MXR Blue Box Level

[rove]

I just built an MXR blue box.  It seems to work just fine, I used the tonepad layout and added all parts shown on the schematic (afterwards I realized that the TP instructions suggest leaving some parts out). Put right before an orange squeezer (another recent build that I am pretty sure I prefer to the Ross, btw) the blue box sounds fantastic within the confines of its abilities, but on its own there is somewhat of a volume drop.  I fixed this in another recent build, the ampeg scrambler, by adding an LPB2 to the end of the circuit; the scrambler is particulary well suited to this as it has a clean blend which allows me to use it as only a clean boost, or as a crazy fuzz.  But I digress.
Anyways, in reading the harmony central reviews of this pedal a number of people mention the removal of a resistor "solving" the gain loss issue of the pedal, but I can't figure out which resistor may do this.  I am not opposed to adding another clean boost stage afterwards, just wondering if anyone has any other suggestions.
At this point I am trying to decide on an enclosure and I really can't see a reason for the Volume pot as I keep it all the way up and that doesn't seem to be quite unity...
thanks!
Paul

[dano12]

It is actually not a resitor, it is the C12 capacitor that takes care of the volume drop. Not sure how the commerical version's C12 matches up the to the Tonepad schematic, but the following should help you match it up.

http://www.noisefx.com/article/mxr_blue_box_mod

I did the c12 mod on mine and it took care of the problem, but it does change the character a little bit. So I put it on a switch. Hope that helps!

[Processaurus]

c12 is a simple lowpass filter, that gets rid of some high end hash.  Whether you want that tonality depends on personal taste, if it were my project and the volume (which is criminally low on this one) was bugging  me I'd stick on a transistor stage like the gain recovery on the big muff, its an LPB1 I hear.  Then you'd have plenty of output, rather than a smidgen more.

I wonder, has anyone tried using FETs instead of BJTs for the crude noise gate you see before the blend knob?  Would that have more output?

Hey, a great mod for the blue box is to look at tim escobedo's circuit snippets, at the PWM, you can insert the last section with the inverters into the blue box, to change the pulse width on the fuzz and/or octave down.  Really helps get that lonesome Atari tone, if you're inclined and have room for a couple knobs.  A switch to go between 1 and 2 octaves down is neccessary too.

[Mark Hammer]

There is a fixed cumulative gain from the two op-amp stages on the order of several thousand, so a properly functioning Bluebox will not produce levels lower than bypass unless the volume knob is turned wayyyyyy down.  The suggested capacitor mod is to bring some treble back instead of filtering off the sharp edges of the octave-down component.  This will, as does any treble boost, increase apparent volume, but the differences you describe stem from something more - pardon the pun - fundamental.  My guess is that the place to look initially is the orientation of the two transistors that form the gating circuit.  On the Tonerpad layout/schem, these are labelled as Q2/Q3.

[rove]

Thanks for the suggestions!  I haven't had time yet to go back over the circuit (I took it out of its proto box so I could try out the tripple fuzz I just built and also need to debug), but plan to do so today.  I will recheck the transistor orientation--I seem to have a lot of trouble getting this right, for some reason I become dyslexic when it comes to ebc, cbe, and so on.  On the Tonepad layout there are actually three caps and a resistor that FP suggests to leave out, they are designated C11, C12, C13, and R24.  The first 2 are 100uf electros and C13 is the one that will give back some treble according to y'all.  Currently I have all these components in place.  Interesting what Mark says about it not typically being a quiet effect as I read a lot of reviews and processaurus' post seems to suggest a typically low volume as well...
Not familiar with escobedo's PWM, but I will do a search.  Was trying to decide on the two octave down mod, worth it?
thanks again!

[Mark Hammer]

C11, C12, and C13 are on the power lines and have nothing to do with the treble.  My mistake.  I was in a rush and thought you were talking about C9, which IS intended to cut the treble and "round off" the sub-octave squarewave.  You can feel free to leave R24/C11-13 on the board.  They are optional, but are modern-day improvements to stabilize/smooth the power lines.

The Bluebox is certainly not in the same league as some of, say, Joe Davisson's ultra high-gain designs, but attaining levels noticeably higher than bypass is easy to do.  Given that Q2/Q3 serve as "gates" to allow the sub-octave and fuzz to be heard when the flip-flop is deemed to be adequately tracking the note, perhaps some folks have selected trannies for those positions that result in less signal exiting the Bluebox.

The stock Bluebox produces a waveform two octaves below that being played.  It can be modded to yield a note one octave below, instead, by tapping the 4013 (i.e., redirecting the flipflop side/end of R11) to pin 13 of the flip-flop instead of pin 1.  I gather this is what you mean.  It's a worthwhile mod IMHO, if only because so many folks play through speakers/amps that won't really handle 2 octaves down from the low notes terribly well.

As for transistor orientation, isn't it remarkable how many ways there are to get the pins of a 3-pin device wrong?

[rove]

As if to illustrate the point, wrongly orientated transistors were the "issue" with my tripple fuzz, which works fine now, near as I can figure, it can get pretty nasty sounding... I like.
And yes I meant the one octave down mod as opposed to the stock 2 octave, going to have to try this.
And maybe I will swap out Q2/Q3 and see what that gets me.  The other thing is that I have some sort of denial about my tendency to put transistors in wrong and only socket them every so often, like not on the blue box.  going to have to make a habit out putting those sockets in...
thanks all, I will post back with whatever I discover.

[rove]

Ok, so I put the blue box back together with the octave switch.  What I think I am noticing is not a lack of gain, but an abundance of compression.  Playing single notes (as this effect is "intended" to be used) I get a a nice boost, but if I play chords, it is definitely quieter than the bypassed signal.  the octave mod works great.  I haven't messed with any transistors yet, but I was thinking of subbing 5089s for the 5088s that I used...
I have tried both a tl072 and a jrc4558 and can't say I noticed much difference.
The tripple fuzz sounds awesome, btw.

[Processaurus]

There is a fixed cumulative gain from the two op-amp stages on the order of several thousand, so a properly functioning Bluebox will not produce levels lower than bypass unless the volume knob is turned wayyyyyy down. 

Definitely there is big gain and big signals at points in this circuit, but the bottleneck is how the two transistor noise gate works, I've looked at it on a scope when I spent time on a tricked out blue box a while ago, and was wondering where it lost most of its output.  There are squarewave signals that go to the rails on the bases of both transistors, and an envelope voltage of several volts on the collectors, but what comes out is surprisingly small, less than a volt P-P if I remember right, and is further attenuated by about 5x by the following passive panning circuit with the blend knob.  Its probably loud enough to match the volume of lower output guitars, but humbuckers or keyboards can be louder.

an envelope/ noise gate stage like is in the rocktave divider using analog switches rather than transistors might have more output with just as many parts.  Still wondering if the FETs would work better, I can't imagine why not, thinking about how boss uses them for electronic switching.  For the casual builder who isn't looking to redesign, though, a LPB1 or microamp tacked on the end would be a satisfactory way to go.