Interfax Harmonic Percolator working circuit DANTONE Pastry Chef

There's a lot of "information" out there about the Harmonic Percolator.

Here is one I spent a few days on working out the kinks that works as it should.

I started out with Iviark's nice small layout (posted below) but I found some problems.

The two blue additions to the schematic drawn below are my fixes, and had to be manually added to the existing layout.  I haven't had time to draw a new layout yet.

The diode lift switch works fine, and is a recommended mod.

The first problem was a fizzy/crackly/velcro kind of decay on the fuzz mode.  Adding the 1 nF capacitor parallel to the clipping diodes near the output solved this problem.  You could experiment with other values maybe in the pF range.  This cap shows up on some of the traced schematics out there.  In my case, it was needed.

I also added a 1 nF cap from the collector of Q1 to ground.  Parallel to the 20K resistor.  This cap is present in the MadBean schematic, as well as a few others.  It increased the transistor fuzz/gain of the circuit and generally made the fuzz sound better.  Without it, the fuzz setting was sort of "insane, bogus" and "fucked up" sounding.  I wasn't too thrilled about it.  With the cap in place, it's a huge, usable fuzz tone.  "Why doesn't my pedal sound like the videos."  In my case, this cap was the answer.

I also experimented with adding 220 uF filter capacitor to the 9V rail, then I tried 47 uF with 100 nF in parallel.  BOTH OF THESE ADDED EXTRA NOISE.  Not what I'd expect, but the filtering actually made the pedal noisier!  I could swear it sounded a little worse that way too.  Not recommended.  I would recommend building with shielded I/O cable from the jacks and circuit board if you have the time and patience.  That might help the hiss a little bit.  You could almost think of it as a pleasant hiss but if I'm being perfectionist I'd rather it not be there.

I used 1N695 diodes with a forward voltage of about 290-350 mV.

You could use 1N4148 if you want the fuzz to be louder.

My Q1 was a 2N404A of 186 hFE.  Q2 was a silicon 2N222 of about 150-180 hFE.  I could have gotten away with a little less gain I guess but these values work fine, they just are a little fuzzier at lower settings on the Harmonics knob.  Doesn't quite work as a "clean boost" it's more of a "dirty boost" with the diode switch lifted.

This pedal does clean up exceptionally well with the guitar's volume knob, like a Fuzz Face.

Diode lift sounds great on this pedal, giving a warm fuzz/boost/overdrive.  Be careful though because it can get dangerously loud.  Watch those cilia.

This is the "stock" or "collective" version.

If you want to build the "Albini" version, check out the MadBean pedals PepperSpray2018 PDF file for the correct values.  I think the "Albini" version is a little lower gain and a little less wild and crazy.  I'd like to build that one too just to see.

Behringer MDX-1200 Autocom Mods (OH Boy...)

I'm not sure where to start this one so this is the GroupDIY post that started it: (those guys are trouble, I tell you.)  One thing to note is the schematic is apparently nowhere to be found in 2018, so this whole thing is based on text advice.

https://groupdiy.com/index.php?topic=39556.0

I got an MDX-1200 on ebay for pennies, and it came to me in apparently NOS condition.   Including original packaging.  However I later discovered that the bypass switches are both intermittent.  Exactly 1 year later I decided to do the work.  On a slow day that needed some activity.

The achilles heel of these cheap compressors (behringer, Alesis) if you go back through my posts is always the cheap switches.  They just don't last.  I've bought and sold a few of these kinds of boxes with the same problems.  You simply can't rely on them.  I may have to even eventually "Mod" this compressor with some regular toggle switches.  For now I just don't touch them.  (If you jam them enough times they work.)  I could also recommend to not buy these.  If you are in a busy studio.

Basically the plan, as usual, is to upgrade the op amps and some of the capacitors.  I made an image list that I will post here:



All the stuff I got from Mouser.  Here are the parts I ordered (I did also order some spare capacitors, so there's extra on this list than what was used.)



The job took about 4 hours with several breaks.  That's FOUR HOURS I could have spent building my Serpent SB4000 Kit!!!  Let's call it practice, I guess.  The main thing is that I got it done and it worked on the first plug-in.  Just testing my skills.  I was careful not to destroy the pots when pulling off the knobs.  Something that went wrong with my Composer and Alesis mods.  I decided to de-solder the red and black wires to the front panel LED's so I could remove the front panel and freely work on the PCB.  I had to re-solder them at the end of the work.  (Pictured, sort of.)

I also learned a new technique of using desoldering braid.  I place it across the 7 holes of the big op amps which are removed already, and gently roll the iron across the braid for about 30 seconds.  You lift up the braid and most of the solder has been removed.  Pretty cool.  My new Hakko tweezers also really came in handy for desoldering the op amp legs after I cut off the op amps.  Stick the iron on there, pull the leg part out with the tweezers.  Then use the desoldering braid.  Then put in your gold pin sockets, and finally your new op amps (after you have done your PSU bypassing with .100uF caps.)

This was also my first time of bypassing the op amps.  You can see that I soldered some small ceramic capacitors to the bottom of the PCB.  What you do is find the + and - voltage pins on each op amp IC, mark them with a black dot (marker), find the nearest ground solder pad with a multimeter (in continuity beep mode) circle that.  Then put 2 caps from + and - to ground.  Apparently this is an oversight in shitty designs like these that should have been there if it were a good design in the first place.  My caps were small enough (MLCC) to fit under the board.  It was kind of fun and I'm glad I learned a new technique.  If you try to use certain high slew rate op amps in an un-bypassed circuit sometimes they will oscillate and possibly even destroy themselves.  This happened to me with one of my Symetrix 501 mods from years ago.  That poor bastard finally got sold on Reverb.com to someone with a better heart than me.

Sound is very high-mid forward and tight, with good bass extension.  It's a pretty aggressive punchy sounding compressor.  There is some distortion or clicking noises on the attack of bass notes.  Knob positions don't seem to help that much.  On bass synth anything below 10ms on the attack knob results in ugly distortion.  I wonder if I need to disable the gates or "enhancer" somehow.  This is making the compressor questionable for studio use.  HOWEVER it sounds pretty killer on snare drum.  Beats my plugin chain.  I think a tiny bit of crunch would just blend in with the snare wires, I'm not sure I can even hear it.  The Midas 512 is also working great on snare drum EQ.  So this goes next to my Composer mod as another "dirty compressor" which is once in a while useful.  I used it on a dirty synth line also and it made it stand out in the mix.

I can't recommend this project based on time spent and shoddy switches.  But if you are the tinkering and learning type it could be worth doing, if you already seem to have and MDX-1200 on hand.  Total project cost including the compressor was about $75 so yeah this is pretty dirt cheap, even after the mods.  Comparing value, a DBX 560A for under $150 or so would be a much smarter investment.  That's a tool that's going to be a lot more useful in any given session, and probably built a lot better as well.  The bottom line is these Behringer compressors are sort of a novelty.  People just mod them because they are so cheap, and the circuit board is easy to work on, once you get it out of the case.  These kinds of basic mods don't really turn them into "pro gear" it's a bit of putting a Hemi into a Pinto kind of result.  If I had to pick one, I'd say the MDX2100 Composer is probably going to be your better bet than this one.  The Composer just turned out better, and I think the earlier generations of these boxes had slightly better switches.

I am even told that the VCA in these are very similar to the ones in SSL compressors, if not even slightly better than the SSL VCA.  Not that they are going to sound the same because of that, but that this compressor might be actually useful to me now.  It's sitting in a sad and happy place in my rack with my Alesis 3630 and Behringer Composer from previous attempts.  Sort of like how Sylvia Massey has her broken Army Man Compressor.  It's 3 rack spaces of semi-working upgraded stuff that does sound pretty cool, but that could never really be sold or recreated.  It's all in good fun, and a good deed never goes unpunished.

Boss GE-7 Noise Reduction Mod and Review

Noise Reduction Modification

The main "problem" with the stock GE-7 is a little bit of hiss.  If you turn the 100 and 200 Hertz sliders all the way down you can hear it.  If you start boosting mids, treble, and level, you can hear it even more.

There are popular "mod kits" out there that involve a bunch of capacitors, but I decided to take a different approach to my mod.

I only want to change the 3 Dip-8 op amps.  Simply clip the left or right 4 legs of the op amps with small dykes, then bend and break the other half.  Remove the remnants of the legs with tweezers and a hot soldering iron.  Clear the holes the best you can with desoldering braid or a better tool if you have one.  Solder in some gold-pin sockets for best results (cheaper sockets are less reliable over time.)  Then insert three NE5532 op amps into the sockets and put the pedal back together.  Make sure they're oriented in the correct direction.  (There's a helpful graphic on the PCB.)  The job took me less than half an hour.

Now when you plug it back in, if it works, the noise should be very low, much quieter than most guitar pickups, or other gain pedals.

The reason I wanted to leave the tantalum capacitors in place is I like the way they sound.  To me this is a "tone" pedal not just a simple equalizer.  I have a cheap Donner EQ from Amazon that doesn't have the character of the GE-7.  It's clean and clear, yeah, but I think the GE-7 has a character that I really like.  It's "warm" to use a wide-ranging cliche.  Vintagey but clean enough.  Hurts my brain a little bit to type that out.

Some famous examples of gear using tantalum capacitors include the Neve 1073 channel strip, the Red Llama and MXR Distortion+ pedals, and the Telefunken ELAM 251 microphone.  If those circuits need to sound cleaner I'll tell you you're crazy.  There is a subtle character related to the caps.  Audio, and especially guitar, is not always about test measurements, but "Tone, mannn."  If the sound is happening then why have a temper tantalum?

It's also not a great idea to swap the caps because any deviation in capacitance value will change the operating frequency of the band it's related to.  If you do change caps, make sure to use exact values.  I just don't see any reason that electrolytic or ceramic caps would be desirable, unless you just don't like the way the GE-7 sounds.  They won't help the noise floor any.  The noise reduction comes from the NE5532 op amps which have a much lower self-noise than the stock JRC '22' op amps from Boss.  NE5532 cost about 35 cents apiece.  They do draw marginally more current than the stock op amps but I don't think it's something to worry about unless you're a battery dude.  I hate batteries.

So the whole GE-7 mod suite, Dantone version, costs about $2.50 and half an hour of labor.  I do hate noise, and tend to "kill it with fire."

REVIEW

My initial reaction to the GE-7 is "Why did it take so long for me to have this."  My 19-year-old guitar brain thought it was some sort of boring "utility" effect but in practice it's so much more.

I've been reading a lot about how Queens of the Stone Age, and Converge, get their "secret sounds."  A big part of both of their approaches is massive mid boosts into cranked up amps.  Kurt Ballou, in particular, is known for using the GE-7, and the EMG 85 bridge pickup, for most of the classic Converge sounds.

It gives a totally different sound than an overdrive, treble boost, clean boost, etc.  It's something you need to hear for yourself to believe.  The resulting tone is fat and crusty, like the amp is about to explode or catch on fire.  You can tailor the brightness, or bassiness to taste as well.  It's nasty.

The cool part is the amp is where the crunch is coming from, not some diode circuit or whatever.  Pushin' doze' toobs makes me evil smile and inspires the playing.  My Marshalls love this effect.  Blues licks sound extra bluesy.  Putting a fuzz in front of the GE-7 is also pretty stoneriffic.

The last picture below shows a typical setting of mine.  Boosted mids with the level all the way up.

Cornish SS-2 and G-2 DIY Pedal Builds and Circuit Notes

Cornish SS-2

First one I built to completion was the SS-2 from the Guitar PCB Super Sonic SS-02 PCB and build guide.

They randomly changed the output pot from 10K to 100K without telling you.  "No no no."  So I went ahead and subbed in a 10KB pot.  (Not a 10KA, I guess I like the B taper better in this circuit.)  One of Cornish's ad slogans is "Low output impedance, capable of driving long cables," etc.  So I decided to stick with the Cornish program.

Diodes are a little "tricky."  I just went quickly with the recommended 1N60P diodes recommended by GuitarPCB and they sound fine.  Wasn't really in the mood for an extended shootout.

1N60P is NOT a Germanium diode.  (Despite what some Chinese eBay sellers will tell you!)  It is a Schottky silicon diode with a similar forward voltage to a lot of the germanium parts, so it is used by a lot of modern pedal makers as a substitution (MXR Distortion+ for example) that meets ROHS requirements.  Whereas the germanium parts would not meet ROHS laws and therefore would not be legally allowed to sell in some markets today.  (Yikes!!)  Maybe that's some of the complaint about the modern MXR yellow pedal.  EDIT: there may have been an older diode with the 1N60P name that would have been germanium, just to add to the confusion.

In my Distortion+ build I find the 1N60P to be a bit brighter sounding than some generic NOS 1N34A in the same spot.  Same in this pedal--it's a little bright, but I like it.  I just have to run the tone control a little bit lower than noon at highly boosted settings.  This is a fun diode trick in the Dist+ and the SS-2.  It would be great to have them on a toggle switch.

If I had thought about it I might have stuck some old germanium diodes in there but I'm leaving it for now.  This was a fairly straightforward build.

Stick with the LM741 op amps it's part of the sound.  They are slow, dirty op amps and that's part of pedals like this one.  They contribute some clipping and sonic degredation that is part of the recipe.

It's capable of massive gain and output volume which can get a bit hissy at certain settings so I added extra filtering on the 9V supply (total of 330 uF.)  Another thing GuitarPCB randomly changed for no reason.  I also shielded some of the I/O wires to help again with noise.

If you crank up your Marshall or other Doomy amp and boost the SS-2, you will get wall-shaking bass out of this circuit...pretty cool.  Once again Pete Cornish is telling it on the tin, "Better low end retention than any other pedal circuit."  -O_o-

I didn't mess with the buffer on this one because buffers are a pain in the ass.

Cornish G-2

This one came second, and I'm glad because it was hard.  And there are a lot of common problems and misconceptions about this circuit.

I believe what a lot of people have come to agree that the original Dirk H. trace was wrong.

When you build this circuit to those schematics it's a muddy, flabby mess of a sound.  Just terrible sounding.  Any quick listen to a decent G-2 demo of a Cornish pedal on YouTube will show you how far off it is.  The video where Shnobel compares it to the Vick Audio pedal (probably built to this wrong schematic) is a good real world demonstration of how far off these builds can be.

Chasing the diode fairy is not the right way to "fix" this pedal.  It's just salt on a burnt duck.  And a significant waste of time.

However I did go with NOS germaniums this time after some swapping, the 1N695 are good sounding and in the general correct range.  You want somewhere between .20 and .30 V on the diode tester on your multimeter for forward voltage.  Mine are on the higher side of that range.

WHAT'S WRONG

I decided to go for the simplest solutions first.  There's a giant filter cap across the transistor on the gain stage right before the sustain pot.  It's a 1 nF cap in a lot of these circuit traces including the GuitarPCB PCB I am building on.  You need to change that to a 100 pF!!!!!!  Picked this tip up deep in the comments on Iviark's blogspot.

Suddenly, a LOT of your highs came back and the tone control is now useful.

The second cap to change is another 1 nF.  On the first clipping stage, change this to 560 pF.  Got this tip from GuitarPCB themselves in their most recent revision, so props to them for that.

Now you have that glassy chime that this circuit is supposed to have, as well as a truly useful range on the tone control.

Two simple cap swaps to go from mud to sweet chimey sustain.

EDIT 4-6-2018.  You're going to want to change that last 1nF cap on the second clipping stage as well.  I just threw a 500 pF on there.  Basically you want it to be much more similar to a standard Big Muff value across the board.  Any number of these 1nF caps will give you a huge muddy sound that somehow is painful to the ears..

After changing all three of the 1nF filtering caps to much smaller values, the whole pedal sounds much more even and is usable at any setting.  The wall of mud things seems to go away with these cap swaps.

God knows how close this is to an original in a side-by-side shootout... but it sounds damn good.  Sort of a sweeter/more overdrive-y muff type of sound.

Guitar PCB told me to change the transistors.  Once again, "No."  This probably doesn't make a huge difference.  I went with the intended BC549.  They also want you to change the tone pot to 100K on the newest board but I'm sticking with the stock 25K tone pot and 10n cap.

I did a buffered bypass and some shielded wiring and it took me a few extra hours to get this wired up in the box.

In review, play with the 3 1nF cap values (100 pF, 500pf/500pF +/- is good here), and you can play with different germanium diodes to get where you like.

It was worth the trouble though, these are some fun pedals.  They are not particularly "amp like" or "organic" or anything of that nature but they sound "Cool and good," as Billy Cardigan might say.  The thing that's notable is that they are indeed pretty far away from your typical dirt boxes.  Cornish has his own thing going on and I like that.

They both seem to have usefulness but the SS-2 seems "special" to me after playing them for a few weeks.

PICS