Monday, September 14, 2020

PETE CORNISH: G-2

The Pete Cornish G-2 is a Big Muff variant:  germanium diodes, double buffered, and sporting a low-pass filter instead of the usual Big Muff tone stack.  

That being said, this is not your standard Big Muff - there's little resemblance in sound, largely due to the lack of scooped mid-range provided by a standard Big Muff tone control, and the germanium diodes change the clipping profile quite a lot.  Do not expect this to sound like a Big Muff.  

There are so many opinions on how this should be built - but keep in mind that the same topology is used for any variation.  The main complaint is that it's too dark-sounding using stock values, so builders try to address this by making various changes.  Please read the comment below by PedalBuilder re diodes, as this is certainly worth some consideration - they may be right in that glass-encased Schottky diodes are used, the diodes may not be germanium given all of the trouble that people have getting this to sound right.  

The arguments usually centre around key capacitor values and the odd resistor.  Apart from being germanium, the diodes used in the G-2 were not confirmed during the trace done by Dirk Hendrik, so this is also a source of much debate.   Here's the original thread on FreeStompBoxes.

What was missed in these earlier discussions was that germanium diodes can leak current - unless you have low-leakage germanium diodes, expect to be disappointed.  Why, leaky diodes reduce gain.

Some people change the diodes too, but once you get away from germanium diodes, you may as well make a Pete Cornish P-2.  Pete Cornish uses the same PCB for both; one uses silicon diodes and is supposedly based on a Rams Head-era BMP. The other is closer to a BMP from the former Soviet Union but with germanium diodes (or perhaps glass-encased Schottky).

The layout below has the same values as the Dirk Hendrik trace; alternate values are all listed below.

This needs the standard Cornish buffer in front, and should be wired as buffered bypass if you want to stick to the Cornish program.  If you don't plan on using the buffer, it's probably a good idea to drop that 220n cap on the input back to 100n.  

NOTE:  It does have one buffer on the board, as the G-2 is double-buffered.  

PETE CORNISH G-2 - VERO LAYOUT

The odd coloured lines running across the vero board are power and earth - ignore...  nothing to do, just a little visual helper for me that I used to include on my layouts.

PETE CORNISH G-2 - VERO LAYOUT



PETE CORNISH G-2 BUILD PICS

Initially, I built this with 2N5088 and BC109C transistors, so if they look the wrong way around compared to the BC549s on layout, that's why.  I ended up swapping the B109Cs out for standard transistors, as they weren't really doing anything special - not even sure why I put them there to start with now.

I dropped the 4u7 coupling caps down to 2u2 and the 1n caps to 510pf caps across collector and base.  I used 1N277 germanium diodes with 220n caps in the clipping section to retain the primary character of the clipping sections.  

It sounds pretty good - I like it more than the stock values for chords.  I find that this pedal appreciates running at a decent level, and it's very dynamic compared to a Big Muff (probably due to the lower gain / less clipping compression).  At wife is at home levels, it sounds constrained, at least to my ears.

It does great lead tones with the stock values, it's rounded and smooth, and is probably not one for the speedy shred masters out there.

Chords are hard to describe as it's so touch responsive - moderate gain levels seem to suit chords better than the higher gain settings.  I found the tone control (low pass filter) is only needed for bridge pickups.  


Pete Cornish G-2 progress pic


PETE CORNISH G-2 - SUGGESTED ALTERNATE VALUES (BY OTHERS, NOT ME)

  • 10n cap on the input buffer - reduce to 1n (brightens the signal from the buffer)
  • 220n caps down to 100n
  • Collector and emitter resistors on input boost stage, change to 18k and 100ohms
  • Reduce 4u7 caps down to 2u2
  • All 1nf caps down to the standard 470pf Big Muff values

You will undoubtedly find more via some googling, along with the heated debates.

2 comments:

  1. Bit of a belated comment, but I've begun to suspect that the G-2 doesn't use germanium diodes. I've measured about 20 different types of germanium diode, and all of them at least 12 times more leakage than I would want for a germanium diode in a transistor feedback loop. Cornish seemingly doesn't have that issue, so I went back and looked at Dirk's original tracing thread for the Cornish G-2. Looking at the trace in turn led me to wonder whether the G-2 might actually use a DO-35 or similar glass package Schottky diode, perhaps something like the HP2835. As your post noted, Dirk was not able to determine what model of diode was used, but the pictures show a glass package diode and measurements showed a Vf of roughly .22 volts. That Vf measurement is on the low side for a germanium diode, but it's well in spec for a Schottky diode. Moreover, using a Schottky would eliminate the leakage issue that is a fundamental trait of seemingly every germanium diode that I've encountered I know that Cornish describes the G-2 as having "warm Germanium qualities," but as Kevin at Aion and others have documented, Cornish also has a track record of exaggerating and making what I would say are misleading statements about the circuitry inside of his pedals. I'd be curious to hear what you think on that point.

    ReplyDelete
    Replies
    1. Hey PedalBuilder,

      I think you make some very valid points, thank you for sharing - you are quite possibly right about all of that.

      I've not played with a G2 in ages, largely because they never sounded as good as the demos. I'll have to have another look.

      Delete