Harmonic Percolator - Perfecting the Circuit

[ManicStrat1982]

Ok, so there may be a thread like this, or at least threads with bit and pieces of info about great sounding Percolators spread out over pages, but I wanted to start my own.....with everything in one place.

Has anyone achieved the "perfect sounding" Harmonic Percolator? (in there own opinion)
What does a perfectly tuned HP sound like? (I know some of these things are objective, just humor me)
What makes a great sounding HP?
How did you achieve it? What schematics did you use? What transistors and gains/leakage did you use?
Did you change anything in the circuit? If so, what?
What makes your version special?

There aren't that many originals out there. It's even said that no one knows what was really in it....that the schematics are guesses. This theory is kind of made believable by the fact that there are so many versions of the schematic. It's also said that the "first ones" had their circuit boards gooped so you couldn't see what was going on.

Please post pictures, sound, video, or just any info that might help shed some light on this circuit.

Perhaps together we might find the key to this mystical circuit! 

[Gus]

Do a search here and the web. it is worth the time to read the threads you find.

There are good threads at this forum that should help with your question.
There is a good youtube with the feedback part demonstrated with a real one

Are you asking people to just give you answers at a DIY page, are you starting a company?

[davent]

Steve Albini thoughts...

Found the rest i think...

[iainpunk]

if you are going on a next build, can i suggest the Harmoinic Instant Coffee, its not ''final'' yet
i posted an Introduction on the other HP thread going on, i encourage you experiment with the resistor values, the capacitor values, the transistors and the diodes.
at this point i have a BC516 as the PNP, paired with a BYX10 diode, and a reverse Beta 2n2222 and a red LED as the NPN side, both with 39k resistors.
with reverse Beta, i essentially mean that i flipped the C and E of the transistor, it still works, but is less linear and the gain of the trasistor is way lower, heavily misbiassing the 2nd gain stage

some ou you might already have read my recent posts and schematics, but i havent seen this thread before, and it seems a good place to post the schematic and name explain:


the 39k resistors are just place holders, if you decide on building it, i recommend experimenting on breadboard beforehand to find the resistors that match your transistors and diodes best.
the 4n7 was initially selected because its the value i have most of, about 60% of all my caps are 4n7, but in lots of diffetent materials and packages, i tried smaller and larger caps, but i keep coming back to the 4n7, it just sounds the best to my ears.
i have also put my NPN transistor in reverse Beta, that gives a really nasty, nasal, noisy tone

i called it the Harmonic Instant Coffee

Code Select Expand


percolator coffee is nice, high quality, fine tuned by the masters of coffee making and soft in aroma's.
instant coffee is mediocre at best, poor quality, the cheapest beans are used and can be really harsh tasting.
Harmonic Percolator is nice, high quality, fine tuned by the masters of pedal building and is soft in tone.
Harmonic Instant coffee is NOT NICE, simplistic, cheapest parts are great for it, and can be really harsh sounding.


cheers, Iain

cheers

cheers, Iain

[ManicStrat1982]

Are you asking people to just give you answers at a DIY page, are you starting a company?

=) My interest is purely academic. I am really intrigued by this circuit....the way it makes the guitar....kind of come alive so to speak. Someone said in a post that it kind of makes the guitar play itself. I totally get what they were talking about....it really does make it seem effortless...at least the one I built. However, mine doesn't sound very much like the demos. Its not very fuzzy...its more like a harmonicly rich overdrive.
I have been building pedals for a little over a year. I built one and now I'm totally addicted...well maybe obsessed would be a better way of putting it.
I have done a lot of experimenting. I know what happens when you do certain things, now I'd like to find out why the things happen.

[garcho]

I love my HP, I've built a few and they've all ended up sounding very similar. Always with the Giblet schematic and Small Bear transistors (2N404A, 2N3565) and never with clipping diodes. It's never left my pedal board, at least, not the one with the HP on it 

[ManicStrat1982]

and never with clipping diodes.

I'll have to try the one on my breadboard without clipping diodes. I've read a lot about that, and people seem to like it.  Also, I tried a couple 2n404 in mine, but I didn't have any luck....even with the 2n3565. it has some different specs than the 2n404a (which I'll have to order)
Right now I've got good results with a MP16b with hfe of 63 and 110ua of leakage, and a 2n2222a with hfe of 256. I haven't tried my 2n3565's in my bread boarded circuit yet...that will be next.
Have you ever tried adding a diode to ground on your input after the 100pf capacitor... like on version 2 of the Hermida  schematic?

[davent]

Deleted... Somehow i put my post in the wrong thread...
dave

[ManicStrat1982]

Steve Albini thoughts...

Thanks for the video posts. Pretty informative stuff about the history and workings of the circuit. I haven't seen all of those videos, just the first one. I never saw the one where he actually played the original pedal.... and the stilts are great haha

[Invertiguy]

I've been playing around with it on my breadboard lately and am really enjoying it. I basically gave it the Fuzz Factory treatment and replaced both collector resistors with 100k pots while making the feedback resistors switchable between the Albini and 'consensus' specs and making the cap between the collector of Q1 and the base of Q2 switchable between 100n and 1uf. The diodes are switchable between a pair of D9Bs, a pair of 1N4148s, and none. As for transistors, I'm currently using an MP16B with an Hfe around 60 and a leakage of around 100uA for Q1 and a 2N3904 with an Hfe around 200 for Q2, although weirdly enough I tried a P416B with an Hfe of around 100 and a leakage of maybe 20uA in Q1 as well and also got very good results, albeit a bit 'crunchier'. This circuit is pretty forgiving of different transistor types and gains and I would highly recommend swapping in whatever you have on hand to see what you like best.

Overall, it's hard to get a bad sound out of this thing. I will say that I prefer things closer to the 'consensus' spec (220k feedback resistor and 20k collector resistor on Q1, 750k feedback resistor and 91k collector resistor on Q2) to the 'Albini' spec (51k feedback resistor and 91k collector resistor on Q1, 3M9 feedback resistor and 91k collector resistor on Q2). The Albini has less gain and is 'smoother' and feels more like an overdrive than a fuzz, whereas the 'consensus' spec is much raunchier and fuzzier (which is how I like it!). I also played around with Dead End FX's 'Splinter' spec (205k feedback resistor and 8.2k collector resistor on Q1, 1M5 feedback resistor and 68k collector resistor on Q2) which was even gnarlier still! Long story short, there really isn't a 'best' version, it really depends on what you're trying to get out of it. I suggest you breadboard it and play around!

[garcho]

It's even said that no one knows what was really in it....

The George Giblet schematic is a trace of an actual unit, that's been around for many years now, not sure who is saying this?

[davent]

It's even said that no one knows what was really in it....

The George Giblet schematic is a trace of an actual unit, that's been around for many years now, not sure who is saying this?

Albini talks about the original he had having epoxy capped,obscured critical parts so when he had someone make copies of it they couldn't identify the correct parts, but in saying that, the video was uploaded in 2008 so i'm sure lots more to the story since then.

dave

[garcho]

the video was uploaded in 2008 so i'm sure lots more to the story since then

Totally. I just bring it up as an example of an actual unit that was traced. Here's this, emphasis mine, from 2007:

Here's my take:

- Definitely at least two versions.

- For the second version Dan's "version 2" mark-up of Alphonso's circuit looks correct (after tracing against the PCB).  The clear things look like resistors close-up

And there's this from Chuck Collins, the person who is arguably the voice of authority on the HP, owns the rights to the name InterFax, and continues to make at least one version of the HP true to one of the original designs:

I have seen incorrect diagrams of this circuit on the internet and they are still posted as of 2020. This is understandable since I have found FIVE different original versions of this circuit and that does create confusion. I make the version I think sounds the best. I hope you agree. All of my experience in the last 30 years shows that every original Harmonic Percolator used a complimentary Darlington circuit. (Sziklai pair) Interfax never used two NPN silicon transistors.

This EXACT duplicate of the original 1970's pedal employs the same obsolete type number Germanium transistor and same number Silicon transistor... It is a Fairchild 2N3565. The Germanium transistor had black paint on the number. They are genuine Texas Instrument 2N404A with original box dated 6-2-66. Also used are the exact same type number 1N695 obsolete Germanium diodes. (They are NOT the common 1N34 or 1N60) These are from BKC Semiconductor in Boston. The same type capacitors also are used. Mylar-Silver Mica, Tantalum, and Ceramic.

[davent]

Thanks Gary, great info to see!
dave

[ManicStrat1982]

I've been playing around with it on my breadboard lately and am really enjoying it. I basically gave it the Fuzz Factory treatment and replaced both collector resistors with 100k pots while making the feedback resistors switchable between the Albini and 'consensus' specs and making the cap between the collector of Q1 and the base of Q2 switchable between 100n and 1uf. The diodes are switchable between a pair of D9Bs, a pair of 1N4148s, and none. As for transistors, I'm currently using an MP16B with an Hfe around 60 and a leakage of around 100uA for Q1 and a 2N3904 with an Hfe around 200 for Q2, although weirdly enough I tried a P416B with an Hfe of around 100 and a leakage of maybe 20uA in Q1 as well and also got very good results, albeit a bit 'crunchier'. This circuit is pretty forgiving of different transistor types and gains and I would highly recommend swapping in whatever you have on hand to see what you like best.

Overall, it's hard to get a bad sound out of this thing. I will say that I prefer things closer to the 'consensus' spec (220k feedback resistor and 20k collector resistor on Q1, 750k feedback resistor and 91k collector resistor on Q2) to the 'Albini' spec (51k feedback resistor and 91k collector resistor on Q1, 3M9 feedback resistor and 91k collector resistor on Q2). The Albini has less gain and is 'smoother' and feels more like an overdrive than a fuzz, whereas the 'consensus' spec is much raunchier and fuzzier (which is how I like it!). I also played around with Dead End FX's 'Splinter' spec (205k feedback resistor and 8.2k collector resistor on Q1, 1M5 feedback resistor and 68k collector resistor on Q2) which was even gnarlier still! Long story short, there really isn't a 'best' version, it really depends on what you're trying to get out of it. I suggest you breadboard it and play around!

I'm also getting some great sounds with a MP16b with hfe of 63 and 110ua leakage for q1. For q2 I was using a 2n2222a, but last night switched to a 2n3565 with hfe of 224 that I like better. I have 1n60 diodes in it, but have tried 1n34a, 1n270, etc. I ordered some Russian d310 diodes that are the Russian equivalent of the 1n695 and am waiting for those. I also ordered a couple Texas Instruments 2n404a to try....and so I can build one with original specs....or at least as close as I can get without the original diodes. I don't have the
mica cap either (I'm not sure if that matters too much). I think the Russian d310's will suffice as well beings they are equivalents of the 1n695.
I will definitely try replacing the collector resistors with pots and I was thinking of trying an eq section that adds around 20db of boost (or subtracts it) to the bass and treble frequencies like on the Keeley Fuzz Bender for a "tweaker" version of sorts. And I'm also trying to figure out a good way of boosting the output on the original circuit without it changing the sound...and maybe a little more gain (distortion).
Overall I am having a lot of fun with this circuit.
As for finding the "best", I guess I should have said " closest sound to the original. 🤘

[Gus]

You can have fun with the basic design of the circuit
a 2 NPN built idea based off the HP
https://www.diystompboxes.com/smfforum/index.php?topic=91764.msg784041#msg784041

If you do a search you will find a lot of information on the HP

[bmsiddall]

Have a play with the 750k feedback resistor.  Winding that up to >2M gave me octave down notes on my build.

[iainpunk]

Have a play with the 750k feedback resistor.  Winding that up to >2M gave me octave down notes on my build.

replace it with a diode bazz fuss style and you get a octave up effect.

cheers

[bmsiddall]

replace it with a diode bazz fuss style and you get a octave up effect.

cheers

On the to-do list 

[iainpunk]

replace it with a diode bazz fuss style and you get a octave up effect.

cheers

On the to-do list 

its not really a 'nice'' octave up, its really chaotic and harsh.

cheers