Thursday, December 1, 2022

TONE BENDER MKI: Component Values Table

Like most people, I've never even seen an original MKI in the flesh, so I'm relying on photos from the internet.  The references below all relate to original MKIs - or pedals that are said to be direct replicas of originals built for Sola Sound, so hopefully the component values are the same.

Please feel free to comment and point me in the right direction if you think there's something I should correct or add - ideally with a link to a reference somewhere.  Unfortunately, there are only a handful of MKIs left in existence, so new reference material is difficult to come by.  

GENERIC MKI TONE BENDER SCHEMATIC

This is the basic schematic for all MKI Tone Benders, with or without R5 & R11and C3 is sometimes electrolytic.  Yes, I know the pots were 47k and 470k, but hey... 

GENERIC MKI TONE BENDER SCHEMATIC

TONE BENDER MKI COMPONENT TABLE

Link to table in google sheets (probably easier to view by opening the link)


TONE BENDER CIRCUIT VARIATIONS

I'm certain that anyone that has found their way to my blog has seen the Dave Main schematics that are the gospel for Tone Benders.  As it happens, there seem to be some other values floating around out there from original MKIs that are not well documented - hence this post.

Apart from a number of small variations that are to be found, I believe that there are three basic types of MKI circuits that roughly match different production periods - for want of better names and to match the existing schematics out there, they are the;
  • Wooden box
  • Gary Hurst
  • Sola Sounds 
There are variations even across these three - despite the variations, they can still be grouped together largely based on capacitor values and production methods.  

The Wooden Box

The wooden box is the earliest of all Tone Benders - and of course very few still exist.  This is what I think sets them apart from others;
  1. Constructed on vero board & gooped  
  2. 3 x OC transistors - exact type unknown, thought to be OC75 or similar
  3. 22n input cap, 22n output cap and 2 x 25u coupling caps
  4. 33k resistor across Attack pot
The wooden box circuit is the basis for a lot of Stu Castledine's MKIs, as he happens to own one that was rehoused in a metal enclosure at some stage.  The information above is based entirely on Stu's MKI builds - there are some minor variations between builds, so take this as general guidance only.

The 22n input cap and the two 25u coupling caps let a lot of additional low-end signal through at the start of the chain, which is then reduced by the smaller output cap.


Gary Hurst

This is the next evolution from the wooden box when production moved into the metal enclosures for perhaps obvious reasons.  
  1. Point to point on phenolic board
  2. 1 x Mullard OC75 & 2 x Texas Instruments 2G381 transistors
  3. 10n input cap, 2 x 25u coupling caps and 100n output
  4. 33k resistor across the 50k Attack pot
  5. 47k resistor across the 500k Level pot
The two large coupling caps allow a bit more bass through - this is certainly noticeable compared to the next version which is the Sola Sound.


Sola Sound(s)

The last of the line - this is when Sola Sounds appears on the front of the metal enclosure.  Very similar to the previous, but with a few changes to biasing and coupling caps.

  • Point to point on phenolic board
  • 1 x Mullard OC75 & 2 x Texas Instruments 2G381 transistors
  • 10n input cap,  25u & 100n coupling caps and 100n output
  • Usually no additional resistor across Attack
  • No additional resistor across the Level pot

  • This version has a more focused mid-range, than the other two as a result of the 100n coupling cap (C3) - perhaps a little more aggressive sounding.


    COMPONENT VALUE VARIATIONS

    C1 - INPUT CAP

    All MKI Tone Benders have a 10n input cap right?  Apparently not.   Stu Castledine's very early version has 22n input and output caps.   Dave Main's Phoenix has a 68n input cap.   

    These changes both make a significant difference to the sound, as more low-end is making to Q2 which is where a lot of the magic happens.


    C3 - COUPLING CAP TO Q3

    Values spotted:  50n, 100n, 25u & 40u

    HPF
    40u     8k2 = 0hz
    25u     8k2 = 1hz

    100n   8k2 = 194hz
    50n     8k2 = 388hz (maybe)

    NOTE: This is just the RC filter calculation - when a leaky transistor is added, these values will increase by varying degrees.   Check out the detailed calculations here

    It's the change from a large electrolytic cap to a 100n cap that really makes a difference.  The higher cut-off of the HPF before that last transistor has a noticeable effect on the tone.


    C4 / R10 - OUTPUT CAP / HPF

    C4 and R10 form yet another HPF - most MKIs have 100n here, with either a 47k or 56k resistor for R10.    

    HPF
    100n   56k = 28hz
    100n   47k = 34hz

    47n     56k = 60hz

    22n     47k = 154hz
    22n     33k = 219hz

    The 22n output caps are only found on the wooden box version, which is paired with larger input and coupling caps - the HPF is removing some excess bass.  It does not sound thinner by any measure.
     

    TUNING NOTES:

    These are my notes from breadboarding and spending time testing a lot of transistors + a bit of reading on the internets.  Your results may vary.

    Q1
    Just a buffer really - there are minor differences with any transistor that will work in this position.  Need some leakage here.  OC75s usually leak like a sieves.

    Q2
    Spend time getting Q2 right - it's the most important transistor in the circuit
    Full attack - voltage will drop from close to 9, down to about 5 (in very round figures).

    Q3
    Affects the general character of the fuzz, including high frequency content.  Lower gain and leakage reduces fizz and scatty note decay.  Despite Q2 doing most of the work, Q3 is still pretty important.   

    BIAS RESISTOR R3
    Anywhere between 180k - 470k have been used.  

    ATTACK POT
    Not unlike the bias resistor - if you have to tune it to taste with R5, there's nothing wrong with that.    Smaller bias resistor (R3) can mean smaller pot value.  Log pots are recommended by some.

    Parallel resistance for the Wooden Box & Gary Hurst versions
    50k pot +  33k resistor =  approx 20k

    The Sola Sounds version does not have the additional resistor across the 50k Attack pot.

    Do not expect a nice even sweep of fuzz across the range.  It can happen, and it's nice when it does, but sometimes it doesn’t work out like this (log pots often fix this).  Note the video below for a demonstration of this, using original Tone Benders.    

    SYSTEM MKI TUNING TIPS FROM SEEKER ELECTRIC EFFECTS

    So nice that this was shared



    REFERENCE VIDEO BY GRAHAM GREEN - ORIGINAL TONE BENDERS





    ADDITIONAL REFERENCES:  RECENT BUILDS FOR MACARI'S (LATE 2022)

    STUART CASTLEDINE OF CASTLEDINE ELECTRONICS




    IAN SHERWIN OF GHOST EFFECTS - GHOST V2



    DAVID MAIN OF D*A*M - THE PHOENIX



    STEVE WILLIAMS OF PIGDOG - SYSTEM 2




    And if you want more, check out https://www.youtube.com/@pinstripedclips

    50 comments:

    1. Woooooow! This is even better than the previous one! I love it! If I catch something or see any possible update I will tell you, but it looks awesome.

      ReplyDelete
    2. Thank you so much for all your effort. This post is a gold mine. How can I help to increase this table?

      I am following Blackstrap on Instagram, and as he has two original Zonk Machines (Normal & Hybrid) maybe his pedals might be a good addition to this table, too?

      ReplyDelete
      Replies
      1. Hello. If they're originals with different values, then sure.

        The main reason for the component tables is for me to note what else is out there, and test variations on the breadboard with the transistors that I have access to.

        Delete
      2. I am not very good at deciphering, but have you seen this listing at reverb: https://reverb.com/item/32882137-original-1968-john-hornby-skewes-zonk-machine-fuzz-pedal

        Delete
    3. This blog is a goldmine. Thank you for sharing all of this!

      ReplyDelete
    4. Nice and useful add! I agree with your findings from the breadboard. I just found that encreasing the leakage on Q3 encreases the sustain and the noise. It's a circuit that needs a fine balancing.

      ReplyDelete
    5. Correct me if I'm wrong, but I think that R5 (33k) it's only in the Gary Hurst version (maybe in some others, but it is not there usually) This is one of the main difference with the Zonk that always has that resistor.

      ReplyDelete
    6. I just want to share something I found on FSB (which they got from the D*A*M Forum) that helps me a lot finding the right transistors very quick:

      TRANSISTOR SUGGESTIONS

      Q1= 60-80Hfe/100-200uA
      Q2= 60-80Hfe/100-150uA
      Q3= 60-70Hfe/<100uA

      Works for me everytime so far. Just built your Marshall SupaFuzz Prototype Layout and it directly worked.

      ReplyDelete
      Replies
      1. I think the biggest problem for people is using transistors with not enough leakage - if you don’t have enough, it just doesn’t work.

        Perhaps it’s a result if different measurement conditions, but most of the ones I like for Q1 / Q3 have had higher leakage - closer to 400

        Delete
      2. True. I find it surprising and interesting that so many different values work. For instance I mostly use soviet transistors and with a leakage closer to 400 they become quiet noisy in these circuits in my opinion.

        Delete
      3. I'll have to try some soviets to see what the differences are, as I've only used european and the occasional Japanese transistor

        Delete
    7. I usually use European/UK or Japanese transistors and in Q1 I used more than 400 with no problem if the gain is between 40-90. Where I see more difference is in Q2 gain (more gain, despite leakage, more intense fuzz) and in Q3, low leakage more gating, and with more leakage more sustain and noise.
      With stock resistors. This circuit is a lot of fun to tune and less bitchy than the Zonk with the RF sizzle that I solved with a capacitor from input to ground (from 470pf to 10nf)

      ReplyDelete
      Replies
      1. Have you tried a ferrite bead on the input? Suppresses very high frequencies.

        Delete
    8. Wow! That's a good idea! I have a lot of ferrite beads that I bought once to do some Spaceman clones and never thought more about them.
      In any case, I found very helpful the fuzz control from the Ghost effects version of the MKI (well, the retuned version wired as the MKIII fuzz control as you suggested) for this. At full I can tune a bit of sizzle but if you don't want to hear it, you can turn the fuzz control down a bit and the sizzle disappears and you still have a looooooot of fuzz to work with, et voilà!

      ReplyDelete
    9. Capacitors to ground on the input are good. Just thought the beads might be worth a try.

      ReplyDelete
    10. I just scrolled through reverb to find more pictures of Sola Sound MKI variants and found these two:
      https://reverb.com/item/64881912-sola-sound-d-a-m-mki-tone-bender-goldie-enclosure?show_sold=true
      https://reverb.com/item/65284574-2022-sola-sound-mki-goldie-gold-what-if-vintage-fuzz-pedal-by-macaris-dam-david-main-like-new-in-box-mkii-mki-rare-fx-limited-edition-of-12

      Apparently they both belog to the "What if" Tone Bender MKI collection. Did you know about these?

      I assume they made this version because of "Goldie": https://fuzzboxes.org/tonebendermki#goldie
      I am just wondering what kind of circuit this is.

      ReplyDelete
      Replies
      1. I think you're assumption is probably right. Those Reverb pics look like MKIs with an extra cap to ground on the input (can't see a value). I'll have more of a look when i have time.

        Delete
    11. And regarding the Stu Castledine Version, I tried to decipher the pictures in the D*A*M forum several times, and I think the 2017 and 2020 Version of Stu Castledine are the same. For instance it looks like he is using two 22n caps and I think I can make out the 3,3M resistor in this picture: https://stompboxes.co.uk/forum/download/file.php?id=27753&sid=bf018a792b81887ae2c2e31411a63e14&mode=view

      ReplyDelete
      Replies
      1. I haven't added it yet, but I've also seen a Castledine from a few years back with 22n on the input and output, but it has a 100n for C3 instead of 25u (C1 22n, C2 25u, C3 100n, C4 22n). OC77 for transistors.

        So I'm not entirely sure which one is the same as his original - either way, both will be interesting.

        Delete
      2. I found another one of Castledine's Versions listed. This time its the 2022 Version, if I am not mistaken: Ebay: Sola Sound MK1 TONE BENDER (Stu V2)

        Haven't checked it thoroughly, but I can make out two 22n and the 3,3M resistor.

        Delete
      3. Thanks - I'd say you're right.

        He seems to use a few different transistor types, which is reassuring to see that the standard OC75 & 2G381 isn't the only recipe available to provide a solid MKI sound - especially given the cost of these now.

        Delete
    12. There seems to be another version I am currently scratching my head about:
      Early metal-bodied Tone Benders - Exhibit B

      Unfortunately there aren't more pictures from the inside (Early metal-bodied Tone Benders - Exhibit B - Inside). What makes me curious is, that apparently C3 is a polarized cap, but a smaller one than C2. If I am not mistaken, C2 in this Tone Bender is a 22uf (look at the 22uf in "Goldie" as a comparison), so C3 must be something smaller (10 or 1 uf perhaps?). C1 and C4 should be 10nf and 100nf.

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

        Delete
      2. Found more pictures, and now I am even more confused. Apparently C3 is 40 uf:
        Insides 1
        Insides 2
        Insides 3

        The pots are A50k & A500k and the transistors are OC75, SG381 (2x)

        Delete
      3. I know - there’s also a photo of the same one on the DAM forum. Looks like a 40uf to me.

        https://stompboxes.co.uk/forum/viewtopic.php?f=6&t=1100

        Delete
      4. Btw I think the version on the pictures is this one: https://youtu.be/YJFnV3Eqy_s?t=752

        Delete
      5. I think I'll build this one as the original seem to have a volume problem. I curious to know what it really sounds like.

        Delete
    13. I found a reverb listing for the DAM 2022 The Phoenix Tone Bender.
      From what I can make out R3 is 270k and R4 is 1,9k.

      ReplyDelete
      Replies
      1. https://reverb.com/item/69024454-sola-sound-mki-tone-bender-the-phoenix-by-d-a-m

        Delete
      2. What difference does it make to use 1,8k, 1,9k or 2,2k for R4?

        Delete
      3. And something else: I get the feeling that for the bias resistor R3 the lower the gain the lower the value of the resistor. Could this be right?

        Delete
      4. Hello. Thanks for sending that through.

        1k8 vs 1k9 will make no difference at all. I think it might be 1k8. All that resistor does is control the lowest point of the bias, where the pedal will either completely cut out, or struggle to work (which can sound interesting).

        Delete
    14. Found something else. Isn't this 2022 Stu Castledine Version missing the 33k resistor (R5). If so, in what way does it differ soundwise?

      https://reverb.com/de/item/69687764-original-sola-sound-mk1-tone-bender-stu-castledine-v2-2022-25-of-just-50-made-mint-boxed-candy-macaris-london?show_sold=true

      ReplyDelete
      Replies
      1. Good spotting. He’s either decided that it works better this way for the transistors used, or he might have a 25k pot instead of a 50k pot (I’ve done this before).

        MKIs can be so picky about transistors and bias, it’s hard to give a definitive answer as to how it affects the sound. It might just be a better sweep for the pot

        Delete
      2. Another change in this version. R10 is 47k instead of 33k. I guess this could sound quite different to the 2017 version.

        Delete
      3. A little bit, but not as much as one might think.

        22n 47k = 154hz
        22n 33k = 219hz

        Delete
      4. But what does that mean? More hz more bass at the end?

        Delete
      5. I see, thank you! Still fascinating because I always had the feeling that the 2017 version had more bass, but they also used a different Amp and Guitar in the promo videos, so comparison was always difficult.

        Delete
      6. It’s always hard to tell with online demos - they really useful, but it can be difficult to be sure what pedals really sound like.

        Delete
    15. Another fascinating post concerning "The Ashtray" (the Original "Phoenix") in the D*A*M Forum: https://stompboxes.co.uk/forum/viewtopic.php?p=375306#p375306

      Apparently Gary Hurst used two OC71 for Q1 & 2 and log pots for this one (I never have used linear pots in this circuit anyway).

      ReplyDelete
      Replies
      1. Great photos. The various versions and differences in construction really do give the impression that he was using anything he could get his hands on / whatever was in the parts draws to make these.

        Delete
      2. Which means, it is up to us to create new variations.
        Input cap: something between 10n and 68/80n
        C2/C3: variable
        R11: 33-56k

        How about a System-Ashtray? :-)

        Delete
      3. It’s certainly very open to interpretation. I think as long as it has the vibe and bones of a MKI, it’s a MKI.

        Delete
    16. What difference does it make to use two instead of one bias resistor like D*A*M did in his 2017 System build (470k and 330k)?
      https://rvb-img.reverb.com/image/upload/s--uR3wnlVY--/f_auto,t_supersize/v1678676089/ifgcid2lwcvyolyaf2i9.jpg

      https://reverb.com/de/item/66844509-solasound-tonebender-mki-d-a-m-fuzz-dam?show_sold=true

      ReplyDelete
      Replies
      1. Correct me if I am wrong, so instead of having a single 220k you end up with roughly 200k by using the 470k/330k combination, right?

        Delete
      2. 330k & 470k = approx 194k, so close enough to 200k given the usual tolerances of old resistors.

        The general school of thought is that Gary Hurst started with a 470k resistor, and if Q2 didn't bias to his liking, a 330k resistor would be added to fix this.

        It's transistor dependant, but generally less gated, and there can be tonal changes. It can also impact the sweep of the attack pot.

        Delete
      3. So it'll sound different, when adding another resistor and leaving the old 470k in instead of directly using a single correct value (like something between 180k and 220k)?

        Delete
      4. Sorry, I think I didn't answer the question properly. Two resistor in parallel will sound identical to the same value resistor on its own. I was talking about the change in resistance in the previous answer.

        Delete