Harmonic Percolator (George Giblet's schem) build report

[bool]

Interesting clips ... anybody from europe here? Where to get 2N404?

@Alex frias : you used metalcan 2N2222 or TO92?

Or can somebody who knows the circuit test BD135/136 ... are they any good for H.P. or a waste of time?

[DougH]

Alex, your clips sound great. I love your playing and you have a really nice sense of touch.

In your clips I'm hearing a little of the "wheeziness" I was trying to dial out of mine. But that may be the way it's supposed to sound, who knows. In any case this is a subjective personal taste issue at this point. (After all the axis fuzz is supposed to sound like that when it is full bore.) I hear the bass growl in yours too, which is what I am getting as well.

I'll do a clip later this week and we can compare notes.

I just ordered 10 2n404a's and some more 2n3565's. I'm going to try to duplicate on the breadboard what I have in the Brick. Hopefully I'll find out if my hfe observations are correct and how easy it is to do a consistent build.

[alex frias]

"@Alex frias : you used metalcan 2N2222 or TO92?"

Plastic TO-92, ordinary ones.

Soundclips: I just listen to the mp3 files I posted, I didn't listen to them after "wav->mp3" conversion, I used River Past, even an annoying aliasing noise can be heard, hummmm... some sound artifacts were created, but you still get part of the picture.

I don't want to look for excuses, but it's not the exact sound of the thing.

I think what you, Doug, are referring as "wheeziness" is part of the percolator's sound character when full cranked, but I think it can be less than mine exhibits right now. What I would like to achieve, if possible, is less "wheezzo", more sub-harmonic (octave down kind of) richness. At the same time retaining the lovely mutating sustain, the weird kind of polyfonic intermodulation and dynamics it has.

One thing some people are pointing at as a significant diferencial in Albinis vid is the singular instrument he is using. If its single coils has less power than is usual I believe it is good candidate to be blamed of "horrible" diferences between his HP-1 and the Barge claimed-clone. I still didn't listen to the Chuck Collins clone.

One thing I'm sure about this circuit: it is one of the most transistor sensitive I have notice. I think it surpasses Fuzz Face on that matter.

[Gus]

I like this thread it is fun and also cool how people are trying different things and posting about them.  The HP is still a mystery because none of us have a real working unit to try and measure voltages from.

Maybe more variations like the brick circuit will result from this and other threads about the HP.

I need to finish building the 2n2222a and 2n2907 one I started

I want to try a two 2n2222a (Hfe about 200) version at .1ma using the two npn circuit I posted
Higher Hfe at Q2
47K collector resistor and 470K C to B Q2 part

10K collector resistor and 1.2 meg C to B Q1 part

.047uf, .1uf and .1uf caps

[DougH]

What I would like to achieve, if possible, is less "wheezzo", more sub-harmonic (octave down kind of) richness

Alex, if you have a 2n404a with higher hfe, give that a try. I suspect that may be what improved the sound of my Brick.

Gus, I don't think collector voltages are that significant in this circuit. I think the current through the feedback resistor and base is more significant. Using large feedback resistors to reach a target voltage is not the correct approach IMO. That starves the base of current and really changes the operating point in a way that screws up the sound. I was thinking it needed around 4.5v on Q2-C but even at 2.1v it's only a 20-25uA difference in Ic- not much. Not enough to worry about IMO anyway. Plus, your voltage calculations are partially based on George's assumption that the NPN Q2 is 100 hfe. Don't know if that's true or not but in the end I don't think that or the voltages are that significant, because Vce (which is significant) is so low.

But by just listening, I'm hearing major improvement in sound so far with a higher hfe Ge PNP for Q1. I noticed it with both the 2n404a's and a 2n614. For some reason I could not get an AC128 or AC188 to work- just didn't produce enough gain. But with stock resistor values, 2n404a's work fine, regardless of Vc. It's just a matter of picking one with sufficient hfe IMO- kind of the antithesis of the lower hfe you want for something like a fuzz face.

Incidentally I tried a 2n5087/5089 pair and it worked fine. It was a little wheezy sounding too and not as smooth, but it did work. I think Ge PNP and the 3565 are the way to go for smoothness.

[Gus]

Doug

Looking at your Brick Schematic.

You have 2.1VDC at Q2 collector so you should have about 1.5VDC at Q1 collector  (9VDC -2.1VDC = 6.9VDC,  6.9VDC / 100K = .069ma,  .069ma X 22K is about 1.5VDC)

So collector to collector you have about .6VDC?

[alex frias]

Well, I did more tests with my Sardonic Percloretor.

1 - A new combination of trannies.

The only 2N404A with significative higher hfe also shows a very high leakage... but sounds good anyway, but with a very reduced gain. So I tried it with a BC108 (hfe= 450). It worked but sounded worse than the 2N404A/hfe=40 + Si/hfe~200, what I liked most and considered as my reference combo.

This time a lot of AC128, AC188 and 2SB54 worked, but none with a sound better or even close to the reference combo. Aside from hfe's I didn't make any other measurement. Doug, I can't agree more: 2N404A is the right path for smoothness, perhaps 2N3565 too.

2 - I tested again several tranny combinations, this time also variating Q1 collector to ground resistance (100K to lower values) and it just gets worse in wheezzo department. Doug, I didn't notice any improvement in octave down effect... And in my opinion, the reference combo with a fixed resistor of 90K worked best, sound wise. Maybe I will try higher resistance values, but that was not the direction indicated by the pics on the net.

I am very happy with the sound as it is at this time, with the schematic I post using the reference combo. But I think it's not THE SOUND of the thing...

Anybody is or has a close friend to Mr Albini? I really would love to see his Percolator guts...

[DougH]

Gus- yes those voltages sound right. I will double check tonight.

Alex- I found that going too high with hfe seems to reduce gain too, at least with Ge PNP's (2n5087 worked okay gainwise). For me it seemed to be a balancing act with the 2n404a, needed high enough for good sound but too high it loses gain.

[alex frias]

At my lunch time I did listen to the ToneFrenzy soundclip. Well, at first I think they didn't cranked up the thing. And comparing with the noises I got from mine version, I prefer mine. I even can't hear our lovely octave down side effect.

A funny thought occurred me: anybody tested this thing with batteries? I'll try it out just for fun!

Doug, I consider your observation very close to the idea that's growing up in my mind in direction to the key of the treasure.

Sometimes I reread our posts and find myself laughable. What an obsession!!!

[DougH]

At my lunch time I did listen to the ToneFrenzy soundclip. Well, at first I think they didn't cranked up the thing.

Yeah, it's a lousy demo really. They fell in love with the low gain sounds so that's all the really demo'ed.

[Gus]

http://www.bargeconcepts.com/bp/bphistory.html

http://www.bargeconcepts.com/bp/hpschem.html

The above links and the video

Has to be NPN NPN? first link 4th paragraph

The 2nd link why have a EF "feeding" a gain stage that is sensitive to the source resistance "feeding" it?  as drawn it will force the top transistor to operate at it's max gain.  The first stage being an EF will remove the interaction with the guitar.  The one we are working from is a part of the feedback setting of the gain stage(like a FF type circuit)

I hope we can make a sticky or something to keep all the posts and threads about this circuit together.

[DougH]

I like the "has to be NPN" with no explanation why. Most of the BC stuff appears to be clones of things designed by others, except for routng and feedback looper stuff. I never understood why they stuck a buffer on the front either.

I made a video of the Brick today and I'll post a link when it's up on youtube.

[DougH]

Here's a youtube clip of the Brick: http://youtube.com/watch?v=1lc85IimYrQ

The video is grainy because I compressed it down to 10% to shorten the upload time (it's 4 minutes +). It's also out of focus.  ( Getting used to a brand new camera. should have used macro...) Anyway, the main point was the audio which was through the camera mic and untouched (except by the youtube manglers probably). There's a lot of natural reverb in the clip because the camera was on the bench pointed away from the speaker. So it's mainly picking up the bounce. (Sounds kind of cool IMO.)

I played around with it more last night, and as Alex mentioned, I think it sounds much better with weaker pickups. With the PAFs on the Hamer it just sounded dull. But with either the single coils on the strat or the retrotron humbuckers on the hollow body epi this pedal sounds really good. It sounds like it comes "alive" with weaker pups whereas hot pups just swamp it. (Retrotrons are similar to the old DeArmonds, a weaker and more trebly humbucker.)

[JOHNO]

Well i guess thats one brick that wont get thrown through a window, it sounds sweet doug. Also watched you amp vid ,nice amp mate. You could have worn a hat though. Oh well you know what they say the more hair you loose the more head you get :o ;D

EDIT; Sorry doug thats dry aussie humour,think ive had too many beers,cheers

[George Giblet]

One idea that might be worth trying is to use a silicon PNP for Q1 then add a linearizing network, comprising of a diode in series with a resistor, across the Base and Emitter of Q1.

The same trick has been used for Q2 on the fuzz face.

[bool]

Interesting ... do you have a schematic (just for the idea you posted)? I'm not sure if I understood correcly.  (just curious: how does it look like in a fuzzface?)

[DougH]

I received my 2n404a's yesterday and measured a couple of them. At least one is hfe 110. So I'm going to test my theory that it requires a 2n404a with hfe ~100 in order to sound "good". If I'm correct I'll do a quickie comparison video subbing a few different hfe values on the breadboard. It's so easy to do videos with this camera it's ridiculous...

Another thing I thought of with the "weak pickup sounds good" issue is it may just be a matter of tuning the input coupling cap to the pickup. I'll try a .022u with my PAFs and see how that sounds. A switch or Gagan-Easy-Face style input tone control could be used to compensate for different pickup types.

[DougH]

Okay, after playing with my 2n404a's this weekend I like the 2n404a hfe to be in the 80-130 range. It's a personal preference thing but below 80 it sounds "misbiased" and above 130 it starts losing "gain" (distortion and harmonics). Also, pickup types come into play. With PAFs, hfe > 130 worked better because there was a hotter signal from the guitar. With weaker pickups, lower hfe may be preferable. Of course you can find a balance that sounds good in both circumstances. However, with PAFs a .022u input cap sounded more defined, whereas the stock .047u tended to "mush out". I didn't notice much difference with/without the 100p, maybe it helps with RF.

After trying different hfe xsistors I measured voltages and didn't notice appreciable differences. I think this is a "current thing". I may build another one and put a switch or "thickness" control on it to compensate for different pickup types. But this concludes my experimentation with it. My advice if you want to build a perc type circuit, would be to get some 2n404a's and experiment, find the hfe you like. I got a 10 pack of them for around $20 on Ebay, delivered in less than a week. The hfe's ranged from 80 - 330.

Check out the video of the Brick, if you get a chance. This is my "reference" sound for this thing:

http://www.youtube.com/watch?v=1lc85IimYrQ

[DougH]

Here's another derivative I had not seen before: http://www.rockandrollworkshop.com/index_files/harmonicannihilator.htm

[alex frias]

Hi, Doug, I saw this stuff in another net place, but when I heard the sample I gave up to believe it is capable to get a Percolator decent expresso out of it.

I think it was a kind of mix of a Percolator with some sound degrading circuit... I don't remember now...

Still no Mr Albini's friend appeared?