to ask.. Where do you get that green stripboard?
      +m

      Delete
    2. yes, the LM13700 version is smaller due to having 4 standup resistors. one of my local electronics shops carries the green vero. it's really high quality stuff. the trace strips are bonded to the fiberglass really well and bright silver plated copper instead of just bare copper.

      Delete
  16. thanks John, i built this today and it worked like a charm except the threshold pot;-(. now i know why.....;-)

    ReplyDelete
  17. So anyone dialed in the perfect values for all 5 knob pots yet?

    ReplyDelete
  18. the pot values posted are the correct ones.

    ReplyDelete
  19. here's a couple pics of mine using Mark's layout with the CA3080 chip. I call it the 'bass comp' since I slightly modded it for more low end. IMO, this is a great sounding and versatile compressor.

    http://johnkvintageguitars.homestead.com/Effects/Compressors/MerlinComp2-01.jpg

    http://johnkvintageguitars.homestead.com/Effects/Compressors/MerlinComp2-02.jpg


    ReplyDelete
    Replies
    1. Excellent stuff, thanks for the pics John. I'm about to buy myself one of those Brother labellers as a Christmas present to myself! Providing I can find a decent local source for the tape. The Brother ones seem a lot more common in the US than over here so all the tape sellers I can find on eBay seem to be in the US.

      Delete
    2. what did you change to get more low end, john?

      Delete
  20. a larger value input coupling cap.

    ReplyDelete
  21. That's the 10n on the bottom left, correct?

    ReplyDelete
  22. I've had a go at building this, but so far think I've made a mistake. Here is a thread with the details and voltages:

    http://www.diystompboxes.com/smfforum/index.php?topic=100697.0

    I'm going to have a look at this tonight, so any more thoughts and advice will be well received!

    ReplyDelete
    Replies
    1. As Merlin mentioned, the different voltages on the IC pins and transistor is very peculiar. The base and emitter are connected to pins 1 and 2 of IC2 via a solid track so it can't even be down to a bad joint. Check the link at IC1 pin 5 though. That connects that pin to the collector and if the solder joint isn't good there, that could account for the different measurement.

      Delete
  23. From my understanding the 3 extra pots are :
    1M for threshold
    1M fot release
    100k for attack
    Linear pots I guess ?
    Thanks very much

    ReplyDelete
  24. It would be any major difference between using BC327/25 or BC327/40 or BC327/16?

    ReplyDelete
  25. This comment has been removed by the author.

    ReplyDelete
  26. I've been struggling with this circuit for about 5 hours now, all I'm getting is an extremely fuzzy low fidelity distortion, with no compression whatsoever. Could it be a bad ca3080? Any suggestions would help. Thanks.

    ReplyDelete
    Replies
    1. IMO, a bad CA3080 is much less likely than you having an error in your build somewhere. first check for solder bridges, misplaced part(s) or a wiring issue.

      Delete
    2. As mind-blowing as it is, it was the CA3080. I finally bought another one, popped it in the socket, and now it works. I thought at first it could be the op amps, I had error checked the board several times and taken voltages, so I used some nice OPA2134s. Now, its a super transparent and hi-fi circuit. I love it.

      Delete
  27. John, could you share the layout for the 5 knob LM13700 version? I don't live in the states and it's hard to find a CA3080 for a reasonable price. Thanks a lot to everyone sharing here.

    ReplyDelete
  28. here you go:

    http://johnkvintageguitars.homestead.com/Effects/Compressors/Engineers_ThumbLM13700.png

    ReplyDelete
  29. I built this using a 2n5087, and it 'works' but it doesn't sound like i have a lot of compression even at max and i have to turn the level up all the way for unity gain... is this normal, or a byproduct of using the 2n5087?

    ReplyDelete
    Replies
    1. Radio Shack has some 2n3906's so i may go grab a couple and try them out. I built the 2 knob version.

      Delete
  30. so this doesnt rely on the ca3080 as much as the dynacomps and such? should i save my nos ca3080 metal can guys for other builds?

    ReplyDelete