blue box build report

[vdm]

hey everyone,

just finished putting together a blue box, and for once i haven't been totally sucked into playing it...

probably my ugliest build so far, on a home made pcb, with very badly drilled holes, and a number of oversized resistors, and parralled/seriesed resistors cuz i didnt have the right values.

it currently has a 100k volume pot that doesnt work overly well, and the 50k blend pot is a trimmer at the moment, cuz i had no 50k pots around. there's no battery switching or bypass, i just threw it together to try and make it work, and it does.... it's just not quite how i imagined......

firstly i'd heard people talk of it having bad tracking, but this is so bad - it's like playing through a harmonizer with more than one note - it just sounds like a mess. though this effect was minimized with the neck pickup on my les paul, and also with any note above the e one octave above the bass string.

then there was the reason i built it. i wanted something really crazy and that would make some crazy noises, but at present all i can make it do is splutery bass that's rolled together, with no craziness at all... any ideas would be sweet though...

but after all that slagging it off, i do have to admit the distortion with no octave down mixed in is quite nice, very thick but bassy tone. with a tiny bit of bass added, it definately makes a nice deep tone, though i wish it had a little more treble in it, but if you're playing in a band with multiple guitarists with no bass, this'd be really good for rhythm.

next is the extreme oct(double octave)down tone, which by itself is quite cool to play bass lines up the octave, reaching down to the proper bass note frequencies. I found the middle range on the guitar was a bit too mellow and the fuzz wasn't as harsh, but if i used the treble setting on the guitar, it'd first go to one octave down, then go to the further one below that... so not overly useful, though im sure it could be used in a very musical way. this problem seemed to go away with the higher notes, which had much more attack.

and finally there's the half and half tone, probably the pick of the litter, i didn't believe it when i heard it at first but as soon as i started playing around the twelfth fret it screamed at me "MARIO BROS." im going to have to learn the full underground theme now... i used to play it on piano.....

i did change a few component values simply cuz i couldnt find the right ones in my parts bin, so i used a 110k for the 150k on the input, and a 68pf cap instead of the 300p (from the tonepad layout) though i doubt this would change the function of the unit at all, it seems to work just as described.

im still in the balance as to whether it'll get boxed up yet... but either way it's a bit of fun. i'd recommend it to anyone who loves mario, for the comical value alone... actually my band plays part of the theme as the intro to a song.... this could be interesting

sorry for the long post... but hey you didnt *have* to read it did you  :wink:

rawk on guys
trent

[Paul Perry (Frostwave)]

The tracking is going to be very dependant on the level of signal in.
Also, any harmonics will throw it. I would experiment wiht an EQ in front of it, also maybe some compression.

[Zero the hero]

If you want a crazy effect, try John Hollis' Crash Sync (maybe together with the blue box).
I've built the blue box with a different layout, and it really sucks... well, those who want to make MUCH noise with their guitar could appreciate this effect... not me, though...

[Mark Hammer]

If you built the "Clone Wars" version (like I have) there are errors in the layout.  Ambrose Chappell's schem and layout do not conform to the MXR factory schem in several ways (the Tonepad schem/layout DO).

Here are a few areas for tweaking that might bring out the best in the box.

As Paul notes, some sort of regular or standard input level will improve tracking.  A compressor can do that but there are other things you can do in the BB itself.  You will note from the Tonepad schem redraw that the output of IC1a not only goes to IC1b but to a fairly standard half-wave rectifier type envelope follower, formed by D2/R15/C8.  C8 sets attack/decay time range, and R15 tweaks actual decay time.  This envelope signal is not used for the same purposes an envelope filter might use it, but rather for gating Q2/Q3 so that sputtering is minimized.

The envelope follower really only has to make an "enough/not-enough" distinction in signal level for these gating purposes, so tinkering with the dynamics of the signal/box make little difference.  Bear in mind, it IS going to spit out a signal whose amplitude is square and fixed so attempting to preserve dynamics does you no earthly good.

In this spirit, consider some soft-clipping in the first stage to provide that compression.  I don't know what sort of envelope voltage the gating transistors "want" but you can consider sticking, say, a pair of 1N4001's or dual back-to-back pair of 1N914's, in series with maybe a 10k resistor or something.  You don't want to fuzz it up THAT much, just basically keep a consistent hot signal going to the envelope-control subcircuit.  Of course, since we have no idea what that optimum signal level should be, the gain of that first stage should be played with.  It is currently set to a gain of (470k+1k)/1k = 471.  The 300pf cap sets the rolloff at around 1100hz.  Consider raising the 470k resistor to 680k, and dropping the cap to 220pf (for a rolloff around 1060hz) or 180pf (for a rolloff around 1300hz).  If you stick a 5k pot in series with the 1k resistor tied to the inverting input, that will let you go from a gain of 681 down to 114.  This pot will be your "Sensitivity" pot.  You want to set it so that you don't have to pick too hard to turn the Q2/Q3 gates "on", but at the same time you aren't all that interested in octave division of the sound of your fingers sliding up the strings.

Although C3 provides some high-end rolloff, it isn't all that steep.  It's also worth noting that so much gain (especially with clipping/compressing diodes in there) is likely to result in additional harmonic content.  When that gets fed to IC1b with a gain of 101 ([1meg+10k]/10k) there is clearly even more risk for nasty fizz and artifacts.  Consequently, a cap in parallel with R6 isn't such a bad idea.  Less rude, but more manageable.  A value of 100pf in parallel with the 1meg resistor gets you a second shallow rolloff just under 1600hz., which is high enough not to impair the unit's ability to detect fundamentals higher up the fingerboard.

That one is feeding the envelope/gate circuit *enough* signal to gate appropriately, is separate from whether the time constants are well suited to your playing style and instrument or signal quality.  I'd suggest changing R15 from the stock 56k resistor to a 22k resistor in series with a 100k pot, yielding a resistance to ground of 22k to 122k.  This will let you reduce the decay time by about half in addition to approximately double.  That may, in tandem with the sensitivity control, permit the optimal gating action for your picking style and signal.

Finally, while there is a blend control, there is no means for adjusting the tone of the octave signal.  Once you've played through a Rocktave or similar unit and had the opportunity to change the octave-down tone, you realize what advantage there is to this.  Although C9 on the output tames some of the harshness of the flip-flop output, that treble bleed is applied to the entire signal - both same-freq and octave down.  A more useful approach would be to apply filtering to the octave signal.  Here the best place to intervene is at the junction of R17/R19.  This is the same junction  where the blend control taps the octave signal and varies how much of it is bled off to ground.  You can stick a cap to ground here.  A .001uf cap will get you a rolloff starting around 2.8khz.  A .0022uf cap will get you a rolloff just under 1.3khz, and 3300pf will take it down to about 860hz.  That may seem pretty dang low for a "treble cut" but remember the Bluebox churns out tones *two* octaves down from what you feed it, so even such a low rolloff will still yield 3rd and 4th harmonics of many of the suboctave notes.

There are a few ways you can implement this.  One is to stick a cap and variable resistor to ground so you can adjust how *much* rolloff at a given frequency.  The other is to use different cap values to adjust *where* the rolloff starts.  Of course, just like the input caps on pedals, if one lead from the cap is left "stranded", it will pop the moment you allow it to bleed off anywhere.  So, I would suggest something like the following.  Take a 2200pf and 3300pf cap and stick them in series, going to ground from the R17/R19 junction.  In series, they yield a capacitance value of just over 1300pf, producing a rolloff around 2150hz.  Use a 3-position SPDT on-off-on toggle to shunt/bypass the 3300pf cap, and the remaining 2200pf cap gets you a 1300hz rolloff.  Flick the switch in the other direction to shunt the 2200pf cap, and the remaining 3300pf cap gets you a rolloff at 860hz.  You can obviously play with values you like, but the principle remains the same: use caps to ground at that point to achieve one of 3 different treble-cut points for the desired degree of octave mellowness.

Again, all of this is intended to reduce the amount of craziness and noise.  Because of how octave dividers work, not all the mistracking and unpredictability will disappear, but some combination of these mods may get you the balance between control/usability and zaniness you desire.

[Arno van der Heijden]

Mark, when is your book going to be released?  

[Mark Hammer]

I do it like Charles Dickens: a chapter a week. :wink:

[vdm]

hey mark, thanks a lot mate, you're always a good source for a plethora of useful information. and ill go through and try a few things you've listed, and search for that mix of useability and zany-ness you speak of...
oh and by the way, do you have any idea how to add more treble to the distorted only output???
thanks,
trent

[Mark Hammer]

Sure.  The cap to ground at the very end of the circuit bleeds treble from both signals.  If you're going to add the treble-bleed caps I suggested to the suboctave signal, then the .01uf "master-mellow" cap at the end is unnecessary.  Lift it or replace it with something like a 2200pf cap and that should give you all the treble you want.  It may be smarter to keep a modest value cap in there for hiss control, rather than eliminating it totally.  In principle, dropping it to 2200pf should raise the rolloff by about 2-1/2 octaves which should restore more bite to the fuzz-only.

Let us know how these mods work out, if you do them.  It's helpful to know when "cures" are less curative than expected, and also helpful to know when component valuescan be profitably played with.

[R.G.]

A much more useful thing to do than any amount of fixes on the Blue Box is to buy PAIA's kit for the Roctave Divider. That baby is the best octave down I've ever found.