"Mr Crunch" booster/OD UPDATED SCHEM

[MartyMart]

http://aronnelson.com/gallery/Martys-layouts-and-photos/Mr_Crunch_V_2

Here's something very useful/simple to build.
Provides a nice boost, and a touch of crunch with a strat or a bit more
from a HB equiped gtr.
You can add a gain pot where the 56k resistor is if required, but rolling
back gtr vol and/or using another pedal to boost the front end will do.
Uses every day components, has a nice bright tone and the 1N4001/1N4148
diode combo works very well here.
As alway's mod to your liking - LESS IS MORE !!

Marty.

[Mark Hammer]

No offense but why would you not provide any floating ground?  Does it work in spite of that absence?

[vanhansen]

Looks nice, Marty.  I love simple circuits.   

[MartyMart]

No offense but why would you not provide any floating ground?  Does it work in spite of that absence?

I'm not following you, unless I've just missed something off the schem ??
Ground's look fine to me and it's working great ....

marty.

[Mark Hammer]

No offense but why would you not provide any floating ground?  Does it work in spite of that absence?

I'm not following you, unless I've just missed something off the schem ??
Ground's look fine to me and it's working great ....

marty.

Not that it is any sort of weakness or sin (far from it), but the circuit is obviously derived from the elements of the MXR Distortion+.  Normally, in such a circuit, the audio AC would sit on top of a +4.5V virtual ground derived by dividing the 9V+ supply in half with a pair of equal-value resistors.  There is no evidence of such a thing in what you posted.  The battery goes to the V+ terminal on the chip and the '+' input of the TL071 has no biasing resistor to Vref whatsoever.

I will openly confess to:
a) neglecting to include biasing schemes in some of my own posted circuits, and
b) not fully comprehending the mysterious ways of op-amps that claim to  be able to work off single-ended supplies.

So, what is it I'm missing here?  Or is it the case that what you are depending on here is what happens when a proper Vref is NOT used?  (again, not that that's a bad thing)  I fell like I'm staring at a picture of a horse with one front leg, one hind leg, and I don't understand how it is that he's still standing up.  Could be my weakness, the horse's strengths, or a trick picture.

[MartyMart]

OK, I see .... well it's not directly derived from anything, though I'll admit it's "a bit"
like a Dist + arrangement
I just decided to try this without any Vref for this opamp and ... well it seems to work
quite well !!
It boosts  ... it crunches .... "It's alive !"
The horse could well fall over in time, tell me if it's rediculous.... but it sounds
damm fine

Marty.

[vanhansen]

Funny that you mention the Dist + because that's the first thing that came to my mind when I look at the schem but I was thinking "Distortion+ Lite".  

In the early stages of my Slapshot, I too left out the 4.5v virtual ground and it still worked fine.  Only when I got further in to it and read more on op-amps did I decide on putting it in.  I gotta try this one out.

[Mark Hammer]

There was a thread sometime within the last calendar year involving the virtues of a variable Vref.  So, for instance, V+ goes to a 2k2 resistor, which goes to the outside lug of a 25k linear pot, which goes to another 2k2 resistor.  The Vref now comes from the wiper of the pot.  What you have there is something which can both replicate the traditional centred ground (exactly half the supply voltage), or something much closer to what you have shown.  Just remember to stick a polarized cap (10uf or bigger) from the wiper to ground.

Knowing your predilection for the interesting and harmonically quirky, give it a try.  It maybe a useful additional to Mr. Crunch.

[MartyMart]

Thanks mark .... I "think" that was a compliment !!
I'll give it a try, "Mr Crunch" is now safely in a very small enclosure
( about 2/3 of a 1590B ) so its too tight for any more parts and he's
all "sealed up" for now.
9v dc  jack only, no battery in there .....
Are you talking "voltage starvation" for pin 7 or adding this pre- "+ input"
as a bias ?

Marty.

[stm]

The bottom line here is that one of the golden rules of using OpAmps is in violation: "every input must have a DC impedance path to ground for proper biasing".

This doesn't mean necessarily a resistor directly connecting each input to ground. For instance, pin 2 (- input) has a DC path to ground through the 56k resistor and the output of the OpAmp.  On the other hand, pin 3 (+ input) has no way of establishing a DC current flow to ground since it is blocked by the 22nF input capacitor, so it is biased "magically".  The TL071 is a JFET input device and requires currents in the order of the pA (picoampere, i.e. 1E-12) to properly bias, so this explains why it is forgiving your lack of bias on pin 3. Even a small leakage current will do the job.

Nevertheless, I would bet the output of the OpAmp is not exactly at Vcc/2 (or 4.5V), but floating somewhere in between 0 and 9 Vcc. In your case this makes a good sound, so the point here is to measure the ACTUAL output pin 6 voltage with respect to ground to gain understanding of what's making it sound good.

After that, you should set a biasing scheme as per Mark's suggestion that allows to set the measured bias point (or any other of your liking) to avoid trouble in the future. Which trouble?  Well, first of all, a different-but-similar chip like an LF351 might not bias at the same point and could sound "different", if it works at all.  Secondly, a bipolar technology OpAmp will 99% surely not work at all, leaving out the LM741, 5558's, and NE5234. And third, there is no guarantee that other TL071's (made by different manufacturers and/or belonging to different production batches and/or chip revisions) would behave identically and consistently in the circuit.

In summary, there is a potential of causing headache to the DIY experimenters that follow your circuit, and after all the solution is simple.

[MartyMart]

I agree, but I tried several chips and got a similar result !!

"Rules" are there to be broken 

marty.

[puretube]

try all those chips again,
each with different caps,
and  each combo thereof in different layouts
(Vero/PCB with large groundplanes/small trace-distances,
no groundplanes/large distances),
at various weather circumstances (barometric & humidity-mojo)... 

[MartyMart]

File has been updated to include Vref bias to opamp
Allowing various single opamps to be used.
It was not stable enough as was and could fail ( it did!! )
390k/470k/560k/1M tested and work as bias "r" with only
slight voltage differences, so use what you have !

Marty.

[stm]

Seeing is believing!

Good you added the bias network to the circuit. Anyways the question regarding the assimmetry on the bias voltage still remains open. A good way of experimenting with it is to place a 25k trimpot in series with ONE of the 10k resistors that form the bias network. In this way you can vary bias from 1/2Vcc to near 2V away from one of the supply rails. This will affect balance bewteen odd and even harmonics just as it is done in the XXL from Tech 21.

The reduction of the 1Mohm to 390k resistor that goes from pin 3 (+ input) to the bias network currently affects input level somewhat due to the voltage divider formed with the 10k series resistor. In this respect you may place said 10k resistor betwee pin 3 and the hi impedance resistor, thus avoiding this divider while keeping the OpAmp input protected.

Another thing about the input impedance level is that usually it is believed that the higher the input impedance the better, with the tendency of having 1Mohm as a "good" value.  In case of non-distorting effects (chorus, equalizer, tremolo, phaser, etc) I agree that a high input impedance is desired to preserve the highs.  On the other hand, when the effect is a distortion/fuzz/overdrive pedal to be put just after the guitar, then a low-valued impedance WILL cut some highs and may improve tone by avoiding harshness. Take a look at the input impedance of the Range Master or the Fuzz Face, for instance; way below 100k!  Some effects even place a 100pF or higher cap to ground at the very input to achieve this...  Of course you can always rolloff your guitar tone as well.

[Mark Hammer]

Well now that it has been improved, perhaps you should "promote" it up a rank: The Captain Crunch!   (Myyyyyy that's shrill)

[vanhansen]

And make sure it has a couple of red LEDs labeled as Crunch Berries.   

[MartyMart]

Thanks chaps ... I always "believed" you,  just thought I'd stumbled on
something that worked .... it did for a few hours, then would "fail" on
power up !
I ended up sticking a 560k from Vref to pin3 and I'm happy with the tone
It meant squeezing a tiny "daughter board" into the case, about 1.5cm square !

Marty.

[petemoore]

As alway's mod to your liking - LESS IS MORE !!
 Thanks Marty Mart, I will do that !!!
 
 

[petemoore]

  Only problem I'm having is the diodes to ground.
  I'm using an LM741, boosts pretty good, biases nicely, but unless I disconnect the diodes to ground the signal dies.
  I tried the diode combo, then with some new 1n4148's, same thing, as soon as I touch them to the 10k coming from the output, no signal...doesn't make sense.

[stm]

If you look at the schematic it will make sense... The 4.7uF output capacitor should be placed between pin 6 (output) and the series 10k resistor instead.  This will fix the problem.

Currently, the 1N4148 diode won't let the node formed by the 10k resistor, 1nF capacitor and diodes go higher than 0.6V, and since this OpAmp's output won't go below aprox 1.5V, said diode is always in conduction, thus clamping the output. I didn't notice this before.