Single Transistor Boost Questions

[Joe Hart]

Hello, all! I built a single transistor boost and came across some interesting (I think) results. It's VERY heavy sounding with a strange octave-like/sputtery sound. I can't really describe it. It sounds great, not like a "yucky" Fuzz Face. It sustains great and trails off naturally -- no gating. It sounds almost exactly like the guitar tone in "Sweet Leaf" by Black Sabbath. Here's the schematic. I added the input cap blend just for some versatility, the original input cap was just .22uf. Also, it sounds more like that tone with a battery at about 7.2V (so I will be adding a battery sag to the circuit).



Now, my question is: did I simply design a misbiased, full-band boost and run it with a low battery or is there something else going on to create the "Master of Reality" tone? Thanks!
-Joe Hart

[Joe Hart]

And if it helps trying to figure out what I did, the transistor has an hfe of 158. And I'm quite sure that it's a silicon transistor.
-Joe Hart

[alanlan]

Not sure if it's mis-biased but if it sounds great then well done and don't worry too much about it 

Looking at the cct you've got only a small voltage at the base, maybe about 0.8V if you're lucky and therefore 0.1 to 0.2V across the emitter resistor (read all voltages as -ve of course and assuming pnp tranny) which amounts to 1 to 2mA of steady state current, but this would have the transistor in saturation since it would pull the collector close to the emitter due to the drop across the collector resistor.  So, heavy clipping or halfish wave operation maybe, not sure.  Perhaps a happy accident?

[.Mike]

[Joe Hart]

Not sure if it's mis-biased but if it sounds great then well done and don't worry too much about it 

Looking at the cct you've got only a small voltage at the base, maybe about 0.8V if you're lucky and therefore 0.1 to 0.2V across the emitter resistor (read all voltages as -ve of course and assuming pnp tranny) which amounts to 1 to 2mA of steady state current, but this would have the transistor in saturation since it would pull the collector close to the emitter due to the drop across the collector resistor.  So, heavy clipping or halfish wave operation maybe, not sure.  Perhaps a happy accident?

I'm not worried about it, just curious if I did something special and/or if it can be replicated. And it is definitely a happy accident. It sounds very cool and I had no idea what I was doing. And still have no idea, hence the post seeking an answer to what I did!

And I think it's more than just heavy clipping, because I have boxes that clip VERY heavily and they don't sound like this. Hmmm.
-Joe Hart

[Joe Hart]

Mike, thank you so much for what appears to be something very helpful!?!? I say "appears to be" because I'm not sure what I'm looking at!!

But first, does it matter that the cap to ground in my circuit is 50uf and you have 100uf? I had tried a few different values and found no audible difference. In fact, even taking the cap out was subtle at best! So, I'm guessing that 50uf or 100uf don't have much effect on the end result.

That being said, I recognize the schematic, of course. Does the simulated oscilloscope show that the waveform is severely clipped asymmetrically (the blue wave)? And I have no idea what the graphs on the bottom half mean.

Thank you very much, though, for the time you spent doing this!! I am very interested in figuring out what I did!!
-Joe Hart

[.Mike]

Basically, it is just a sim of the circuit. It is being fed a 100mV p-p signal @ 1khz. V- is -7.2v. Transistor is a generic 2n3906. Volume pot set at 50% (linear).

The raw signal is being fed to the first channel of the scope (yellow). The effected signal is the second channel (blue). Yeah, looks like an inverting asymmetrical clipper to my uneducated eye.

The bottom chart is the frequency response. The circuit is fed a wave from 1Hz to 100Khz, and the response is plotted. So at 10Hz, it looks like the signal is boosted by 10dB, at 100Hz it is boosted 20dB, and then pretty much flat all the way up past 100kHz. There's apparently a way to run a distortion analysis, which would probably be much more helpful, but I haven't figured out how to work it yet.

Here it is w/ the 50uF:



I did it using a trial of Multisim and a virtual PC. 

Mike

[Tony Forestiere]

I am curious too. Joe's schematic looks like a negative supply, while the scheme in the simulation appears to to be a positive supply.
(Did I get the new nomenclature right?)

[.Mike]

Nope, it's negative supply. Look closely. There's a minus sign before the voltage.

It was covered by a label in the first image. I moved it in the second so it is clearly visible.

I sim dual supply opamp circuits doing it this same way all the time, and I have taken virtual measurements to confirm it puts out a simulated negative voltage.

Mike

[Tony Forestiere]

Duh..I missed that I have a tiny monitor and missed the tiny negative symbol. (I also have a tiny brain, feet. nose, hands...um. Let's stop while I can.)
I do have some 3906's. I would probably play with this over the weekend. The output waveforms and shelving are definitely interesting. Thanks for the sim.

[EATyourGuitar]

so I can do this with germanium if I get it 0.1v above the turn on? like 0.8v at the base for silicon or 0.4v for germanium.

[Joe Hart]

I am curious too. Joe's schematic looks like a negative supply, while the scheme in the simulation appears to to be a positive supply.
(Did I get the new nomenclature right?)

That was in another thread about possibly shorting out "positive ground" and "negative ground" pedals by having them touch (with the enclosures being part of each pedal's grounding scheme), and R.G. mentioned that he didn't like those terms. His reasoning being that "ground" is always 0V, so it is neither negative nor positive. So I'm going with "negative supply" and "positive supply" from now on (when I think of it!).
-Joe Hart

[Joe Hart]

Here it is w/ the 50uF:

So my ear was right. It seems to make no difference at all.
-Joe Hart

[Joe Hart]

The raw signal is being fed to the first channel of the scope (yellow). The effected signal is the second channel (blue). Yeah, looks like an inverting asymmetrical clipper to my uneducated eye.

The bottom chart is the frequency response. The circuit is fed a wave from 1Hz to 100Khz, and the response is plotted. So at 10Hz, it looks like the signal is boosted by 10dB, at 100Hz it is boosted 20dB, and then pretty much flat all the way up past 100kHz.

So it's pretty much a full-range booster with inverting asymmetrical clipping? So could that be done a number of other ways, right? Although I am totally happy with the sound, I didn't really "discover" anything new?

And I still don't really understand what the circuit is doing, or do I? I think what's happening is since there's such a small voltage at the base, the transistor is freaking out and totally distorting the signal? Is that about right?
-Joe Hart

[Tony Forestiere]

That was in another thread about possibly shorting out "positive ground" and "negative ground" pedals by having them touch (with the enclosures being part of each pedal's grounding scheme), and R.G. mentioned that he didn't like those terms. His reasoning being that "ground" is always 0V, so it is neither negative nor positive. So I'm going with "negative supply" and "positive supply" from now on (when I think of it!).
-Joe Hart
[/quote]
Yup, I agree with RG. I do prefer referencing the supply polarity because Ground is Ground is Ground. I would hate to have my commode flush counter-clockwise if I touched my pedals together.  

I see the VCC at -7.2 volts, and your comments of a sag control. I wonder what the sim would show at higher and lower voltages as the voltage increased or decreased from that level?

EDIT: I always botch up trying to selectively edit quotes. I might get it right someday.

[Joe Hart]

I do have some 3906's. I would probably play with this over the weekend. The output waveforms and shelving are definitely interesting. Thanks for the sim.

Let me know what happens. I tried some other transistors, but they either didn't work at all or didn't sound close to the one I ended up using. I think I'm understanding what's going on with this circuit and maybe what happened was like randomly swapping transistors in a Fuzz Face. If you do that, you could/would end up with a great Fuzz Face sooner or later, but you'd be much better off getting the correct gains, then tweaking the resistors to get the correct voltages. Maybe my circuit just needs any transistor with similar hfe (158) and the "correct" voltages. But let me know what you come up with. Thanks!
-Joe Hart

[Joe Hart]

I see the VCC at -7.2 volts, and your comments of a sag control. I wonder what the sim would show at higher and lower voltages as the voltage increased or decreased from that level?

Maybe Mike will have nothing to do but keep sim-ing this circuit! :-)

When I used a new battery, it sounded "nicer" -- read that as less heavy and less evil and less cool!!! Maybe it brought the voltages on the transistor closer to "ideal" and I got less of the asymmetrical clipping? It sounded good, but it just didn't have the "ratty, torn speaker" kind of sound that the lower voltage did.
-Joe Hart

[LucifersTrip]

Maybe my circuit just needs any transistor with similar hfe (158) and the "correct" voltages. But let me know what you come up with. Thanks!
-Joe Hart

what are the actual ebc voltages with a 7.2v and 9v battery...and what transistor # did you use?

thanx

[Joe Hart]

All I have on hand is a new battery, but here are the voltages (if I measured correctly and have the correct pins labled):

Batt: 9.53V
E: 28.2mV
C: 6.55V
B: .643V

Does this explain anything to anyone??
-Joe Hart

[Joe Hart]

Actually, this looks pretty close to what Alanlan estimated. I don't know enough about electronic theory to understand what it means, though!
-Joe Hart