Tuesday 10 March 2020

Keeley Fuzz Bender

Verified by AlexS


Demo by MTK


68 comments:

  1. Is the voltage getting boosted on this?

    ReplyDelete
    Replies
    1. No, the gyrator tonestack which consist of IC1 and IC2 runs both on normal power supply (9v) and negative power supply (-9v). I guess the reason is to give the tonestack more headroom to work on. The fuzz section (Q1-Q3) runs on normal 9v power.

      Delete
  2. Ok, thanks. That makes sense.

    ReplyDelete
  3. Do you know what trannies Alex ended up with? Might start this one soon.

    ReplyDelete
    Replies
    1. I think he used 2N2222's for the silicons and a GT404 for the germanium.

      Delete
    2. No problem for the silicon transistor, but a germanium NPN is difficult to find. Do somebody know where find any good replacement? Did somebody try, just for a test, a silicon transistor instead of a germanium?

      Delete
    3. Yes. I used what fusshead said. I just followed his previous advice on converting fuzzes to NPN, but you can try others.

      I just tested it out with silicon (2n222 & 2n3904) and it works for testing purposes, but was more gated than normal. You would want to get the germanium for sure.

      Delete
  4. More "common" NPN germs is AC176/AC127/AC187. Q3 needs some leakage to sound good and bias correctly. One trick you could try is a low gain silicon with a 1M-2,2M resistor across base and collector to simulate leakage. Tried it in a Tonebender Mk 2 and it works fine, at least in that circuit.

    ReplyDelete
  5. I'm seeing something weird. This is the schematic: https://www.freestompboxes.org/download/file.php?id=34330&mode=view
    Pin 1 of the 1044 should be not connected, but in the layout is not "isolated".

    ReplyDelete
    Replies
    1. The scematic is a bit unclear regarding pin 1. But pin has to be connected to 9v, otherwise the charge pump will cause lots of noise and oscillation since the internal boost function is running free and isn't bypassed by the voltage.

      Ses that those crosses at pin 1 and at one half of the Gate pot?. Both these points is connected to 9v.

      Delete
    2. Ok, thanks. Indeed, it's no so clear. But it's ok if the layout si verified and without any issue.

      Delete
    3. Comparing the schematic with the stripboard above, is the IC1 using 1,2,3 pins so that in the schematic 1,2,3 and 5,6,7 should be swapped numerically?

      Delete
    4. Sorry.. but if the X:es in the schematic go to 9V, and pin 1 and 8 do too, and the Gate 3, then the Red "9V" seems misplaced?
      Shouldn't it be 3 rows "higher", where the D2 has the stripe? Where the current "9V" red marking in the layout appears should be ground instead? Or the D2 is upside down, but then I am confused again. Sorry, I am a bit new to this, so please be patient if this is just a noob question...
      Love the tones and the demo, will try to make my own once all parts arrive.

      Delete
    5. ... and pins 6 and 7 on the 1044 go to X on the schema, but they go to sort of nowhere in the layout, should they go to 9V as well, or best left "hanging"? (Or it does not matter...)

      Delete
    6. The schematic at FSB is a bit confusing at a first glance with all those unessecary "x's"...
      Pin 6 and 7 should not connect to anything. Pin 1 and 8 connects to 9v through a polarity protection diode(D2).

      Delete
    7. Thanks, then the "red 9V" notation in the Layout picture above should be to the right of row Q instead of current T?
      (That would make 1 and 8 meet the D2 diode on the "stripe" side, and the link that takes it to Gate 3 as well all go to 9V.)

      Delete
    8. No, D2 is a series protection diode so the current source must be fed through the anode.

      Delete
    9. So it is upside down in the schematic diagram?
      There anode seems to go to ground... and pin 1 8 gate3 seem to meet at 9V.
      Sorry, it just does not match with 1 8 gate having 9V directly. Instead they connect to cathode (stripe side) of D2 and D2 gies to 9v. Whereas the schematic does not feel close to that. Sorry for asking so much what is obvious to you and others. I am teying to learn. Sometimes I take schematic and design the layout myself, something I could have thought is ok here. But that would never have worked then? I read it had some unclear bits, but the point of 9V, I can't see being the same un the two representations.
      I assume the layout has been built and confirmed so it has to be correct.

      Delete
    10. Oh, I see what you mean regarding D2. Keeley is using a parallel polarity protection which is not best solution IMHO. I almost always sub it for series polarity protection in all my layouts on the expense of a tiny voltage drop. The circuit will work the same until you plug in a power supply with wrong polarity.

      For pin 1 & 8. I almost always use the more common (and cheapest from reliable sellers) charge pumps avaiable like the one written in the notes under the layout. For these, both pin 1 and 8 must be joined together, otherwise you will dissconnect the Boost pin (pin 1) and will end up with squealing and noise. Slightly less common charge pumps like LT1054 and TC1044 almost always has pin 1 hanging free.

      Delete
    11. Many thanks! I thought the schema and layout were from the same, but now I understand. And the ref to Xes led me to that as well.
      I plan on getting TC1044S, and might as well keep it connected anyway as I interpret you as it is ok.
      Again thanks!

      Delete
    12. Any idea if "LTC1044CN8 (DIP)" would work for TC1044S? Or betyer with MAX1044 and connect pin1?

      Delete
    13. Seems to have max 9.5V vs tc1044 having 13V

      Delete
    14. Saw the price, will keep searching

      Delete
    15. MAX1044 or ICL766S will do just fine, they are not particulary expensive. Just stay away from Ebay or Tayda when you sourcing charge pumps.

      Delete
    16. Will try TC1044S from Das Musikding eventually

      Delete
  6. Holy shit this pedal rips! Super thick fuzz tones all the way to velcro city. Cleans up pretty decently with the guitar volume, but who's running this anywhere but full tilt? The Treble and Bass controls are insane. Out of all the pedals I've built, the low end on this things blows them away by miles. If you watch the video, use good headphones or speakers to hear what I mean. Laptop/phone speakers won't do it justice. Wow! I'm in love. Ignore that I've got 2 controls for the bias, I don't have a dual-gang A250K right now and just used 2 singles.

    Build - https://imgur.com/a/MdtrYOa

    Demo - https://www.youtube.com/watch?v=UXPY3ZGFJwU


    Thanks Anders!

    ReplyDelete
    Replies
    1. Holy moly, the low end rumble is sick! When you turned up the bass knob I couldn't help but laughing...

      Added you demo to the post.

      Delete
    2. It's incredible. I first tested this at my desk with a little 8" practice amp and it sounded like it was going to die with the bass up all the way. I want this EQ in every pedal.

      Delete
    3. The eq comes from the Boss MT-2, just a pair of different value, I think.
      What transistors in your built?

      Delete
    4. 2N2222, 2N1304, and a 1N270 for the diode. I'm going to try out some different stuff later today, after work.

      I've got some low gain 1613's in the 60 range and some 1302, 1304, 1306, 1308's.

      Delete
    5. What kind of range is good for a 2N1304?

      Delete
    6. http://www.njsemi.com/datasheets/2N1302%20-%202N1308.pdf

      Delete
  7. I have really difficult to find the NPN germanium transistors all of you are mention.
    I could buy AC127 and AC176K in my usual shop on line, with random hfe, or I have to spend more for a selection. (Probably just for the hfe they could find.)

    ReplyDelete
  8. I moved the bias over to 2 trimmers and tucked them away. I'm not going to be messing with the bias anyway. Set it and forget it. I reused an old 1590B enclosure that used to have a Dirt Transmitter in it. It was a very tight fit. But it's good to go. I ended up using a 2N1304 and staying with the 2N2222's.

    https://i.imgur.com/x9ajIaM.jpg

    https://i.imgur.com/mOKsri6.jpg

    No fun graphics for this one.

    ReplyDelete
    Replies
    1. Thats a tight fit for sure. Impressive!

      Just finished mine with 2N1711 for silicons and a 2SD72 for the ge. This thing rips!

      Delete
    2. Just boxed up mine, I used a 404b and 2n1711's. Gonna try some others but so far this thing is great! So much low end if you want it. Built this, the Pink Goat (which I love), and a Black Ash over the last few days. Now I gotta troubleshoot a colorsound overdriver that sounds like trash and that Coda Black Hole. I HATE fault finding. Hope everyone stays safe!

      Delete
  9. So I landed on 2 2n1711's with an hfe of 142 each and a leaky mn38 with an hfe of 120. Love it!

    ReplyDelete
  10. Hi another, satisfied build here. 2x 2N3440 for the Si and an OC141 for the Ge at the moment, but looking forward to seeing what else works.

    One thing - setting the fuzz at full CCW cuts the signal completely on my build. Is this the case for all y'all as well?

    ReplyDelete
  11. How are the lugs numbered for the dual gang pot?
    123
    456

    Or

    135
    246

    ReplyDelete
  12. Did I create the scheme correctly?
    https://disk.yandex.ru/i/VfN8PswmlBBN3Q

    ReplyDelete
  13. Hi! I just built this and I'm getting a significant volume difference when turning the bias to the right. Any idea what I should check? I used a 2SD352 Germanium PNP.

    ReplyDelete
    Replies
    1. Shouldn't it be NPN ? (No idea if this relates tonyour problem though)

      Delete
    2. My bad! The 2SD352 that I'm using is an NPN Ge (no idea where I got the idea for a PNP germanium).

      Delete
    3. Phew, one less thing. I didn't get around to build mine yet. Did you put a trimpot inside instead of a fixed R?
      When I've done that I dial it in and do not turn it much after. Except on Rift Octafuzz (Aion kit) where I made it an externa pot. Still ends up at 12 o'clock.
      Sorry, not much help. Let's hope some others can be more creative and helpful.

      Delete
    4. No problem and thank you for your help! Managed to solve it after finding a minor error in the Ge NPN pinout of the schematic I was using!

      Delete
    5. Great! Good to hear. Was it this schematic? And did you use a trimpot for bias instead of fixed R? If so, can you share details on what you used and how useful it is? Cheers.

      Delete
    6. I used a very similar schematic from another builder that made his own PCB and by mistake he used the wrong footprint/component (still a great PCB tho!). I used a 250K pot with 180k resistors as used here. I would definitely recommend having the Bias/Gate knob on the outside instead of a trimpot/fixed R - it can entirely change the sound of the fuzz and produce different sounds.

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

      Delete
    8. Thanks! So the 180k in the layout above (guessing top left corner, not the one by gate 4 and 6), you put 250k (linear or log/audio?) And mounted on the outside instead?

      Delete
    9. I think I see now. The "gate" here is called "bias" in original. A 250k dualgang. I saw another forum where a version with the dual split in 2 x B100k for more options. Got the 250 already and plan it on outside. I misunderstood your earlier post, but all is well now. 🙂

      Delete
  14. Finished building, but it does not work. sounds as if one or more wires are not well soldered so I tried re-soldering most of them first. Then it got quiet instead. Started probing, but that will take some time and I matched the schematic from FSB with the layout above. I know there are some differences that have been discussed and sorted above. 2 questions remain:
    The "Gate A, pin3" in FSB goes to an "X" but Gate 3 here goes to 9V? I can only assume the layout is correct and I'll try again to ignore the Xs in the FSB schematics.
    Second issue is I was stupidly thinking yellow caps are ceramic and red are Film Box, So I assumed C12 (at S2-T2) was 100pF. Which it is not. So will swap that out, but fear that it will not resolve my issue.
    I have some varying Q3s (from TV, Radio and phone sets and what not) that I am trying out so that could be an issue as well, although doubt all 4 would act up (on is an AC176k so it should be fine).

    BTW, the probing sort of had signal allover the place but in some places none, It was not like the signal got "this far and no longer", more random. But is was not very thorough so that remains to sort out.

    ReplyDelete
  15. Forgot: The 500pF, does it matter if I use 470pF instead? Or should I try to find a 510pF?

    ReplyDelete
  16. After going through the connectivity multiple times, I think it just manage to "settle in" and work. Only small worry is that the ICs get quite hot. Especially the 1044/IC3. Not burning hot, I can touch it. But not for very long. not sure if this is any concern.
    I also find that the Fuzz and Bias knobs do not produce all that well in their full swings, some settings seem to be just too much. I doubt this is wrong, more of a feature. I read in some other forum that the Bias knob affected the overall volume or tone, and I can confirm that. No biggie for myself though.

    ReplyDelete
    Replies
    1. Decided to disconnect lug 3 on the Gate so that TC1044S pin8 no longer went there and the heat is not gone but a lot less. I cannot detect any audio differences so I will keep it that way.

      Delete
    2. After some time, playing I noticed it was glitching again but somewhat differently. So before resoldering the wires to Fuzz pot I probed it. Much to my surprise it turns out to be a bridge between the copper traces around the 220nF cap by the Fuzz 1,2,3 side.
      I scratched the space between with a knife and the crossover connection stopped. And it works. I have never experienced a brodge like that forming seemingly over time. So probably something that was there all the time and being intermittent. Still strange IMHO

      Delete
  17. What sort of gain ranges should I be looking at for this? I've got 2n2222's from 110 to 200 hfe. I would normally have everything similar, including the Ge, but i just wondered if there is a general consensus before I get tinkering. Cheers, Ben.

    ReplyDelete
  18. I used the Si ones I had. Didn't measure them. Then I went through som 5 or 6 different Ge laying around until one was best. The difference was quitr minimal so prehaps a case om me wanting one to be better than the others.
    I nother forums and discussions it seems that hfe is not as decisive as one might think. Perhaps leakage? If you have the option then try out the ones you can and as long as the best is not "too little" of whatbyou want, then I'd box it.

    ReplyDelete
  19. I ended up leaving the "X"es alone, unconnected in the schematic I used. That was the part I was unsure about. It reduced the heat on the 1044 if I remember. There should be a link to it above.

    ReplyDelete
    Replies
    1. I found a schematic online that seems to match. My build is quirky. Volume issues and the Treble pot kills signal when turned all the way up. I'll keep futzing with it. It sounds cool but it's not quite right.

      Delete
    2. I will not have access to my hard drive with schematics until next week. Which schematic are you refering to?

      Delete
    3. Yeah, no worries. I found one that matches up. Only difference is how one of the 4558's is assigned. It must be my mistake. I did get the Zap Machine going. That one sounds great. But I'd love more control over the gating, like this one.

      Delete
    4. Oooh, I missed that you verified the zap Machine! I'll head over to that post and tag it.

      Delete