Bubble Font Green Russian Muff with mil spec resistors? (gut shots)

[claytushaywood]

I recently fixed my buddies russian muff.  It is a tall font (or bubble font maybe) green russian muff, and it is one of the coolest muffs ive ever heard.  I absolutely love this one with my setup... I NEED to clone it.

Now i'm a firm believer in component compostion selection when it comes to fuzz.  and this one is interesting.

maybe you all can help me out?  Are those metal film resistors with the color bands?  and the green and red ones have the exact value listed on em instead of color coded bands (and they are actually very tight tolerance wise) i've only seen the number labelled on mil spec resistors... the solid

are the caps MLCC?  or what?

and anyone have an idea if the transistors are still sourceable for these muffs?  maybe not... all they say is "E H" on the flat side and they have a little triangle on the corner.

Id like to get as close as possible to this guy so any help or any other tricks would be appreciated.  I read that the guy building the wren and cuff tall font clone said one of the transistors biasing resistors ended up being a different value than on the schematic.

Any known quality schematics and/or experience is appreciated!  I had never played a green russian until the other day and I really cannot stop!!! so much fun!  and ive played and built a LOT of muffs!  who else is a fan?

[claytushaywood]

Here's a pic of the components!

[claytushaywood]

try again!

[culturejam]

Quite a mix of parts there. I assume Mike was doing what he always did and just go for whatever was most available and least costly. 

It's hard to say if the ones with the values printed on them are mil-spec. Some of the values are less common for carbon film 5% (20K, etc), but they are available. The rest of them are 5%, as they have four color bands. So they are carbon film almost certainly.

The caps look to all be multilayer ceramic.

[BillyBoy]

I've built several Sovtek Muff clones – Red Army OD / early Civil War, later Civil War / Tall Font, and Bubble Font.  I've also "upgraded" some Black Russians to Civil War / Tall Font.  If you want to stay true to their technology, it's a little tough to determine the type of those USSR components.  Maybe more carbon film than carbon comp and metal film resistors, although I think all were used.  Maybe more ceramic caps than film, although both were used.  Electrolytics seem to have only been used for the power filter cap.  I've built with metal film resistors and film caps (or silver mica) to cut down the "bad noise" for high-gain playing.  I've built with carbon comp and ceramic for more "good noise".  And I've built with carbon film and ceramic.  And other combinations.  They all sound OK but different and seem to me to impact noise, possibly harmonics, more than base sound.  The carbon comps and ceramic might suit me more, but it might depend on what I was trying to do. 

For Tall Font, Civil War, or Bubble Font, use KT3102E or KT3102EM transistors.  E is the metal can version, EM is plastic.  I can't tell the difference.  That's what was in the originals.  You can find them on ebay.  HFE is usually in the 400's, sometimes a little higher.  Doesn't seem to matter much.  I expect you can get there with 2N5088 or 2N5089s, too.  But you might want to stay in the 400's, maybe 500's.  You can use 1N914 diodes, or go for KD521A or KD521V for Tall Font and Civil War or KD522B for Bubble Font.  All available on ebay.  Again, doesn't seem to make a huge difference which you choose.

The main, and maybe only, significant difference between the Red Army OD / early Civil War, later Civil War / Tall Font, and Bubble Font was the values of three caps.  Red Army OD / early Civil War used 430pF silver mica (big chunky square brown ones).  Later Civil War/Tall Font used 500pF, ceramic or film (each was actually a pair of 1000pF because 500pF was apparently unavailable).  The Bubble Font used 470pF, ceramic or film. 

The difference in those cap values is small, within the tolerance ranges of each other.  So if you just randomly pick some of each value, especially ceramic, you might not hear any difference, or not what you expect.  Caps rated at 430pf and 470pf could both be 450pf with only about 5% variation.  And 10% off on 500pf could also get you 450pf.  Ceramics can vary by 20% or more.  I think that is why people prefer particular Sovtek Big Muffs over others – they got the magical cap values for those 3 caps in some builds, along with magic values on other components within their tolerances.  Even if you get the exact values for the caps, the sound is different, but not wildly different, between the three versions.  Some builders go up to 560pf on Civil War / Tall Font builds, to smooth them out some more.  But I don't think they were ever built that way on purpose (although maybe within tolerances).  560pf sounds a little too bassy for me.  The lower values (430pf) are a little ruder, grittier, and have more highs (aggressive) and the higher values (500pf) are smoother with less highs (creamier).  Bubble Fonts at 470pf are in between.  I have a couple of them that sound great, but I've gravitated toward the 500pF caps while I was off chasing my approximation of Gilmour's tone.  If you want to control the results, you have to buy a bunch of caps and sort through them to get actual values that match the version you want.  Note that a lot of DMMs don't have the accuracy to really sort caps in the pF range.  I had to get a decent LCR meter to do that.  Check your DMM specs before you waste some time and get disappointing results.  But then, maybe your Bubble Font values are really in the Tall Font range or vice versa.  You wouldn't know unless you take that pedal apart and measure them.  Don't do that!  Worst case – socket your caps and play around until you get a match with what you have.  To start, maybe pick sets of caps at around 450pF and 520pF and see which sounds better.  Then start tweaking from that value.  You should hit your target somewhere in that range, or maybe just a little higher or lower.  It is tedious, but if you stay methodical, you should hit it eventually.  Of course there is still the variation in technologies and tolerances on all the other components, and aging, ...  And I don't know if USSR manufacturing kept parts within spec tolerances reliably.  They also had a couple different versions of the tone stack they used, but you can determine which you have by looking at the component values.

You can tell if yours is a Bubble Font or Tall Font by looking at the graphics on the front.  Or you can look at those 3 caps.  If you have 3 pairs of 1000pf, you have a Tall Font or later Civil War.  If you have (3) 470pF, it is a Bubble Font.  It is definitely not a Red Army OD or early Civil War.  The caps in those are about the size of a piece of Bazooka Joe bubble gum, and brown. 

[BillyBoy]

I just realized I could get a bigger version of your picture if I clicked on it again...

I don't know the resistor and capacitor types for sure.  The larger green resistors might be carbon comp or metal film.  If I had to pick, I'd guess metal film.  The other solid color resistors might be metal film.  The ones with bands are probably carbon film.  I'd guess ceramic for the caps.  Should be 23 resistors:  (1) 1.5K, (1) 39K, (4) 100K, (1) 2.7K, (1) 22K, (3) 10K, (4) 470K, (3) 390R, (3) 12K, (1) 1K, and (1) that is either 20K or 39K (I see a 20K).

Pots should all be 100K, probably linear, but I usually put an audio on the volume.

KT3102EM transistors.  Diodes look like KD521A.

I can't see cap values, but just by counting them, it looks like a Bubble Font.  I think it would need two more caps to be a Tall Font or late Civil War.  If that's the case, then your 13 caps should be:  (3) 470p, (6) 0.1u, (2) 0.047u, (1) 0.01u, and (1) 0.0039u.  Plus (1) electro that is probably 22u, maybe 20u, for power filter.

That might help you figure out which is which if you can't see all the markings.  Luckily, most of yours are turned so that you can see them.

If you are building a Mudbunny, it looks like that would be the "Green Russian" version in the build doc, with some different/better power filtering/protection and a pulldown resistor added. 

[juansolo]

I wouldn't get too hung up on the construction to be honest. Though the caps look like film (a couple of the smaller ones may be ceramic). What would be handy is that pic with all the values on it. It'd be easy then to turn it into a schematic and clone it.

Transistors you'll have to compromise/guess on regardless unless you fancy pulling and metering them. So being ultra-anal on the cloneage I find to be a bit of a waste of time.

[GrindCustoms]

I wouldn't get too hung up on the construction to be honest. Though the caps look like film (a couple of the smaller ones may be ceramic). What would be handy is that pic with all the values on it. It'd be easy then to turn it into a schematic and clone it.

Transistors you'll have to compromise/guess on regardless unless you fancy pulling and metering them. So being ultra-anal on the cloneage I find to be a bit of a waste of time.

+1 on that...

You'd take to Rusky Muff with serial numbers that follows each other and they will likely sound different... ...once tolerances piles up... who knows what will be the character of the said unit.

[BillyBoy]

Another +1 on that.  That's what I found in my builds and mods.  Technology, brand, etc matters less than you might expect in this circuit. 

Best and shortest path to happiness is to talk your friend out of that one!!!

To get close with a build:

   follow the schematic and stated values – this on its own gets you into the Sovtek tones
   use transistors with HFE in the 400's, maybe 500's, 2N5088 is a decent choice, or 2N5089 for a little less noise, others should work OK, too
   use metal film resistors and film caps
   use ceramic caps for the 3 key caps, and screen their values carefully for the version you like (eg 470pf for BF).  They represent 100% of the intentional/designed tone difference between RA OD, CW, TF, and BF versions.  Maybe socket these on a Mudbunny build.
   source KT3102E/EM and the USSR diodes if you want, but it isn't as critical as some seem to believe.  If it was, then they would all sound more the same, but they do vary.  It might improve your odds a little bit.  Or at least make you feel better about it and remove nagging doubts about your build...

It's nearly impossible to control the tolerance slop with that many components having tolerances from 1 to 20%, so don't expect to get a 100% match.  You should get 85% goodness, maybe closer, if you are lucky.  Since the Sovteks consistently stuck to the schematic and stated values (unlike NYC builds), the resulting range of tones is a bit narrower.  You could even say that the 3 versions actually have a characteristic sound, even with the tolerance slop.  I've ended up with pretty good Sovtek clones by doing that.  Mine are close enough that I rarely play my expensive vintage units – mostly just use them as a comparison.  But I did have to be careful with measuring and selecting those 3 caps since their tolerance ranges can cause their values to overlap.  I try to stay within about 5p of stated values on those to get consistency. 

If you want to try to tighten up your match a little bit, then also get close on the 47n and the 100n cap values.  I just use tight tolerance film for those and don't worry about screening them.  You can also get film caps in the 430 - 560pf range for the 3 key caps.

Best bet for build happiness... Listen carefully to that pedal and decide what about it makes it appeal to you more than other Sovteks.  Aggressive, more highs, grittier, in-your-face is the low end of the range for those 3 caps.  Smoother, creamier, less highs, more refined and gentlemanly fuzz is at the high end of the range.  Take a guess at a starting point like I described above and then (if necessary) tweak them together in about 10-20pf increments until you get a similar sound.  Or in this case start with 470pf and go from there.  Then don't think about it anymore, and happily play your killer hand-tuned build and think what a bargain it was compared to buying 20 vintage Sovteks to find another perfect one! 

[claytushaywood]

I appreciate all the response.  I do understand the importance of tolerance in components, especially in these ceramic caps.  Ive found most all components I get are consistently tighter in tolerance than they're listed... except ceramic caps, even multilayer ceramics.  Ive never had a multimeter that could measure capacitance using the test leads (that function has never worked on my DMM- just a broken function on it from the jump) But is measuring capacitance in circuit not doable for the most part?  Itd be worth it if i could measure half the capacitors in circuit just to get an idea of how far off the values are from labels. 

With my muff builds I always take care in selecting the 3 400-500pf caps.  I always socket and tune them to the transistors- which is probably the most useful thing you can do with any muff build for real.  I am also one of the ones in the camp that component composition does make a difference in fuzz builds.  There is something going on that maybe isnt measurable.  and it definitely is a cumulative approach.  Like using a couple metallized film caps isnt gonna change the way a circuit sounds, but going from all metallized film caps to all ceramics will make a difference.  Maybe i think small value/tolerance changes dont matter as much because i'm not auditioning them as a cumulative effect.  This is of course beyond transistors and the 3 key cap values. 

On the transistors, the ones in this pedal say E H on them- i just double checked to make sure i wasnt seeing E M.  Was there an EM version of these transistors?  cant find anything on it. 

And on the cap values- most all of them say "u10" and 2 say 47n, 3 say n47 (470pf i suppose) 1 says 3n9, and 1 says 10n

Interestingly, with the diodes.  in each of the two pairs- one side read ~.570 and the other read ~670.  Would that be worth taking into consideration?  after I remove these transistors and measure!

[BillyBoy]

Sounds like you're on top of it. 

From what I can tell, the KT3102E (metal can packaging) was used mostly in the RA OD and early CW types.  The KT3102EM is the same transistor in plastic packaging, and seems to have been used in CW, TF, and BF.  At some point, perhaps in the later BF models, definitely in the Black Russians, they started using other transistor types.

The letter after the 3102 denotes different specs of the 3102, just like normal.  All but the E and EM versions of the 3102 usually have HFE values too low for this circuit.  They made the 3102E/EM for a number of years, so you'll find them with slightly different markings, fonts, etc.  Often, the EM version was labeled as E, since you could tell it was in plastic.  The Russian alphabet is different, but also similar in some ways.  Some of their letters look like our letters, but aren't the same.  It can be confusing if you get into it.  Luckily, the E/EM versions have letters that look like an E.  So if it is labeled 3102 you can probably safely assume it is an E (or EM if plastic), since those are the only 3102's with consistently high enough HFE.  I think that the 3102E range of slightly lower HFE contributes to the overall Sovtek tone.  Lots of other transistors used in Muffs had considerably higher HFE values.  The attached file shows some of the spec info.  Note that the specs for the versions ending with "M" are the same as the specs for the ones without the "M" (eg EM and E) since the M is just a packaging modifier.

The ones with weird markings, like colored dots, may have appeared in some late BFs, but are definitely present in Black Russians, along with transistors clearly marked as other types (eg BC547C, BC549C).  I don't know if those weird ones are 3102s or not.  I think by then that EHX was getting back on its feet and headed back to the USA.  Maybe they were already starting to make changes.  They did that all the time in NYC.  The first post-USSR version used 5088 and 5089 transistors, amongst a number of other changes.

I guess I was splitting hairs on the technology not mattering that much and muddied my point a bit.  I mostly think it doesn't matter that much to what I call the "base tone" in a Sovtek build, specifically.  I can get that with different types of parts.  But it does matter for noise/clarity, maybe harmonics, or maybe some other subtlety that is difficult to identify.  If you change nothing but the noise level in a pedal with that much gain, it can make a noticeable change in the overall sound.  The base tone can then come through cleaner or clearer or tighter or something, whatever the right adjectives are.  I'm hopeless at describing sounds.

I'm not sure I could tell the difference between all metal film and all carbon film.  But all carbon comp makes a difference I can hear (and like).  That sounds good in Triangle and Rams Head builds, too.  I could probably tell the difference when I use all film or all ceramic caps.  But since I usually end up with a mix of film and ceramic and I'm picking values carefully, it's not so noticeable when only one or two get changed.  I've never bothered measuring the diode voltages since the few diode types I've tried have all worked OK with random picks.  If I listen very closely to the clipping in the tone, maybe I could hear the impact of the diode voltage differences.  But with sequential clipping stages, I bet it would tend to even out the tolerance variations in the diodes.  It certainly matters in some other circuits, though.  Guess it just hasn't been an issue in this circuit for me.  I just measured some of my KD521As – they were all around 0.65v.  Since it is just as easy to build with one technology as another, I often stay as close to what I guess the originals to be, just for peace of mind if nothing else.  And also because I usually have some original components because I usually at least try that if I can. 

[nieradka]

Ive gotten many transistors sold as kt3102em from Ukraine that arrived and were the color dot ones (white/green?) Its probably just a different factory's marking. They test in the same ranges and sound the same to my ears as kt3102EMs marked with writing.

[BillyBoy]

That's good info to know.   

Were your KT3102 with dots all the same marking, or did different dots mean different HFE ranges or anything you could tell?

[nieradka]

All same markings 400-700 hfe (most ~500).

[JC103]

Time for a resurrection! How did your build turn out? I'm going down the Civil War path as we speak... Ceramic or Mica, hmm... choices.