Harmonic Percolator clone debugging help

[entropicp]

I'm a relative newbie. This is my fourth build and the one with the worst issues I've had so far. I spent 6 hours yesterday and  last night debugging and was getting very frustrated. After taking the voltage measurements I think I've found an issue with the diodes since they aren't getting any voltage. But I was wondering if it was an issue with the parts or the layout and also if there are any other obvious issues with the circuit that I should be aware of. I already checked for solder bridges and if I was following the layout correctly.  Any help would me much appreciated and will probably save me a trip to  the psych ward.

Here's the info:

It's a Harmonic Percolator clone. Schematics from here:

http://www.geocities.com/george_giblet/effects/percolator_later_rev1_1.png

I built it on stripboard bases on the layout from here but made some minor changes, although no changes to the circuit itself. This is my first stripboard build:
http://www.fredric.co.uk/misc/interfax%20harmonic%20percolator%20stripboard%20veroboard%20layout.gif

I made a diagram of the actual layout I soldered with DIY Layout creator. I wired the transistors with both emitter tabs facing each other to the inside. Emitter, collector and base are marked on the diagram.




The LED Turns on/off and the sound works when bypassed. I'm fairly sure there isn't a switching issue because when touched I circuit board with wires when turned on I got a crackling sound.


Here are the substitutions I made:

R1-680k
R4- 20k
C2- 50 nF
Q1- 2N1303 germanium
Q2- 2N2484 silicone

And finally the voltage readings. I was getting 0 from all of the components on the ground strip, C1, and the lead  of C2 connecting to C1 as well as the diodes. Stupid question, but should I get a 0 reading from the ground strip?

Battery= 8.42

V - battery= 7.07
V + battery=8.16

Q1
C =1.23
B =1.07
E =1.08

Q2
C=2.05
B=1.79
E=1.12



D2
A (anode, the non-band end) =0
K (cathode, the banded end) =0

D3
A =0
K =0

c4
Q2 emitter =1.5
ground= 0

Let me know if additional info/readings are required. Many thanks.

[anchovie]

should I get a 0 reading from the ground strip?

Battery= 8.42

V - battery= 7.07
V + battery=8.16

The ground strip should read 0V, because you are measuring voltages with respect to ground. Which is why those battery readings don't make sense...

What points are you referring to by "V - battery" and "V + battery"? The battery terminals? If your battery measures 8.42V then you should see 8.somethingV on the + terminal when loaded by the circuit and 0V on the - terminal, as it should be connected to ground.

[entropicp]

Thanks for the reply anchovie.

Code Select Expand

The ground strip should read 0V, because you are measuring voltages with respect to ground. Which is why those battery readings don't make sense...

That makes sense and clears things up a bit.

What points are you referring to by "V - battery" and "V + battery"? The battery terminals? If your battery measures 8.42V then you should see 8.somethingV on the + terminal when loaded by the circuit and 0V on the - terminal, as it should be connected to ground.

Yeah, I mean the battery terminals. I left the black probe on the ground lug of the output jack and tested red from where the battery terminals connect to the middle lug of the stereo input jack (-) and the 9v AC socket (+). Just rechecked twice and got a 0V reading on the battery - . I must have screwed up the first time.

[entropicp]

Okay, I managed to get the thing to sort of work finally! I had somehow twisted the base and collector wires around and had put the collecter to where the base was supposed to be and vice versa.

However it sounds really brittle, harsh and buzzy, especially on the lower strings and seems to be cutting out all of the lower end. Would this have something to do with the Ge transistor I used (2N1303) or the resistor substitutions I made? I also used film caps instead of ceramic for C3 and C6. Would that make a difference?

Also when I set the harmonics knob (P1 100K) to null the volume cuts out completely and there is some volume loss until it reaches about  a quarter of the way up. Is that normal for this pedal?

[anchovie]

The Harmonics control simply sets the input signal level, so there will be no sound if it's set fully anti-clockwise. It does the same job as the volume knob on a guitar.

My HP sounded rubbish and severely mis-biased with the stock resistor values. I tried tweaking some of them but in the end got bored and decided that running a LPB into a Fuzzrite sounded far more awesome.

[entropicp]

The Harmonics control simply sets the input signal level, so there will be no sound if it's set fully anti-clockwise. It does the same job as the volume knob on a guitar.

So basically the knobs work like preamp and post gain? I'll try messing around with different value resistors then and see if that changes anything. I am definetely underwhelmed so far. Not exactly the Shellac tone I was hoping for.

[edvard]

Yes, you could re-label the knobs "Fuzz" and "Volume" and it would be the same thing.

Your tone problems may or may not have anything to do with your transistor.
I built one with a 2SB175 in Q1 and it sounded pretty cool but, once again, not the Shellac tone you're looking for.
You might start by using a smaller cap on the input, like a .01 or .02 uf.
Put trimmer pots on the collectors and see if you can bias it into the territory you're looking for.

Use a pick-sized piece of sheet metal for a plectrum... 
(just kidding... sorta...)

From all I've been able to glean, Steve Albini's tone has a lot to do with the rig he uses besides the Percolator.
Aluminum neck guitar w/soapbar single coils played with a copper pick, noise gate, solid state preamp, generic power amp into homemade 4x12 cabs, it's all in there. Plus I'm pretty sure he doesn't use the Percolator all the time, it's just one flavor in the whole recipe.

Meanwhile, focus on what you can get out of what you have. I've found the Perc to be quite a capable pedal for all kinds of noises, not just Albini's tone.
Like a thousand circuits after it, it can be tweaked and modded to your desired flavor of awesome.

[entropicp]

Naw, I don't need all that stuff to sound like Shellac... I just have to wrap my guitar strap around my waist and I'm a plane.

I'll try changing the input cap (C1 right?). How would using 820K instead of 680k for R1 affect the sound theoretically? I don't have any 750k resistors. I also have to get some trimmers to try biasing them via the collectors. It's a pretty tight stripboard layout so it might be a bit tricky to fit it all into the enclosure but I'll have a go at it.

[edvard]

Naw, I don't need all that stuff to sound like Shellac... I just have to wrap my guitar strap around my waist and I'm a plane.

It's my art...

I'll try changing the input cap (C1 right?).

If you're following George Giblet's schematic, the input cap would be C2 (.05uf). C1 is the 100pf one.
It sounds like you're losing low end as it is, so maybe leave that one alone.

How would using 820K instead of 680k for R1 affect the sound theoretically? I don't have any 750k resistors. I also have to get some trimmers to try biasing them via the collectors. It's a pretty tight stripboard layout so it might be a bit tricky to fit it all into the enclosure but I'll have a go at it.

Changing that value would change the input bias point as well as varying the negative feedback.
So (again, theoretically...) higher value = less negative feedback = higher gain (and vice versa), BUT it will also affect the collector bias point.
That's where the collector trimmers come in.
Adjusting collector resistance will affect gain a little, but based on your comments, it sounds like your build is a bit off-bias anyways, so set it where you lose the buzziness and you should be good.
'Buzzy' is what bad bias sounds like and for many fuzz effects it's totally on purpose but I don't think that is what the Percolator is supposed to be doing.
I would do all this on a breadboard first, and measure the trimmer resistances so you can replace them with permanent resistors when you do your stripboard layout.

Hope this helps...

[entropicp]

Thanks Edvard. Don't think I need higher gain. I'm ordering some trimmers and I'll try breadboarding it when I get them. I think I have enough parts to redo it except the Ge transistor. I'll get some transistor sockets too so I can see what it would be like with a different transistor as well. I'll try to figure out how to do the mod where you can use an spdt switch to by pass the clipping section too.

[edvard]

If you don't need higher gain, then a 680k in place of the 750k shouldn't be a problem.
If it affects the sound too much, try different values for Q2's collector.

No need for SPDT. SPST should work because both diodes tie to the same point in the circuit.
Try silicon diodes too. It adds a little more volume and 'aggressiveness' to the sound without going into raw transistor drive territory.

Right now I'm working up schematics for some mods I've tried on my Percolator build.
I'll post as soon as I get everything finished.

[DougH]

All bets are off if you don't use the 2n404a Ge transistor as specified. The biasing on this is very weird and unfortunately device-dependent. Tim E. has a version of this with COTS Si transistors with the biasing networks adjusted appropriately. Otherwise, you should be able to find the 2n404a's and 2n3565's on Ebay.

[edvard]

All bets are off if you don't use the 2n404a Ge transistor as specified.

I totally agree. If you want to faithfully reproduce the sound, you gotta have the original parts and set voltages correctly.

The biasing on this is very weird and unfortunately device-dependent. Tim E. has a version of this with COTS Si transistors with the biasing networks adjusted appropriately.

Are you talking about the Harmonic Jerkulator? It was ok, I didn't like it much though.

That said, I contend that the HP is a wonderful design that can be coaxed to produce some quite charming noises with a few tweaks and part swaps.
I'm posting now...

[entropicp]

All bets are off if you don't use the 2n404a Ge transistor as specified. The biasing on this is very weird and unfortunately device-dependent. Tim E. has a version of this with COTS Si transistors with the biasing networks adjusted appropriately. Otherwise, you should be able to find the 2n404a's and 2n3565's on Ebay.

That is very useful information.The George Giblet schematic said 'et. al' for the germanium transistor. Guess that wasn't entirely true. I had looked on UK Ebay for the 2n404a's before starting on the project and there weren't any available. The only ones I saw were from the States and it got really expensive with shipping. The best I found was AMZ FX webpage which had 2 for $12. So i'll order from them. I used 2N2484 that were specified in the schematic but I  got some 2N3565's just to be sure.

Right now I'm working up schematics for some mods I've tried on my Percolator build.
I'll post as soon as I get everything finished.

Many thanks. That would be awesome. I saw the diode clipping switch mod  here and I thought it added some really nice variation:

http://www.youtube.com/watch?v=m1Y6HdU7PlY