super-fuzz schematic differences or errors? second set of eyes welcome

[jessenator]

I'm having deja vu  : P  (swear I've asked this before)

So I've made more progress on my revised "Sardine Tin" (super-fuzz) PCB: caught a few errors, and was almost ready to get these made, and I had a minipanic when I started comparing board component bits. I don't know if these are alterations made by various makers, or just an easy flip flop mistake.

I'd appreciate a second set of eyes here.

So here's my schematic: https://www.dropbox.com/s/qp8mtjeofluf0ld/sardine-tin-rev-2.0-schematic.pdf —apologies in advance for the <X| bifurcation. I'm weird and want my schematics to fit on paper sizes, so this one is US Tabloid (17x11) size  : P

Here are the big questions I've got. "?" is for "appears on one schematic, not on another" and the others have specific callouts or diagrams:


Also, sorry for the crop, but they relate 1:1 with my Eagle schematic linked above.

The biggest one is that cap polarity... I think I got confused by the OG schemo and just went with bipolar caps, but I'd like to get the danged thing right this time around.

I've looked at the original schematic and a couple others drawn by other board makers to make the comparisons/cross-checking on the construction.

Anyhow, thanks for the help!

[mauman]

BTW, the link to your schematic is broken.

The polarity of C1 doesn't matter, it's isolated from anything except ground by C4 and C6.

R10 is a pulldown resistor to prevent popping when switching active/bypass, optional.  Anything from 1M to 2.2M will work if you use one.

R19 sets a minimum value for the Drive pot, so that the output level doesn't go to zero when you turn Drive all the way down. No effect otherwise.

That circuit near C8 could be a capacitance multiplier, if the Q emitter is supplying voltage to something downstream.  Those are usually seen with large cap values (1000 uF or more) to stand in for an even larger cap, using it with 1000p (is that the label?) would increase the effective capacitance but why not just use a larger cap? 

[jessenator]

BTW, the link to your schematic is broken.

blerg... fixed now

The polarity of C1 doesn't matter, it's isolated from anything except ground by C4 and C6.

R10 is a pulldown resistor to prevent popping when switching active/bypass, optional.  Anything from 1M to 2.2M will work if you use one.

R19 sets a minimum value for the Drive pot, so that the output level doesn't go to zero when you turn Drive all the way down. No effect otherwise.

Thanks for the clarification!

That circuit near C8 could be a capacitance multiplier, if the Q emitter is supplying voltage to something downstream.  Those are usually seen with large cap values (1000 uF or more) to stand in for an even larger cap, using it with 1000p (is that the label?) would increase the effective capacitance but why not just use a larger cap?

Yeah, the 1n/1000p cap was from one of the schematics I looked at, while the 2n2/2200p cap was from another... Here's that section (though the link to the PDF is fixed)



Thanks for your responses!

[WonkoTheSane]

Older units have the 1nF cap, and newer ones have the 2n2. I've built a few and changing the 22k resistors on the balance pot to 18k makes biasing up the differential pair a little better, but it depends on what transistors you decide to use. The 2sc537gd is unobtanium but the 2sc828 is still pretty easy to find. I've never seen a unit that had the minimum gain resistor.

[jessenator]

Older units have the 1nF cap, and newer ones have the 2n2. I've built a few and changing the 22k resistors on the balance pot to 18k makes biasing up the differential pair a little better, but it depends on what transistors you decide to use. The 2sc537gd is unobtanium but the 2sc828 is still pretty easy to find. I've never seen a unit that had the minimum gain resistor.

Interesting. I don't know how I feel about ponying up Ge levels of money for Si, but hey, discontinued parts are just that, and anymore command R@R3 prices.

There are "Panasonic" 2sc828s from CN-based sellers, but my trust in their authenticity is ...lacking.

I got my boards and built one of them, so at least I know it works hahah https://www.madbeanpedals.com/forum/index.php?topic=34374.0

I'm thinking about getting some 2N3903s (in a lot) and seeing if I can get some of the lower-end hFE values there. So far, with just the 5088s and a couple of 4011s it sounds pretty decent. In a very early version I made, the biasing was really only eliminating noise, but that was all 5088s in it, so perhaps there's more going on with more true-to-original NPNs in there.

[WonkoTheSane]

In the original units the "bias" is actually a balance control to make the octave come out, as they didn't bother matching transistors. I kind of prefer mismatched transistors hfe wise and making it a control labeled "texture". As with most old fuzz pedals 100-250 hfe will work best in the circuit. I don't think 2n5088 is a good choice, 2n3904 would be better.

[jessenator]

In the original units the "bias" is actually a balance control to make the octave come out, as they didn't bother matching transistors. I kind of prefer mismatched transistors hfe wise and making it a control labeled "texture". As with most old fuzz pedals 100-250 hfe will work best in the circuit. I don't think 2n5088 is a good choice, 2n3904 would be better.

I did wonder, to be honest. I couldn't hear any semblance of "octave" in my previous builds (this board and an earlier one) so this could definitely be why. My recent build (to test the boards I've had manufactured) aren't any lower than 236 hFE   : |

I did order some NOS Motorola 3904s which arrived today (and some Motorola 3903s are on the way). Just got done testing and the majority of them are around 141 hFE. the rest only hover around 146-170 which works out. I want to wait for the lower value 3903s to experiment with the placement suggested by Aion as well.