Friday, 1 March 2013

Analogman Sunface BC108

Info from Analog Mike about his silicon original:

We have some BC108 and BC183 silicon transistors, and have been making silicon Sunfaces for several years now. The BC108 are in small metal cans while the BC183 are plastic. Silicon fuzzfaces came out in about 1969. Dallas Arbiter used both of these types in their Fuzz Face pedals. The silicons are fuzzier than the germaniums, and brighter. The silicon transistors have much higher gain. They don't get totally clean by turning the guitar down like the NKTs but they do clean up quite a bit and are sensitive to your dynamics. Check out the sound sample below to hear this.

We can make a SUNFACE with no extra charge for silicon. There is a small extra charge for modifying and converting an existing germanium fuzz pedal to silicon. The sundial is not really needed, as the silicon transistors are not temperature sensitive. But you can use the sundial knob to dial in different sounds if desired. We do put the BIAS trimpot on the inside of the two-knob sunfaces so you can still tweak it if desired. The Sundial knob may cause more noise on a silicon (high gain) sunface, due to the extra wires, so it may be best to not order the sundial if you want to keep the noise down.

Silicon fuzzfaces are negative ground, so you can use almost any power supply without fear of melting it down. But it still may sound better with an old style battery. The FUZZ knob on the silicon version usually sounds and works best if you turn it down a bit (on the germanium I usually like it up full).

The BC183 transistors we got in 2011 are my favorite for a silicon fuzzface. They are a little warmer than the BC108 and still have plenty of fuzz. We got a large box of these, each individually hermetically sealed in a foil pack, with the same paint stripes on the back that they used in the original 1969 Fuzz face pedals. I think Eric Johnson has BC183 in the red fuzzface that he used for years, and the new 2012 Dunlop Eric Johnson signature Fuzz Face seems to use these too.

Jon Carin (Floyd, The Who, Roger Waters, etc) contacted me in 2012, looking for a fuzz that sounds like the giant lap steel on One Of These Days by the Floyd. I recommended the BC183 Sun Face for him and I think he likes it!

I'd probably suggest using a 10K for the Sundial.  5K + 2K2 doesn't even equal the usual resistor in that position in a Fuzz Face (8K2) and so I think you want to give yourself a bit extra.






82 comments:

  1. My main shop doesn't carry 250k pots. Is it better to use 500k or 200k?

    ReplyDelete
    Replies
    1. 200K may cut some more low end, 500K would let more low end through, so make a decision based on what would be preferable to you.

      Delete
    2. I used 100kA as the output pot and it still sounded great. I used a slightly larger output cap (12nF instead of 10nF).

      Delete
  2. Mark- should the output on the board be volume 3? also, is there any "factory" setting for the trimpot, or just set it by ear? i built this with a 10k for the sundial and some BC183's and it sounds real good so far, so you can tag it. its getting late here so i'll play around with it some more tomorrow when i can actually turn it up and see what it can do.

    ReplyDelete
    Replies
    1. The default setting for the trimpot would be all the way clockwise, basically it's just acting in the same way your guitar's volume pot would. Since Fuzz Faces are known for their ability to clean up nicely with the volume knob Analogue Mike obviously thought it'd be a good idea to set a trimmer on the board to mimic this so you could get the same effect without having to use the pot on your guitar. Personally I prefer these circuits with it rolled back a little to give something that's a little more like an overdrive than an all out fuzz, plus it cuts a little bass so gives the impression of more mids and highs which is good if you play with a band. Hope this helps.

      Delete
    2. Oh and yeah you're spot on with the output to Vol 3 :D

      Delete
  3. Wow, requested yesterday, layout today, awesome !
    Thanks a lot Mark

    ReplyDelete
    Replies
    1. What sort of HFE should the q1 and Q2 108c's be? The datasheet shows 108C to be in the 100 to 170 range which seems a little on the high side for a FF circuit. Cheers

      Delete
    2. Reading the description, I think they're meant to be high gain..

      ''The silicon transistors have much higher gain. They don't get totally clean by turning the guitar down like the NKTs but they do clean up quite a bit and are sensitive to your dynamics. Check out the sound sample below to hear this.''

      Delete
    3. 108s and 109s can be great in FF even with a high hfe. What I've found is that the relative gains are important...so Q2 in a silicon FF should be about 50 higher than Q1. So 120 and 170 will be great. As long as Q2 biases around 4.5 and Q1 around 1.6 you should be fine. 109s are even higher gain and can sound wonderful too.

      Delete
  4. Thanks Vince, will breadboard it first and try a few 108's.

    ReplyDelete
  5. I did this one this morning before going to work. I used 2N5088 si transistors as I don't have BC108s yet. They seem to have similar specs (hfe 300-900).
    Also I socketed the trim and used a 10k resistor instead (I really need to source trimmers asap...)
    And I used a 5k pot for the sundial
    So, the effect works, but I would not tag it yet for the following :
    - The sundial pot does nothing at all (will try with a 10k pot)
    - With the fuzz control below 3/4, very nice low gain and soft fuzz, then suddenly in the last 1/4 it becomes high gain and very trebly/harsh
    Might be the 2N5088, I'll try the BC108C as soon as I get them (about 10 days from now). And maybe a log pot would do the job better ? I'll try and let you know
    Denis

    ReplyDelete
    Replies
    1. I replaced the sundial pot with a 10k, now I can definitely ear a difference, i really like how this control smoothes the sound
      OTOH, I put a 50k lin pot for the trimmer, can't ear any difference on all it's range
      I tried MPS18 trannies which sounded pretty good

      Delete
    2. That trimmer may well not make too much if any audible difference with the fuzz turned up high. It's used to create a setting in conjunction with the volume pot of your guitar to try to fine tune the low gain clean up when you roll back the guitar volume to the level you want. Roll back your volume and then mess with the trim pot to see if the sound gets dirtier in one direction and cleaner in the other.

      Delete
    3. I'll try that thanks Marc

      Delete
  6. This thing rocks big time! Following on from the sunface germanium build I put this on the other half of the board. I did mess with it to get the bias spot on ;in my opinion). I used a 10k sundial as suggested, and swapped out the 2.2k resistor for a 4.7k to make the sundial more usable. I also swapped out the 470 ohm resistor for a 220 ohm to help with biasing.
    The trannies I used are an odd bunch but the work brilliantly, I was shocked at how good it sound...seriously. I used a 2n2102 in q1: 89 hfe, and a bc109c in q2: 470 hfe. Odd mix but sweet. Thanks for all your help guys! I'm over the moon, maybe to the dark side even.... Cheers

    ReplyDelete
  7. I found this far too much of a pain in the arse to be worth it. Sounds just like a standard silicon fuzz face but a hell of a lot more messing about. Biasing is ridiculous and if not right it's noisy as hell.

    I would run a mile from this and make a standard silicon. Just seems like another pointless re-hash.

    ReplyDelete
  8. It does require a bit of fine tuning but I love it. Mine was noisy as hell until I got the biasing down, it stared me in the face and screamed at me at full fuzz settings, but now she's super quiet. I quite suprised at the difference actually. I think its well worth it. I learnt a heap about fuzz circuits and biasing which is half the point to me :)
    TONE BENDERS NEXT........

    ReplyDelete
    Replies
    1. So Jeff, can you explain how you biased this one ? (Quiet is not exactly how I would define mine :-)

      Delete
  9. The way it originally was set up I couldn't get the voltage I was after on the collector of q1 ( Approx. 1.4v if I remember correctly. I may have to get back to you with that voltage). I was using two bc109c with around 470 hfe each. Sounded super bad. I then stumbled across this fuzz face calculator and thought, what the hell... http://www.diystompboxes.com/analogalchemy/emh/emh.html
    I punched in what I had and found according to the program, my q1 hfe was too high with the 33k resistor on its collector. I plugged in some lower hfe figures for q1 and found that I was probably better off using a q1 hfe around 90. To make it work I also replaced the 470r with a 220r.
    Powered her up and bingo, q1 voltages now spot on. Measured q2 voltages and adjusted sundial for around 5.4v and marked the pot so I had a reference. Sounded great after that. Very quiet. The 2n2102 could be helping too, not too sure. Hope this helps mate.

    ReplyDelete
  10. adding my view to that of Jeff Lauder:
    I also had a bad experience with 2 BC108C so I swapped the first one for a 2N2369 and the magic happened.
    So...
    1. 2N2369 (hfe=80) and BC108C (hfe=450).
    2. 10kB sundial trimmer
    3. `00kA Volume pot
    4. 12nF output cap.

    You get instant Dark Side of The Moon guitar Solos and Jimi Hendrix Band Of Gypsys |(Gypsies).

    http://fuzzquest.blogspot.co.il/2012/11/the-pink-face.html

    ReplyDelete
  11. Yeah, tone is great with BC108Cs but at maximum fuzz, the artifacts are unbearable.

    Will get some BC183s on monday, and try with 2N3904 in Q1 as well

    BR

    ReplyDelete
    Replies
    1. Just a comment. Purchased a bunch of BC183C. HFE goes from 500 to 800, as with BC108C.

      When people talks abou HFE 120 for Q1 and 170 for Q2.... which trannies are they talking about????

      Delete
    2. Well,

      Finally BC183s: Q1: 490hfe Q2: 525, tow "low" gain (if we talk about C rated trannies) and Q2 collector set at 4,6v.

      Not by far one of my favourite silicon fuzzes. Too harsh and very noisy when fuzz knob set full CW.

      I know that with a lower hfe Q1, would sound quieter and surely better, but then would be not an Analogman Sunface, just an "X" silicon fuzz face.

      Actually, to my ears and taste, when talking about silicon fuzzes, you can't find anything that gets near to a DAM Meathead Deluxe.

      I don't think this one will get boxed :(

      BR

      Delete
  12. My take on this one :
    At first I didn't like the sound : way too harsh and very noisy.
    I wanted to stick with BC108 as it was a request from a friend.
    So following Jeff's post, I took a look at the calculator on diystompboxes and here is what I came with :
    33k --> 6k8 (to match Q2's gain, ie 506)
    470r --> 220r (to stick to the calculator)
    Sundial 5k --> 10k as suggested
    1µf input cap --> 2,2µf
    Output cap 10nf --> 100nf
    Q1 is hfe 470 and Q2 506
    Sounds much much better now, less thin (thanks to the cap swaps) but still bright, as this is to expect from this circuit.
    I also added a 47µ filteing cap between the +9V and the ground. Now the circuit is very quiet, it did the trick
    I might swap the fuzz pot for a log one (all the fuzz amount arrives at the end of the pot course) and the vol one for a lin

    Now I need to learn how to play time's solo :p

    http://nsa34.casimages.com/img/2013/04/06/130406073406563326.jpg

    ReplyDelete
    Replies
    1. Very nice looking build, thanks for the info

      Delete
    2. Thanks Mark
      Finally I'll stick with the 1k lin and 250k log pots

      Delete
    3. Damn this looks good! I'me going to do the same applications you did.
      Will this work with germanium transistors however? I used AC128 and liked their sound better than those bc108's

      Delete
    4. The Sunface NKT275 is the germanium version and can be found here:
      http://tagboardeffects.blogspot.co.uk/2013/03/analogman-sunface-nkt275.html

      Delete
    5. I know. I meant to ask if the same modifications would work well for the germanium version too.

      Delete
  13. Very faint volume. When I hook the circuit up, it all actually works - but it's the tiniest, weakest output you've ever heard. It's as if something in there is doing just the opposite of amplifying. I have tried different sets of BC108s, swapped transistors, double and triple checked each resistor, cap, pot and wire placement. It is all there. In fact, all the knobs work appropriately as well. I even tried BC184s. Very, very odd - i even took the trimpot out and tried some regular fixed-value resistors in there

    This has been a strange build. I purposefully built a few smaller projects before I took this one on and my soldering and other basic techniques are down. I've checked all the points for shorts, checked direction of the electrolytic, all solder joints check out, all wires are securely soldered to the vero board. I'm stumped. Given that the circuit "seems" to be behaving as it should - albeit with a huge volume cut, are there any common places to check that would have that effect?

    Thanks a bunch,

    ReplyDelete
    Replies
    1. Assuming you don't have a part failure, the vast majority of times this happens, it's down to a cold joint somewhere. Check them all, reflow if necessary, including all offboard including pots, switch and sockets.

      Delete
    2. i'll reflow everything. it's a small build :) I'm using a test board i made up where I just run the wires into busses on a breadboard that are already connected to jacks, led and stompswitch - those offboard things are all verified. It's got to be something on the board somewhere. Thanks for the tip.

      Delete
    3. Just this weekend i debugged a friend's build of lunar module that was working ok, but extremely quiet. Probed it to find out that one of the resistors was 3.3R where it should have been 2.7K. I'm thinking he was trying to sub 2.7K with 3.3K, which would have worked.. But most of the signal just died in the voltage. So i would say that another common error could be a wrong component value in some place...
      +m

      Delete
    4. I had the same on the crunchbox
      My pulldown resistor was 220R instead of 1M...

      Delete
    5. i will double check all the resistor values. I did find that I had a 1u cap where I was supposed to have a .01u :D but oddly enough, even fixing that it all sounds the same. no output. I didn't see any pulldown resistor in this circuit - unless i just don't know where to look. The elec. cap is 22u right? It's kind of fuzzy.

      I reflowed everything, double checked the pinouts of the transistors, Even had the pot lugs all reversed at one point, so i've tried them both ways haha. It may just be something I wind up having to rebuild. It's a small circuit at least. I socketed the transistors so they wouldn't get any heat damage. Very strange indeed.. Appreciate all the tips.

      oh - i guess I should check voltages at the transistors? I did trace the voltage around the circuit - i was getting around 4.5 on the base of the right-most transistor.. can't remember what I was getting at the base of the other. I assume one transistor is responsible for the clipping and the other is part of the output amplifier, so i was wondering if something was wrong there. You have to twist the legs on the first trans - so i've made sure they aren't touching or shorting out..

      Delete
    6. well, after much procrastination - i just rebuilt the thing, checking the first build while i made the second. Strange, because it is absolutely identical. reflowed everything, even used the BC108s from the first build in the new build as well as all the original pots and offboard assy and it works like a champ. Who knows. Very nice sounding fuzz so far, although it's late and I can't really crank anything up :)

      Delete
  14. Replies
    1. Thanks for the heads up, video changed

      Delete
  15. From what I can see ditching the internal trimmer will have no impact on the circuit except I might need to toggle my guitar volume a bit more? It's only trimming the input signal right? In which case consider it ditched haha

    ReplyDelete
  16. Any suggestions on what would be a good pedal to box with this
    was thinking a Les Lius ??

    ReplyDelete
  17. If I wanted to use a trim pot instead of the sundial for biasing, how would I go about doing that

    ReplyDelete
    Replies
    1. Just get a 5k trimpot and wire it the same way. One less hole to drill. :)

      Delete
    2. Should it be on the vero board

      Delete
    3. Sorry stupid question, I'll just raise the amount of rows by three and use links

      Delete
    4. That's certainly a decent option, if you have enough space in your enclosure.
      If not, just use short wires and leave it hanging offboard and use plenty of electrical tape to isolate it. Worked for me on more than one build. :)

      Delete
  18. Ok so if I don't want the sundial knob at all how shoupd I change the build? Speak to me like an ape because I'm really new at this stuff.

    ReplyDelete
    Replies
    1. Swap the 2K2 for something between 8K2 and 10K (you'll have to experiment for the best value, socket the connections so you can quickly swap and change it if you like), then put a wire link between the Sundial 1 and Sundial 2 & 3 connection points.

      Delete
    2. And when you're done, have a banana. Those are yummy. :D

      Mmmm... now I want one... lol

      Delete
  19. Just finished this circuit
    Thanks for the layout
    Getting some oscillation with the fuzz cranked up
    Used the 10 k sun dial
    Anything I can do to clean up the oscillation
    Thanks
    Ed

    ReplyDelete
  20. Tried a few different size caps between base and collector on Q 1, 100 pf did the trick
    Also used shielded cable for input and output for extra measure. No more oscillation , I still get a bit of minor noise and his when fuzz is cranked at the highest volume levels, I can live with this as the sweetest fuzz tones are not at full full.

    ReplyDelete
    Replies
    1. This comment has been removed by the author.

      Delete
  21. I have a sunface and would like to add an push/pull on/off volume pot. Is this a simple mod?

    ReplyDelete
  22. My try on that one ended with too little gain/fuzz... and high sounds that can cut heads. I'm sure I did something wrong as what I ended up with sounded nothing like the demo.

    ReplyDelete
  23. I may be asking on the wrong thread but if I had npn germaniums could I use them in this circuit or do I need to tweek the nkt275 circuit to make it non instead of pop; sorry if this is really obvious I'm new to all this!

    ReplyDelete
  24. That should read npn instead of pnp -I could murder predictive text on smartphones!

    ReplyDelete
  25. Well I suppose it makes more sense to use the silicon NPN layout so all the connections with polarity are correct. But I'd use the trimmer shown in the germanium version rather than this one which is multi turn. In fact I'm going to edit this layout to change the trimmer.

    ReplyDelete
  26. Cheers thanks for that, it'll have to be my next project (blue delay on the go at the mo), will let you know the result of how it sounds.

    ReplyDelete
  27. What's about to put in this circuite the BC108 not Bc108C?

    ReplyDelete
    Replies
    1. Yes it will work fine although the current gain in each won't be as high. Some people prefer lower gain transistors in the Fuzz Face anyway, so give it a go and see what you think.

      Delete
  28. Did anybody here experience any RF pickup or a whistle when the trimmer is set to 0k?
    It's funny but only after adding 5k in series did I get the whistle to stop.
    Why is that?

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

    ReplyDelete
  30. tried 3 times and got nothing

    ReplyDelete
  31. One transistor is aligned as it should C-B-E, with base in the middle. But the other one has E-C-B, Base at the bottom? Is that correct?

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

    ReplyDelete
    Replies
    1. the transistors in the layout are in the correct orientation. you may need to twist the pins of your transistors to put the correct legs into the correct places.

      Delete
  33. Hi, thank you very much for the layout! I built this, used AC126 transistors and get no sound at all. I looked at the NKT275 version and it is with a volt converter and the switched polarity for the capacitor. So do I have to do this one with a volt converter too?

    ReplyDelete
    Replies
    1. NKT275 transistors are PNP Germanium, as are the AC126. If you plug AC126's into the BC108 layout it will not work, as the polarity is reversed, BC108 transistors are NPN Silicon transistors. you will need to reverse the electrolytic capacitor, and reverse your power and grounds.

      Delete
    2. Ah great! I was thinking about that but since it is my first fuzz I didn't know about the difference between the two transistor types.
      Thank you very much!

      Delete
    3. any time man. for future reference there are two types of power, positive ground which use PNP transistors, and negative ground which use NPN transistors.

      Delete
  34. Trying to give this one a chance, but I'm not having much luck. The only change I made was to run a jumper from the bottom pin of the trimmer to the top right, next to the 1uf cap.

    Anyway, I tried the first time with BC109s since the place I was ordering from didn't have 108Cs. The hfes were ridiculously high. 592-613 hfe for 9/10 in the bag and the last being at 397hfe.

    I biased it using the site mentioned above and ended up having to use some pretty weird resistances. It came out sounding like garbage. Well, not entirely. When it picks up my guitar signal, it sounds pretty mean, but then it sputters out completely and I have to pick rather hard to get a sound. I can't lower my guitar's volume to get a clean signal at all.

    I built it a second time, this time with 2N3904s with hfes of 349 (Q1) and 354 (Q2). It sounded better, but still sputtered out when the volume hitting it got too low. I tried a buffer before the fuzz and the signal got better, but not by much.

    I'm thinking the hfe's are still too high, especially in Q1. I went ahead and ordered ~10 2n2369 (supposedly w/ a range of 70-110) and ~10 BC182K (supposedly w/ a range of 120-150) So I'm hoping I'll get it sounding right using the 2n2369 at Q1 and probably the BC182K in Q2. Just gotta wait for them to come in.

    Just in case I'm being an idiot, running a jumper from the bottom pin of the trimmer to the one next to the 1uf cap wouldn't screw up my signal, right? I'm assuming it's the 'Michael Fuller' Mod and running the jumper just effectively omits it.

    ReplyDelete
    Replies
    1. Solved! Ended up being a problem with the guitar (Disc on the 5-way wasn't aligned properly and my guitar wasn't outputting it's usual amount).

      Ended up switching the trannies out anyway to 98hfe on Q1 (2n2369) and 151hfe at Q2 (2n2219). Sounds absolutely fantastic!

      When it's biased it has a bass-y sort of mid-2000s sound to it ala maybe Dead Meadows and is dead silent when I'm not playing. Great for riffing!

      When it's mis-biased it sounds like a more lively version of the Roger Mayer Axis Fuzz, hinting at some Hendrix live in 1970 type stuff (read: broken amp and fuzz sound). A little bit of white noise and picks up radio when the volume is at 0, but still sounds great.

      Not sure which sound I like better. I might eventually rebuild it with trims for volume and fuzz and just have a gigantic bias knob and sort of tune it in so all the way CW is biased and all the way CCW is mis-biased to where I like it.

      I'm having fun putting my Axis fuzz first in line to use as cleans/slightly overdriven and then using the Sunface for lead tones.

      -----

      For future reference, I found the easiest arrangement was to use a B5K still as the bias knob but switch out the 2.2k resistor with something like 3.3k. Socket the 33k resister and trannies and then use the calculator mentioned up higher to find R1's value. I also liked a B2K for the gain knob in place of the B1K.

      Delete
  35. I cannot obtain the unity gain with this layout, both NKT and BC version.

    Even though the volume is in the max position, the effect volume is much less than unity gain.

    Any advice?

    ReplyDelete
  36. New builder here. It says to solder Volume 2 to Output, but I don't see output listed on the board. Where do I solder it??

    ReplyDelete
  37. New builder here. It says to solder Volume 2 to Output, but I don't see output listed on the board. Where do I solder it??

    ReplyDelete
  38. You solder a wire from volume 2 to the output connection on the stomp switch. Check the offboard wiring tab at the top of the page here to see the switch wiring

    ReplyDelete
  39. Hi , how would I make the SUNFACE BC108 using Axial Components what changes do I need to make on the Vero board.

    ReplyDelete