Let's talk about the Jordan Bosstone. An odd one...

[John Lyons]

Interesting, thanks for the rundown. I'll have to think on that...

John

[R O Tiree]

In simulation, I got oscillations, and taking C2 out seemed to stop it. The voltage at that junction swings up and down quite dramatically without the cap; with the cap, the voltage only goes up and down half a volt or so, which holds the base of Q1 higher than it "should" be, so the gain goes up dramatically. However, with the cap oscillation can take place for the reasons earthtonesaudio mentioned above.

Next, if we remove Q2 for a second, then the north end of C4 would bounce up and down between 4.5V or so and 8.5V. Put Q2 back in and we find that the voltage drop across R6 (refer to the schem at the start of the thread) is constrained to be about 0.6V or so. This means that, as Q1 conducts heavily, the collector falls to almost 0V. That collector is connected directly to Q2's base, so Q2's emitter drops to about 0.7V. On the up-swing, Q1's collector shoots up to about 8.5V. This happens with signals greater than about 30mV. With smaller signals, Q2's base and emitter continue to just track up and down 0.6V apart in time with the voltage at Q1's collector and don't really have any effect apart from that.

C4 is not simply there to de-couple the output. If you take C4 out, the circuit simply refuses to output any signal at all. If you simply replace C4 with a wire, you now have 9V ->18k (R2 in the original schem) -> diode ->GND, so the emitter of Q2 and the "north" ends of C3, R3 and R6 are all held firmly at about 0.6V, so absolutely nothing happens. Putting C4 in there means that the rest of the circuit can swing up and down to its heart's content. Take the diodes out for a second, so we can think about what happens at the right hand end of C4. As the voltage at Q2's emitter swings high, C4 charges up. Then the Q2E voltage takes a nose-dive. That charge has to go somewhere, so the voltage at the right hand end of C4 goes below zero. As stated, the 100k pot is not exactly a short-circuit, so we get a weird, but quite smooth waveform at the output. Put the diodes back in and clipping occurs any time the output exceeds +0.6V or drops below -0.6V. So, outside those bounds we get hard limiting of the output signal at + or - 0.6V as the diodes conduct heavily. Inside those bounds, the current increasingly has to flow through the 100k pot to ground. This is where the very weird waveforms come from with this circuit.

[brett]

Hi
Yes. And that further explains why I haven't had the oscillations.  I replaced that cheapo and outdated 2x560k and a small cap with a more modern 1M and a 22uF decoupling cap.

Yep, the waveforms can be really weird.  I *think* much of that is because the output impedance is different on the + and - sides of the signal. 
cheers

[John Lyons]

Just putting the schem on this page for easier reference.



Thanks for the simulation R O!
I'll have to digest this.

john

[Sir H C]

Nashville version:





Alhambra version:

[edvard]

Hey now!
I dig the ceramic cap "pancake stack" thing they have going.
Perhaps that's to induce some parasitic capacitance within the stack to cut the feedback?
Or just to make room?

Hmmm....

It also appears that the number of capacitors gives away the variety:
3 caps = Nashville
2 caps = Alhambra

Thanks for the shots Chris!

[R O Tiree]

If my soldering looked like that, I'd be ashamed of myself...

OK, done a bit more simulation and it appears that placing a 22k resistor between the north end of C2 and the junction of R3/R4 damps out the oscillations without affecting the response unduly.

I also simulated a guitar pickup at the input complete with Vol and Tone, rather than a simple ideal AC signal source, because it occured to me that the resonant circuit that they make up might be the culprit if this was the first or the only pedal in the chain. Didn't make any difference one way or the other. Without the damping resistor it oscillated, with it, it didn't.

I'd be interested to see if someone who has a "howler" could try this damping resistor thing and see if it actually works.

[brett]

Hi

it appears that placing a 22k resistor between the north end of C2 and the junction of R3/R4 damps out the oscillations without affecting the response unduly

Although that will reduce or eliminate the oscillations, it also minimizes the power supply filtering.  C2 is an old-school troublemaker that does nothing tonally. Ditch it and connect a 10/22/47 uF cap from V+ to ground.
cheers

[earthtonesaudio]

Hi

it appears that placing a 22k resistor between the north end of C2 and the junction of R3/R4 damps out the oscillations without affecting the response unduly

Although that will reduce or eliminate the oscillations, it also minimizes the power supply filtering.  C2 is an old-school troublemaker that does nothing tonally. Ditch it and connect a 10/22/47 uF cap from V+ to ground.
cheers

C2 cuts negative feedback from the collector to the base, which increases gain above the 3dB rolloff created by R4/C2 (12Hz) and also increases the input impedance (though not significantly because R5 dominates).  It also filters power supply noise.

Adding the 22k in series with C2 make a voltage divider in conjunction with R4 so that the gain is (neglecting active load details) R4/22k.

[R O Tiree]

The trouble with the 1M/22µF mod is that the gain drops unacceptably. C2 might not affect the tone, but it sure affects the gain in spades.

Adding the 22k in series with C2 does, indeed, reduce the gain but, as I said earlier, "not unduly". I've done some more tweaking and 4k7 might well be enough to damp things out and not affect the gain at Q1's collector more than a hair.

C2 is a voltage divider to AC signals in conjunction with R4 in any case, so it's just a matter of the degree to which the extra resistor affects things. Looking at the numbers, the %age gain change is small in the range 82Hz - 1.3kHz (low E fundamental to 24th fret on high E fundamental). Above that frequency, C2's reactance becomes quite small, so the new resistor begins to dominate. At 82 Hz, C2's reactance is 88k. At 400Hz, 18k, at 1kHz, 7.2k and 2kHz it's 3.6k. It's hard to say what effect this limit on the higher harmonic content's gain will have. Let's face it, the output waveform is totally FUBAR'd in comparison with the input waveform in any case so who can say what's normal and what isn't? With that 4k7 in there, a mere +/-1mV input signal @ 800Hz gives +/-625mV at Q1's collector. That's an insane gain of 625! Short the resistor out and you get a gain of 640. Take C2 out altogether and your gain drops to 120.

As to power supply filtering? This thing draws only 350µA or so. I'd run it off a battery, I think. TBH, a standard 9V battery will last months of "normal" use at that current draw. If you really wanted to run it off a PSU, why not a 10µF || 100nF (or similar) from +9V to GND as you suggest, Brett? I agree totally.

But why hamstring this insane pedal by overly limiting its gain (taking out C2) vs giving it just a little more damping than was originally designed into it and otherwise letting it do its funky thang?

[Rickofitall]

I'm on my 3rd Bosstone build and my results are somewhat confusing.

The 1st 2 were great distortion pedals but not fuzzy at all. This 3rd one is fuzzy but has no volume.  I swapped transistors, checked all my connections, made sure my soldering was good, checked my pots... even went as far as to build another board with all new parts... the problem remains.

I've tried a few different trannys for Q1 but have pretty much stuck to the 3906 for Q2.

I don't even know if I'm looking for an answer here or just maybe someone else who has built this thing wth less than fuzzy results.

[John Lyons]

Yes, the bosstone is more of a distortion/fuzz.
If you used all new parts and a new board then
it's got to be the schematic or something wrong
that you are doing when you build it.

There aren't any other variables unless the parts
are bad somehow.

john

[mac]

I was looking the factory schem, r2:15 and c3:50uf

IMHO it works better with lower gain transistors as brett noted, no more than 150.
And can be fine tune by tweaking the 150k resistor. Put a 250k trim to find the sweet spot.
I went even further, I used Ge transistors

mac

[rnfr]

i've been doing some interesting things with the bosstone lately as well.  messing around with Ge trannies and using a variation of mac's diode biasing scheme.  i'll post a schematic when i get home later along with some waveforms that i got.  i found a VERY odd quirk within this circuit as well.

[rnfr]

this is the schematic of what i have been messing with.  a silicon booster stage into a Ge bosstone sans diode clippers.  i tied a pot to the C and E of Q2 in order to be able to vary the shape of the wave and the phase of the output.  some cool sounds can be had this way.  you can see that on Q3 i used mac's diode biasing method to get the circuit to conduct properly.  the odd thing that i found was that when the collector of Q3 was lifted, i actually got more output out of the collector.  matter of fact it sounds really nice.  with the collector grounded, you get a waspy sixties fuzzrite kind of sound, lift the collector and you get a nice full fuzz.  the signal off of the emitter is less distorted, but sounds good none the less.  with the phase adjustment and the lift switch you can easily get 5 or 6 very different tones.  below are the wave forms.  can you guys tell me what is happening when C of Q3 is lifted and why there is a much stronger signal off of the collector?  i'll post some of the waveforms so you can see what i am talking about.



COLLECTOR GROUNDED, COLLECTOR OUTPUT.

COLLECTOR LIFTED, COLLECTOR OUTPUT.

COOLECTOR LIFTED, EMITTER OUTPUT.

COLLECTOR GROUNDED, EMITTER OUTPUT.


sorry for the crooked waves, i just got the scope and it needs adjusting.

[mac]

sorry for the crooked waves, i just got the scope and it needs adjusting.

at least you have  a scope!!!

mac

[rnfr]

yeah, i got it for 40 bucks cheap!  it's a lot of fun.  these were actually the first waveforms i got out of it.  any idea what's going on here?  is the third trannie just acting as a clipper when lifted??  maybe the 18K resistor is reducing the gain? that batman looking curve actually sounds really nice.  very harmonic percolator-ish.  the others aren't bad either.  the emitter outputs sound nice and overdrivey- very coll with clippers added too.

[mac]

it always seemed to me that the second transistor, or third in your schem, is acting like a clipping diode, and that the 18k to vcc makes it conduct a little.
a kind of bazz fuss + output clipping diodes.
but unlike the bazz fuss the base-emiter path is not between C and B but in series with another 18k. there is a post about something similar, a very simple circuit like a BF but it has a 10k and a diode and the output taken from the junction of both. it produces an octave up fx.  sorry i do not remember the link, i have to search for it.

mac

[kinski]

Okay, awesome thread. I'm about to start building a Bosstone. I'd like to order all the parts at once. I'd like to order the part(s) needed to rid of any oscillations, should they arise (as they seem to be common).

I was going to use this layout:



Any word on what I would need to add to this circuit to stop oscillations? Yes, I read the thread, and there is a ton of tech info that is a bit over my head. Is there a definitive answer on how to do this?

Much appreciated!

Thanks!

[kinski]

Or perhaps I should use this layout? Whats with the 2k7 resistor? and the 18k wired to Q2?