Marshall - Supa Fuzz (196x) [gut shot picture thread]

[JVanDe7]

Just to recap, because I'd really like to get this right -

This is what I'm working with...
Q1- 70 hFe (leakage 31)
Q2- 88 hFe (leakage 62)
Q3- 121 hFe (leakage 116)

And I used this schematic... except I've omitted the .01uF cap to ground at input, everything else is just as shown.
Oh but I'm using a negative ground layout.



So I used the 8K2 to bias by changing it to a 20K trim. I had it set so Q3c was 4.5v and it was sounding pretty tasty in my opinion. But I do want to get the correct vintage Marshall Supa Fuzz sound so I tried to use Phibes' suggestion and put the 8K2 back in place and use a 100K trim instead of the 47K on Q2c. So when I tried to go that route I'm only getting 2v on Q3c no matter where the 100k trim is set. Why would this be?

[JVanDe7]

Alright. I'm a dumbass. I had the trimpot wired incorrectly. It's early, I haven't had my coffee, and I was in a hurry. That'll teach me...

I also read the MKII thread here at freestompboxes and I read a couple over at the DAM forums. Lots of good info. Thanks fellas. I think I should be good to go now. I'll give a shout once I have it all sorted out and sounding perfect.

[JVanDe7]

This is without a doubt, the most badass fuzz I've built. Period. I had a 16mm B50K pot laying around so I used that instead of the 47K resistor, for bias, and it works perfectly. I drilled and mounted it on the enclosure. Which, by the way, is a Zvex size enclosure from SmallBear. Tiny little bugger. http://www.smallbearelec.com/Detail.bok?no=960
They are listed as size B but I have a size B and this one is even smaller. Tight fit inside. That's why I used the smallest layout I could find (vero 11x11). But you know, with 3 big ass germanium transistors on the board, plus two 24mm pots and one 16mm pot... I'm a little bit proud that I was able to fit it all in there. It's a tiny pedal that packs a HUGE punch. The bias knob is super usable too. And it worked out so that the proper bias range is when the knob is set somewhere between 11 o'clock and 1 o'clock. It's nice to be able to fine tune the bias without opening up the enclosure.

THANK YOU to phibes and all of you guys! Seriously. Biasing via Q2's 47k resistor is definitely a better way to do it and I never would've known.

Here's a couple pics to show what I mean about the size difference. So glad I got it all to fit in there (that's what SHE said)

[phibes]

Glad you got it! Nice looking box too! Sometimes it's really easy to fuck up wiring the trimmers right.

Yeah the Q2 trimmer is definitely the way to go. I know Dave Main was the first to start implementing that as I didn't see it in any boxes before his, he knows Tone Benders so credit definitely goes his way. Really though, I don't know if I mentioned it in the other thread but there's a pattern that always got over looked. Go to Fuzzcentral for example (yes I know there not all 100 percent right) but if you look at the MKII's, Vox Tone Bender and Fuzz Face, you'll see that the 8K2 never changes but the other collector resistor varies from 10K-100K. Dave's hinted at that before. They we're obviously changing those for a reason.

[JVanDe7]

Got it labelled. Gonna get a smaller/different knob for bias though.

[LucifersTrip]

Stu was kind enough to provide measurements of his vintage units on the D*A*M forum: http://stompboxes.co.uk/forum/viewtopic ... 99#p110299

just breadboarded this with 3 X 2SB459's (hfe 186, 193, 213 respectively, leakage ~ 100 - 150uA)

With no changes I got a very good fuzz right off...no gating...no poor decay.
9.69v Danelectro supply.

EBC

Q1: 0, -.01, -9.63
Q2: 0, -.12, -.36
Q3: -.24, -.36, -7.70

not exactly what Stu got, but close...I'll screw around with it later. one note: I got the highest (-7.70v) Q3 C voltage with the 3 in those positions.
Q1C may be too high.

[LucifersTrip]

For the above setup (hfe 186, 193, 213 respectively) I adjusted Q2 C resistor. The sweet spot was at 34.5K which put Q3 C at -7.38v

Q3 C is good between ~ -6.7v to -7.8v. Below -6.7v is raspy and starts to gate (btw -4.5v = sh*t as expected). Above -7.8v started to thin out the tone, get weaker and I was getting more pick percussion.


Next: different transistors...

[JVanDe7]

I'm wondering if 4.5v sounded good to me because I am using much lower gain transistors? I know what good fuzz sounds like and it wasn't far off from what I am able to achieve now with the added bias control. My transistors have much lower gain and leakage than what I've seen about original units on the web. Or does that not matter or make any difference? Or maybe because my battery is only 8.8v?

[LucifersTrip]

I'm wondering if 4.5v sounded good to me because I am using much lower gain transistors? I know what good fuzz sounds like and it wasn't far off from what I am able to achieve now with the added bias control. My transistors have much lower gain and leakage than what I've seen about original units on the web. Or does that not matter or make any difference? Or maybe because my battery is only 8.8v?

Sorry...I just realized I reversed the results and wasn't really clear enough about the -4.5v

Below -6.7v it gets thinner and even thinner all the way down to around -2v, then starts to gate around -1.2v, then turns off right under that. At -4.5v it wouldn't have sounded bad if I didn't hear what it sounded like at -7.38v, but compared to that it was terrible...thin & trebly with no fat richness. Above around -7.8 it also starts to get weaker & thinner, then get clicky (pick percussion) when even higher. 100K took me to ~ -8.6 to -8.8v

I just spent another 3 hours experimenting with different gains & leakages.

The first thing I did was try 3 transistors in the 200 hfe range with higher leakages (200-300uf) and it was a flop. There was no hiss, but much weaker in all respects.

Then I placed the original 3 back in and subbed a higher leakage (300uf+) Q into each of the 3 spots one at a time. In Q1 it seemed to actually increase the gain. I got more feedback and a fatter, thicker sound, but lost tightness and there was an increase in pick percussion. This was a tough decision because chords sounded great with the added fat, but the pick percussion killed it. In Q2 it totally killed the sound and actually started gating. In Q3, there was not as much change as I expected, but not any better...if anything, a little worse.

...so you can get away with some leakage in Q1 & Q3

Then I placed the original 3 back in and subbed a lower gain one (hfe 120) Q into each of the 3 spots one at a time. In Q1, it was the worst. It was actually like turning down the fuzz 50%. In both Q2 & Q3 it simply made it much thinner and trebly losing the richness.

...and I hate to say this, but the last thing I did was try the FF +1 (hfe 80, 80, 120 respectively, all low leakage) and again, a flop. Compared to the 3 x 200, it sounded like I turned the fuzz down half way. I also had a way smaller margin to get anything useable when adjusting the resistor on Q2 C.
==================

Regarding your 8.8v...I'm pretty sure all your readings will be a bit lower. Try one thing just for the hell of it. Put a high gain into Q1 and see if it jacks it up. I think I'm almost done (amazingly with the 3 I picked initially), but I am going to experiment with even higher gains (low leakage) in the 3 positions before soldering.

[JVanDe7]

I don't have a ton of Ge trannies but I do have a few. I think I have one or two that are around 170 but the leakage is also kind of high (200+) if I remember correctly. I should try one of those in Q1 you think? I'm leaving for work now but I can give it a shot tonight after I get back home.

[LucifersTrip]

I don't have a ton of Ge trannies but I do have a few. I think I have one or two that are around 170 but the leakage is also kind of high (200+) if I remember correctly. I should try one of those in Q1 you think? I'm leaving for work now but I can give it a shot tonight after I get back home.

Yeah, why not? Later today, I'm going to put the 80, 80, 120 back in and see if I can supe it up.

[Electric Warrior]

I'm wondering if 4.5v sounded good to me because I am using much lower gain transistors? I know what good fuzz sounds like and it wasn't far off from what I am able to achieve now with the added bias control. My transistors have much lower gain and leakage than what I've seen about original units on the web. Or does that not matter or make any difference? Or maybe because my battery is only 8.8v?

Sorry...I just realized I reversed the results and wasn't really clear enough about the -4.5v

Below -6.7v it gets thinner and even thinner all the way down to around -2v, then starts to gate around -1.2v, then turns off right under that. At -4.5v it wouldn't have sounded bad if I didn't hear what it sounded like at -7.38v, but compared to that it was terrible...thin & trebly with no fat richness. Above around -7.8 it also starts to get weaker & thinner, then get clicky (pick percussion) when even higher. 100K took me to ~ -8.6 to -8.8v

I just spent another 3 hours experimenting with different gains & leakages.

The first thing I did was try 3 transistors in the 200 hfe range with higher leakages (200-300uf) and it was a flop. There was no hiss, but much weaker in all respects.

Then I placed the original 3 back in and subbed a higher leakage (300uf+) Q into each of the 3 spots one at a time. In Q1 it seemed to actually increase the gain. I got more feedback and a fatter, thicker sound, but lost tightness and there was an increase in pick percussion. This was a tough decision because chords sounded great with the added fat, but the pick percussion killed it. In Q2 it totally killed the sound and actually started gating. In Q3, there was not as much change as I expected, but not any better...if anything, a little worse.

...so you can get away with some leakage in Q1 & Q3

Then I placed the original 3 back in and subbed a lower gain one (hfe 120) Q into each of the 3 spots one at a time. In Q1, it was the worst. It was actually like turning down the fuzz 50%. In both Q2 & Q3 it simply made it much thinner and trebly losing the richness.

...and I hate to say this, but the last thing I did was try the FF +1 (hfe 80, 80, 120 respectively, all low leakage) and again, a flop. Compared to the 3 x 200, it sounded like I turned the fuzz down half way. I also had a way smaller margin to get anything useable when adjusting the resistor on Q2 C.
==================

Regarding your 8.8v...I'm pretty sure all your readings will be a bit lower. Try one thing just for the hell of it. Put a high gain into Q1 and see if it jacks it up. I think I'm almost done (amazingly with the 3 I picked initially), but I am going to experiment with even higher gains (low leakage) in the 3 positions before soldering.

High leakage for all three should work out well. Try different transistors with rather high leakage for Q2 to find the one that gives you just the right amount of gating. You want it to gate only subtly at room temperatures, so it doesn't get too bad when it's warmer.
I think a bit of gating is part of the character of the MKII and gives it a really cool attack.

Here's a a really wild sounding vintage example with lots of gating:



Picking transistors for an MKII sure ain't easy. I keep redoing mine

[JVanDe7]

High leakage for all three should work out well. Try different transistors with rather high leakage for Q2 to find the one that gives you just the right amount of gating. You want it to gate only subtly at room temperatures, so it doesn't get too bad when it's warmer.
I think a bit of gating is part of the character of the MKII and gives it a really cool attack.


This is why I love the added bias knob so much. I've had several fuzzes with internal trims for bias adjustment and that's a nice feature - you can open up your fuzz and maybe get out your multimeter and dial it in. BUT, there is a distinct advantage to being able to play your guitar while you adjust the bias, which is not so easy to do with a trim pot and screwdriver. I'm going to put a very small knob on that bias pot, so it won't get bumped/used very often. I really enjoy having that feature on this pedal. Just like you mentioned in the above quote, the amount of gating is important and the bias knob is great for dialing it in, or just for experimenting. I can get some Zvex Fuzz Factory esque velcro/gated sounds by fine tuning that bias knob. Usually my preferred setting is just before the gating starts getting noticeable.

[JVanDe7]

Oh, also, I'm thinking of maybe changing the two 10uF caps to 4.7uF and then putting the .01uF to ground cap (at input) back on the board (it's currently omitted). I'm just curious to see how that version of the Supa Fuzz sounds. I like the boomyness of the fuzz right now as is, but when I use the neck pickup it can get pretty extreme. But I guess that's just what the Supa Fuzz is all about? I still have my Soul Bender clone if I want more "normal" tone bender sounds...

[Electric Warrior]

Shouldn't be much different.
Left: 5µF, cap from input to ground, OC81D; Middle: 5µF, cap from input to ground, Impex S3-1T; right: 10µF, no cap from input to ground, OC75;

[LucifersTrip]

High leakage for all three should work out well. Try different transistors with rather high leakage for Q2 to find the one that gives you just the right amount of gating. You want it to gate only subtly at room temperatures, so it doesn't get too bad when it's warmer.
I think a bit of gating is part of the character of the MKII and gives it a really cool attack.

Picking transistors for an MKII sure ain't easy. I keep redoing mine

That's probably because many of the originals sounded very different because of ge inconsistency. Yes, the gating can sound cool, and it'd be better if you could dial it in, rather than have it permanent. Though, it's doubtful that the original designers wanted it gating full time.

I tried high leakage all over the place in this and in no instance did it actually make the fuzz any better...but yes, it can make it different, which is up to personal taste. As I wrote, I could get away with it in Q1 & Q3, but in Q2 it was a real fuzzkill.

The other problem with dialing in the gating was that it didn't anywhere gate near the best fuzz. The best fuzz for my current setup is at -7.38v, but to get it to gate, I'd have to get it down to ~ -2v or up to around -9v....or of course, throw a leaky Q2 in there and lose fuzz. I haven't found any combo yet that gives me a good fuzz near the gating (probably since gating is the result of severe misbiasing)

...but I'll try more...

I'm sure the best solution for MKII's is to just build a bunch of em. They're cool enough to make it worth it.

[JVanDe7]

Didn't get a chance to mess with the transistors like I planned. I got caught up building a Mosrite Fuzzrite instead.

But anyhow, just wanted to say, it's funny how jazzed I was about this Supa Fuzz when I first built it. I guess the honeymoon is over now because it still sounds good, but not as good as my Soul Bender clone. It could be the transistors though. I built the Soul Bender back in October 2008, and the germanium transistors (SmallBear) are Hitachi brand I think? I forget the hFe on them but man oh man it sounds so damn delicious. I didn't socket the transistors though (doh!), so that was foolish of me.

So yeah, I still dig this Supa Fuzz but not as much as I was a couple of days ago when I was raving about it in this thread. That's how it goes with new pedals though I guess...

[LucifersTrip]

I'm finally done (well, at least this build) and thanx to DAM's 200 hfe recommendation, I will barely have to make any change. It is absolutely killer with the 47K reduced to 36K...that's all.

I furthered my leakage tests with a much more controlled experiment. I substituted transistors with the same number (2SB459), almost same hfe, but different leakage...into each of the 3 slots. If you are going for straight out smooth saturated screaming fuzz, then low leakage won out in all cases. If you want a more lower gain, dirty, gated fuzz like in EW's youtube vid, then throw a higher leakage one into Q2. In Q1 a higher leakage reduced treble, sharpness, a little volume, and gave it a little fatter sound (not needed with 10uf input cap). In Q3, it reduced the overall attack.

In my previous version I had hfe 186, 193, 213 respectively ...but I upped the ante on the screaming fuzz by increasing Q1 further and keeping the leakage low. So, my final version is hfe 244, 193, 213 ...all leak under 150uF

If I want to get some gating, I can raise the C voltage on Q3 to a little lower that I initially thought...to around -8.3V. This way, you'll have a gate along with a rich, saturated fuzz. I can also get it to gate by lowering the voltage to around -2v, but the tone is way too thin and wimpy. As I said before, if you want the gate to by dirty (with not nearly as much rich saturated fuzz) simply increase the leakage of Q2.

I like monster fuzz, so I'm going with low leak hfe 244, 193, 213 and a trim/cutoff on Q3 collector to give me from -6v to -8.5v

i I was going to mod it, I'd just put in a switch to alternate between a high and low leakage Q2. That way I could have everything: rich saturated fuzz, a dirtier low gain tone and gating with both.

[JVanDe7]

Alright so I tried messing around with the transistors a bit. I don't have many spares so I couldn't get anything to sound better than it already did. So I just left it alone in the end.

I also converted from the 10uF version (with no .01uF cap to ground at the input) - to the 5uF version which DOES include the .01uF to ground at input. I like the results! Not a big difference but the 5uF version seems to have more treble and a bit less boomy/woofy-ness. The biggest difference is that I no longer pick up radio signals. Without that .01uF cap I was picking up crazy strong interference. So I used a poly film .01uF and now it's good to go. I might change the 50k bias pot into a 100k just to give it a bit more range, but other than that - this thing is rockin' like Dokken.

I think the 10uF version is the later version, right? And the 5uF version was the first/earlier version?

[LucifersTrip]

Alright so I tried messing around with the transistors a bit. I don't have many spares so I couldn't get anything to sound better than it already did. So I just left it alone in the end.

you tried a high gain in Q1?

I also converted from the 10uF version (with no .01uF cap to ground at the input) - to the 5uF version which DOES include the .01uF to ground at input. I like the results! Not a big difference but the 5uF version seems to have more treble and a bit less boomy/woofy-ness.

then I believe you're closer to a SOL TB MKII

The biggest difference is that I no longer pick up radio signals. Without that .01uF cap I was picking up crazy strong interference.

radio should go away after boxing

I think the 10uF version is the later version, right? And the 5uF version was the first/earlier version?

viewtopic.php?f=19&t=751&start=100#p179392