LPB2 distorting a little

[erick4x4]

I just finished throwing this together on a blank pcb board.

Seems to work fine, goes from silent to really loud. Seems to keep the tone fine.

With humbuckers I can get it to distort. Is this normal? I can't tell if it is distortion or misbias, its subtle so it could be either.

Voltage is 5v on the tranny. Supply is 8.9v, so it is a little more than half. Before I go putting a trim in there, could that .5v be it?

Also, I didn't have 43k and 470k for the voltage divider so I just used 47k and 470k, would that be the cause?

Thoughts?

[PerroGrande]

5v on the collector with quiet inputs is about right for the default circuit.  Can you post the other voltages (E and B, quiescent)?

I don't know how "hot" the output from your humbuckers is or how aggressively you're playing.  It is conceivable that you're producing enough output from the coils to drive the unit into clipping.

I'm a little confused about your biasing question -- the original circuit calls for a 43K and a 430K resistor.  Did you do 47/430 or 47/470? 

These values *will* change the operating point a bit.  47/470 comes noticeably closer to the original circuit behavior than 47/430 would.

[erick4x4]

Sorry, yes I used the 47k and 470k instead of the 43k and 430K.

It seems like I am just pushing it into clipping, it is just a hard clipping.

I have been reading about changing to the Orman values to increase the input impendence, which might fix the highs just a touch.

Jack lists 830k and 100k, but that isn't 10:1. Thoughts? Why not 830k and 82k or 1M and 100k?

I also read about changing the 390 ohm to 1k. What does that do?

I think I might lower the input cap (or make it switchable), this would make it a better boost to saturate my distortion, and cutting a few lows, would probably stop the clipping anyway.

Thoughts on the voltage divider, or the 390 to 1k change?

[GibsonGM]

I've done the 390 to 1K, and it increased the "fidelity" of the pedal (not that it's very hi fi to begin with, lol!).  It also decreased my gain, being on the collector end of the transistor.  I'm either going to put a 1k trim in there to dial up a sweet spot, or just go back to the 1K...I don't feel I'm getting enough boost to kick it up over my band (I use the LPB for lead boost). 
Playing with the input/output cap will get you more or less bass, but might not remove the 'additional distortion'...this circuit is a single transistor, and is meant to be driven a bit - it will color your sound, that's it's nature.  Putting in or in/out on a switch is a good idea if you want a treble booster, though!  ;o)

[wampcat1]

Honestly, I don't think it's all that uncommon to run it into clipping if your input signal is a bit hot. I would run a volume type control in front of it and use that to control the clipping.
bw

[RDV]

How do we know the LPB is distorting and not the frontend of whatever you're running it into?

RDV

[PerroGrande]

Well, lets take a quick look at some of these changes.  Remember, gain and clipping prevention are often working in opposition with one another. 

The 100/820 combo that Jack suggests is good, especially in combination with the 1K mod.  The bias points still look okay and you'll get a bit more headroom.  It also works to increase the input impedance a fair amount, so this, too, is a good thing.   The sacrifice in this case is some gain. Increasing the emitter resistor does this, as the gain is given by the ratio of the collector resistor to the emitter resistor.  So you go from a voltage gain of around (-)25 to (-)10  (- because the stage inverts)... With the reduced gain comes some additional headroom. 

You may be able to squeak a little more gain and headroom out of the circuit by using 120K/820K in the biasing network, a 1K emitter resistor, and an 11K collector resistor.  In some tests, these values showed some (expected) clipping on signals with an amplitude of around 400mV (800mV peak-to-peak). (The stock circuit starts to clip at around 200mV, give or take.)  With this combo, I've raised the bias point on the base to around 1.1v, thus giving the emitter a little more room to wiggle.  The 11K resistor nudges the collector a little bit closer to the midpoint of the supply (4.5v).  The quiescent current rises by about 40 microamps, which shouldn't cause any trouble.  The biasing network is still stiff relative to the base current, so we're okay there.  The penalty, again, is gain.

Either way, you're still dealing with a single-transistor gain stage and the fundamental limitations involved therewith.