Debugging SuperFuzz

[Canucker]

Hopefully I did this correctly. Its my first attempt at collecting this info on a build. http://ustomp.com/?p=12 Thats the layout I used. The only change I made was put the input cap ahead of the 22k resistor and turned the input cap around (as advised with this layout in more then one place and shown in other layouts for the Superfuzz). I get sound out of the effect but its super quiet and clean....turned on the amps distortion and it still gave a clean bassy sound.I haven't yet hooked up a stomp switch so the "effect" is always on.

Thanks in advance for any help/advice.


Q1
C =7.35
B =6.75
E =0.356

Q2
C=7.99
B=7.35
E=5.98

Q3
C=7.10
B=6.53
E=0.513

Q4
C=5.99
B=7.74
E=0.009

Q5
C=6.01
B=7.65
E=0.009

Q6
C=3.576
B=1.012
E=0.29
.
.

D1
A (anode, the non-band end) = 0.527
K (cathode, the banded end) = 0

D2
A = 0
K = 0.527

[LucifersTrip]

here's from an original
http://www.diystompboxes.com/smfforum/index.php?topic=64068.0

        C        B        E
Q1  6.01    .66      .117
Q2  8.88   6.01     5.41
Q3  6.23   3.27     2.72
Q4  3.1     1.7      1.11
Q5  3.1     1.67     1.11
Q6  5.75   .98       .361

yours are far enough off that I would double check component values & pinouts first...and it really goes awry at Q4 & Q5

remember, e < b < c

[R.G.]

What he said, but more so.

1. What transistors did you use? Did you try to use the original Japanese type numbers? If so, they definitely have a different pinout. So do the metal can types recommended in the blurb and questions at the web site. The number one problem with new builds is incorrect part insertion, or changed pinout of substitute parts.
2. Q1 is dramatically off, and I'm surprised it passes any audio at all. For bipolar transistors, the transistor cannot be working as an amplifier if the base and emitter are not 0.4V to 0.8V apart. Period. That's the range over which the base-emitter junction diode is on. Your voltages show Q1 base at 6.75V, the emitter at 0.35V. That is not possible for a non-damaged NPN inserted in the circuit correctly if the other parts leading to it (resistors, etc.) are also correct. Q2 shows much the same problem. So do Q3, Q4, and Q5. Q6 ... might... be OK, but 0.7V is higher than I'd expect for a small signal device in a low-power circuit.

So something is quite wrong. What devices did you use, did you look up the datasheet for them and did you verify which pin (e, b, c) went into which hole?

As LT said, E, B, and C. What order?

[Canucker]

Thanks for the quick reply. I used 2n222 transistors. I used this for the pin out http://www.hamradioexpress.com/catalog/images/mp2222_small.gif The knobs don't make any noticeable difference to the sound. I started a second one to see if I have the same problem but I'm testing every part before I install it to ensure no incorrect values. I may have to pull a few parts out of this bad build though because I don't think I have enough of everything to build two. Thanks again for the assistance!

[R.G.]

The only change I made was put the input cap ahead of the 22k resistor and turned the input cap around (as advised with this layout in more then one place and shown in other layouts for the Superfuzz).

Turning the cap around was good, it's shown incorrectly in the layout.

But why did you put the input cap ahead of the 22K? That means that you're not getting the 1M pulldown outside of the cap, and the order of the 22K and cap should not make a difference in the signal getting into the input.

Q1
C =7.35
B =6.75
E =0.356

It is possible that the 1.8K resistor from emitter to ground is open, and your meter is completing enough of a circuit that you get a reading on the emitter. But when you connect to the base, this doesn't happen, so the base appears high. Turn power off, set the meter to ohms, and measure ohms from the emitter LEAD OF THE TRANSISTOR to ground. Use the lead on the transistor to catch a possible bad solder joint or trace.

Q2
C=7.99
B=7.35
E=5.98

Can't tell anything from here, other than the base is connected to Q1 collector, as it should be.

Q3
C=7.10
B=6.53
E=0.513

Q4
C=5.99
B=7.74
E=0.009

Q5
C=6.01
B=7.65
E=0.009

... and there's that high emitter again. Is it possible that the GROUND lead to Q1-Q6 is not connected to power ground? Try continuity/resistance readings between ground points to try to discover an open.

Q6
C=3.576
B=1.012
E=0.29

This one looks kinda right.

D1
A (anode, the non-band end) = 0.527
K (cathode, the banded end) = 0

D2
A = 0
K = 0.527

It's odd that one of the ends would be at half a volt, the other at 0V. There should be no DC flowing through there to make this happen. It's suggestive of a board trace flaw or soldering problem, consistent with the above suspicions about ground.

Try those tests, take two aspirin, and post back what you get.

[Canucker]

I noodled around with things...checking resistor values and changing some of the wires for the pots while simultaneously etching an enlarged board (to ensure no soldering errors)....no values appeared to be wrong.....but with changed wires I did manage to get a louder clean tone when I tested it out...the "balance" and "expander" both increased the clean volume while the bias did nothing...it wasn't quite as loud as the clean when I plugged my guitar direct into the amp but it was much closer then before....needless  to say I'm glad I etched a second board because I'm going to give that one a go before I hurt my brain more on this one...Thanks for the expert help! Even though I didn't make this work, you guys are an amazing support system! 

[Canucker]

I did my rebuild and got some sweet sounds with my three way switch in the up postion (connecting 1 and 3)....didn't have enough fresh transistors so I had to reuse one from the old build....it was suggested that I had damaged some transistors in the first build and the not so sweet sound coming out with my three way switch in the other position would back up the damaged transistor idea.....also discovered that the three way switch is in bad shape though it doesn't look it...you really have to push it hard into the working position. The 10k pot i was using was also in rough shape....so there was more then one thing wrong the first time around. In the working positing this thing kicks ass! It doesn't sound like an octave fuzz though as the site I took it from suggests....its just an awesome fuzz....which I super prefer!

[digi2t]

It doesn't sound like an octave fuzz though as the site I took it from suggests

The closer the trannys are matched, the more pronounced the octave effect. Especially when you use the neck pup, and dial down the tone.

You could add a trimmer in the octave section as well, and fine tune it if your trannys aren't balanced. Example;



With properly matched trannys, the octave can be more pronounced. Check Jimi's Superfuzz revisited video.

[Canucker]

Thanks for that bit of advice. I used all 2n2222 transistors and didn't test those...but I did test all the resistors and caps. Now I know what to avoid! Yet again I'll bitch bitch bitch about people uploading videos where they can't play any full chord rhythm patterns!!!!!!!!!!!! Single note stuff really doesn't give you much sense of how a distortion/fuzz/overdrive sounds when you really give it. Almost anything sounds decent with single notes....even an out of tune guitar sometimes!

[pinkjimiphoton]

bitch away.

fuzzes tend to sound almost identical when playing chords, where as when used in single note lines and doublestops, you can hear the TONE of the fuzz interact with the instrument.

i would suggest you don't watch any of my videos. you'll be dissappointed, cuz i never play louie louie over and over while twisting knobs.
best,
pink

ps pn2222's are way too hot to work well in a superfuzz circuit.

btw...full chord rhythm patterns? why? very rarely do we play "full chords"...ever...at least not us guys who do it for a living.

damn, now i'm bitching. lol. rock on dude.

[Canucker]

bitch away.

fuzzes tend to sound almost identical when playing chords, where as when used in single note lines and doublestops, you can hear the TONE of the fuzz interact with the instrument.

i would suggest you don't watch any of my videos. you'll be dissappointed, cuz i never play louie louie over and over while twisting knobs.
best,
pink

ps pn2222's are way too hot to work well in a superfuzz circuit.

btw...full chord rhythm patterns? why? very rarely do we play "full chords"...ever...at least not us guys who do it for a living.

damn, now i'm bitching. lol. rock on dude.

Made you bitch made you bitch! If fuzzes tend to sound the same when playing chords why would it matter what transistors I use?

[Arcane Analog]

Most Super Fuzz variants I have looked into use transistors with relatively low (for silicon) HFE. Anything over 300 is too much for my tastes. I would personally aim for mid 100s. Experiment with a few low gainers and see what you prefer. I do not think the 2N2222s were very high gain but measure them and experiment. I would also recommend matching the HFE for the pair of transistors used for the octave up - as I believe pinkjimiphoton mentioned in his demo.

[pinkjimiphoton]

bitch away.

fuzzes tend to sound almost identical when playing chords, where as when used in single note lines and doublestops, you can hear the TONE of the fuzz interact with the instrument.

i would suggest you don't watch any of my videos. you'll be dissappointed, cuz i never play louie louie over and over while twisting knobs.
best,
pink

ps pn2222's are way too hot to work well in a superfuzz circuit.

btw...full chord rhythm patterns? why? very rarely do we play "full chords"...ever...at least not us guys who do it for a living.

damn, now i'm bitching. lol. rock on dude.

Made you bitch made you bitch! If fuzzes tend to sound the same when playing chords why would it matter what transistors I use?

they sound different playing power chords...which i almost always do in video demos. add in 3rd as in a "full six string chord" and suddenly they all sound like poo.

it matters because if you use gain too high, the only way you'll pass signal is using a jackhammer for a pic. transistors that are too hot will need biasing adjusted to get them to work, or they won't work at all.

in reality, it DOESN't matter what transistors you use, as long as the pinout is right...what matters most is the gain. and as dino said previously, matching them for gain in the octave section (as demonstrated in my video) makes a huge difference in the way the effect works and tracks.

btw...octave fuzzes and full chords? all ya get is a really crappy fuzz sound...you won't HEAR the octave. (tho on some, if you roll the volume way down on the guitar, you can get an interesting rhythm sound similar to some hendrix...not really useful for most people tho).

so..use the transistors you prefer, we all do. but...if searching for "the sound" of the original, your best bet is to get something original, or something close to the specs of the original.

hit tayda or something, and compare a bunch of transistors there (they're cheap there) ...look at the data sheets,  for a charachteristic called Hfe....that's the gain. generally with fuzzes,  you want it around 100. some designs can use higher gain (often real noisy) but the classics have definite sweet spots usually between about 60 and 180 or so.

since this is silicon, it's easy...get a cheap meter with a transistor checker, and pop 'em in and see what they read. find a bunch that are close, and a pair that are REAL close...those are the ones you want in the octave part. the closer, the better.

sockets can help, but my advice is when you find what you like, take out the socket and solder the transistors in...this prevents failure on the road, them sockets don't hold the tranny very well once they've been used a couple time.

rock on dude.

[digi2t]

Most Super Fuzz variants I have looked into use transistors with relatively low (for silicon) HFE. Anything over 300 is too much for my tastes. I would personally aim for mid 100s. Experiment with a few low gainers and see what you prefer. I do not think the 2N2222s were very high gain but measure them and experiment. I would also recommend matching the HFE for the pair of transistors used for the octave up - as I believe pinkjimiphoton mentioned in his demo.

The original Superfuzz reputedly used 2SC828Q's. The "Q" is the lowest gain range for that tranny, but impossible to find. What I gave Jimi are "R", which are the next step up in gain, but by testing them I managed to assemble a set that crosses into the mid "Q" gain territory. If I remember correctly, the hottest one in there is around 180 or 190.

[Arcane Analog]

Most Super Fuzz variants I have looked into use transistors with relatively low (for silicon) HFE. Anything over 300 is too much for my tastes. I would personally aim for mid 100s. Experiment with a few low gainers and see what you prefer. I do not think the 2N2222s were very high gain but measure them and experiment. I would also recommend matching the HFE for the pair of transistors used for the octave up - as I believe pinkjimiphoton mentioned in his demo.

The original Superfuzz reputedly used 2SC828Q's. The "Q" is the lowest gain range for that tranny, but impossible to find. What I gave Jimi are "R", which are the next step up in gain, but by testing them I managed to assemble a set that crosses into the mid "Q" gain territory. If I remember correctly, the hottest one in there is around 180 or 190.

Those HFE ranges sound pretty good to me. Really, anything sub ~250 would pobably sound great. I personally like the sound of transistors ~150 HFE and lower. My experience is such that when using high gain devices biasing/tuning the circuit with something close to the original values becomes much more difficult if not impossible.

[Canucker]

The reason I ended up using the 2n2222's was because someone in the comment section mentioned them on the page I got my layout. http://ustomp.com/?p=12 I had one or two on hand so I knew I could get more at the place I shop. I used my meter to check the hfe of other transistors I have on hand and I have a number of 2n3904's that rate between 125 and 150. How "matched" do they have to be to be considered a match? I have two different brands which I guess accounts for the variation.

[Arcane Analog]

The reason I ended up using the 2n2222's was because someone in the comment section mentioned them on the page I got my layout. http://ustomp.com/?p=12 I had one or two on hand so I knew I could get more at the place I shop. I used my meter to check the hfe of other transistors I have on hand and I have a number of 2n3904's that rate between 125 and 150. How "matched" do they have to be to be considered a match? I have two different brands which I guess accounts for the variation.

Matched would be as close to the same as possible with the transistors you have on hand. That said, the layout you posted also has a control to adjust the bias of Q4/5 which help to dial in the octave. That will help to further 'match' the transistors.

For fun, try your lowest and highest HFE transistor in Q6 and see what you prefer.

[Canucker]

They also mention the 2n2222's here (2n2222a actually)  http://mudhoney.net/vincent/pedal.htm and the alternative they list is 2N3391. Poking around the net I found that they have a high hfe rating. Its a Mudhoney page and I guess you could say they're my inspiration for building this pedal. Maybe this weekend I'll go socket crazy and test everything I have on hand.

[Arcane Analog]

The circuit does not need alot of HFE to get thick. I found with too much it becomes far too mushy, loses all definition and simply sounds bad. Check the datasheets for the original transistors. Even with low gain devices you should be able to get a huge sound.

Socketing is a definitely a good idea although I prefer a breadboard before commiting to solder. I avoid sockets in builds as much as possible as they are a risk for transistors falling out over time.