Vibrato

[Mark Hammer]

This scanned article from Polyphony shows how to mod a fairly "normal" LFO to produce alternate waveforms: http://hammer.ampage.org/files/hypertriangleclock.gif   The part to pay special attention to is the segment with the 2N3819 JFET and the diode from gate to ground.  The rounding off of what starts out as a triangle waveform can be accomplished at either or both peaks, depending on whether the diode links gate and drain, gate and source, or both.  I gather the extent of rounding can be tinkered with by means of diode choices, biasing, etc.

[gez]

It's worth pointing out that any 'shaping' circuits are best set up using a scope and best run from a regulator/regulated supply so that results are consistent (battery flattening will alter wave form).

[byoc]

ANY phaser, chorus, or flanger can be teased into providing vibrato, simply by lifting the dry signal from the mixing stage.  Unfortunately, for the vast majority of commercial pedals, the way that electronic switching is accomplished the "effect mode" consists of turning a single FET on to connect the wet signal to the mixing stage, and "bypass" simply turns that same FET off so you are only left with dry signal.  In such circumstances, installing a "dry-lift" switch results in no signal whatsoever when you go from effect to bypass, using the stock system.  Of course, if you build your own pedal and use a mechanical (rather than FET) switch, you can still use a dry-lift to achieve vibrato since the DPDT or 3PDT will sidestep the entire circuit.

RG is correct in noting that the VB-2 is likely the only commercial pedal (at least widely distributed one....there are ALWAYS regional or small-production manufacturers that 95% of us here have never heard of because of limited product distribution) targetting vibrato alone.  What makes it a targetted product, rather than merely a chorus with a dry lift is 2 central features: a) bypass circuitry tailored to retaining only wet signal when in effect mode and subbing dry signal when in bypass, and b) incorporating a ramp-up feature that increases vibrato intensity over time to simulate the way musicians produce both finger vibrato or vocal vibrato as the note is held.  I personally do not know of any other pedal that has this, and it is one of the strongest reasons to build the BYOC VB-2.

Just as an aside, we have never, to the best of my knowledge, discussed adapting the VB-2 ramp-up feature within the context of a flanger or chorus here.  That could be either by a) adapting the stock VB-2 to be able to achieve ramped-up chorus effects by mixing some dry signal in effect mode (as noted, normally this is absent), or b) adding the ramp-up feature to another circuit.

One of the more interesting vibrato effects out there is produced by means of using ONLY the wet signal from a phase-filter adapted 4-stage phaser circuit using OTAs.  This could be either a Small Stone or "black" Ross Phaser or a clone of either.  The DOD FX20 also has potential in this regard.

Normally, cancelling the dry signal in any 4-stage phaser, regardless of how the phase shift is produced (whether OTA-based, FET-based, CMOS-switched, or LDR-based) produces pitch warble or vibrato.  Generally speaking, the more stages of phase shift, the wider the warble, so a 4-stager will produce more robust vibrato, the same way it will produce more intense phaser effects.  In the phasefilter modded version, 2 of the phase-shift stages are converted to lowpass filter, while two remain in phaseshift/allpass mode.  This is VERY easily done with OTA based phasers, since it simply involves redirecting one end of a cap in each stage from the input to ground.  A single DPDT toggle and you're in business.

When the two stages are switched from phaser/allpass mode to lowpass, and the dry signal is lifted, you get a cyclical swept treble cut, along with a cyclical pitch warble.  Although the pair of lowpass stages do not specifically cut the volume, because there is this swept treble cut, the guitar signal appears to lose and gain volume as a psychoacoustic effect of the change.  The resulting sound is this deliciously swampy mélange of vibrato, tremolo, and autowah that can sound ridiculously sexy at certain speeds.

Given that, my own recommendation for vibrato is to build or mod a 4-stage phaser into a souped-up modulation machine that can do phasing, 4-stage vibrato, or phasefilter (w/ and w/out dry signal mixed in).  Charlie Barth has his "Frankenstone" pedal to illustrate over at the Moosapotamus site, although I don't think the phasefilter samples do justice to that particular effect.

All of that being said, the VB-2 clone has much to commend it also.

At first glance, the VB-2 schematic may seem like a relatively complicated circuit to pick apart and implement into "original" designs.  But if you study it closely, you'll see that 50% of the circuit is nothing but buffer.  Once you block out all the unnecessary circuitry it becomes easier to understand what's going on.  Personally, I would give the buffer section of the VB2 a bit of an amputation for the BYOC PCB because it creates noise, but people want it to be as close to the original as possible.

The first thing you need to know is that the BA662A VCA chip is extremely rare and expensive.  "Pulled" BA662's sell on ebay for $45 - $70....mainly to Roland TB-808 bass line cloners.  The BA662a is a Roland proprietary chip and was designed to replace the CA3080 in many of Roland's synth and drum modules.  Who ever designed the VB-2 thought the BA662a would be ideal for to use in the ramp feature.  Fortunately, Rohm, the foundry that produced the BA662a for Roland decided to continue producing their own version of the design after Roland pulled the plug on production.  The Rohm version is the BA6110.  www.smallbearelec.com now carries these for about $2 each.  Unfortunately, the BA6110 is not exactly "equivalant" to the BA662 and it is also not pin for pin.  The main short coming of the BA6110 is that it needs a stronger input signal than the BA662.  This can be overcome by adding an extra transistor array to ramp up the input signal.

So once we get over the obstacle of sourcing a VCA for the ramp up feature and we look at the schematic again, we can see that the VCA is just a "gate" that controls the amplitude of the LFO to the MN3102.  I have read posts from many people saying that they have cloned the VB-2 without the VCA by simply omitting it and running the LFO from the TL022 straight to the MN3102.  They are either lying or had some "happy accident" because the signal from the TL022 coming off the output pin, as is typical, is not strong enough to drive the MN3102 as is, nor is it the ultimate output of the LFO.  None the less, the thing that makes the VB-2's pitch vibrato unique from any other chorus that has had the dry signal removed is it's pure sine wave twin T LFO.

Anyways...I guess Mark's post inspired me to write a long winded reply that could be summed up in a few sentences.  The bottom line is this:  All you need to get the VB-2 sound (aside from the ramp feature) is a sine wave LFO and no dry signal mixed in.  You could probably modify many less complex chorus circuits to do this.  Also, IMHO, you could not achieve anything that sounds remotely like the VB-2 with OTA or opto based circuitry....only BBD based.

Of course, if you aren't worried about copping the VB-2 sound, then none of this matters.

[Mark Hammer]

Thanks for the reply.  Yep, that 662 is a very rare bird.  If your uncle had a drawer full of some of those, some SSM2040s, and some MN3010s, you could retire right now!

Though we should be clear to emphasize the difference between tremolo and vibrato, I should point out that the ramped-up modulation feature has been nicely implemented by RG Keen in the context of the EA Tremolo.

I suppose the other thing that was not mentioned by BYOC was the feature on the original VB-2 (and I assume implemented on the clone?) that allowed for either "free-ranging" vibrato without the ramp-up, versus unlatched momentary/selective swell/ramp.  The VB-2 was one of those Boss pedals where, depending on how long you hold down the foot pedal, the same switch would do one of two things.  The DF-2 works the same way, and I imagine there have been other pedals like that throughout the past 25 years.

[byoc]

Thanks for the reply.  Yep, that 662 is a very rare bird.  If your uncle had a drawer full of some of those, some SSM2040s, and some MN3010s, you could retire right now!

Though we should be clear to emphasize the difference between tremolo and vibrato, I should point out that the ramped-up modulation feature has been nicely implemented by RG Keen in the context of the EA Tremolo.

I suppose the other thing that was not mentioned by BYOC was the feature on the original VB-2 (and I assume implemented on the clone?) that allowed for either "free-ranging" vibrato without the ramp-up, versus unlatched momentary/selective swell/ramp.  The VB-2 was one of those Boss pedals where, depending on how long you hold down the foot pedal, the same switch would do one of two things.  The DF-2 works the same way, and I imagine there have been other pedals like that throughout the past 25 years.

Sometimes I think it would be more prudent to invest in NOS components rather than stocks or mutual funds....I'm dead serious.

Another thing about the VB-2 that is total different from all BOSS pedals from that era (except for maybe the DF-2.  I am not familiar with that one) and earlier is the bypass.  Instead of the standard two transistor flipflop, it uses a single chip - the BA324.  It's a SIP5...pretty odd.  I suppose they had to use something different to get the bypass function and the ramp function from the same footswitch. 

It would be interesting to use the EA trem LFO with a BBD and RG's ramp mod to make a vibrato that is comparable to the VB-2.  Give it a much smaller buffer with just a dual op amp and I think you'd have a vibrato project that would be more accessible to the general DIY community that likes to source their own parts and make their own PCBs.

[Rodgre]

I know I'm not answering the initial question, but I wanted the world to know that the VB-2 isn't the only commercial vibrato out there. I think the Electro-Harmonix Wiggler (and to a lesser extent, the Worm) is a great dedicated vibrato pedal.



Roger

[Mark Hammer]

Sometimes I think it would be more prudent to invest in NOS components rather than stocks or mutual funds....I'm dead serious.

Another thing about the VB-2 that is total different from all BOSS pedals from that era (except for maybe the DF-2.  I am not familiar with that one) and earlier is the bypass.  Instead of the standard two transistor flipflop, it uses a single chip - the BA324.  It's a SIP5...pretty odd.  I suppose they had to use something different to get the bypass function and the ramp function from the same footswitch. 

It would be interesting to use the EA trem LFO with a BBD and RG's ramp mod to make a vibrato that is comparable to the VB-2.  Give it a much smaller buffer with just a dual op amp and I think you'd have a vibrato project that would be more accessible to the general DIY community that likes to source their own parts and make their own PCBs.

Doesn't even have to be a BBD.  Could also be a cluster of allpass stages, in either a FET-based or OTA-based phase shifter. 

If I'm not mistaken (and I'm rushing here, eating lunch before I have tor un back to an all-day meeting in 2 minutes), RG's circuit is envelope driven, where the VB-2's ramping circuit is user-actuated via a footswitch, such that vibrato can be ramped up independent of actual playing level or pick intensity.  In some respects, they can achieve the same outcome, but in other respects a different way of incorporating modulation into one's playing.

[Nasse]

I have cheesy old casio synth that has a vibrato "delay" parameter, the note starts to "vibrato" only after a short time, makes solo violin more realistic or bit less mechanic. I wonder if any vibrato fx for guitar has ever done this feature...

I´m not sure if it really was a real fx but I remember me talked about vibrato fx for vocals with my dance musician friends, it was years before digital fx, someone claimed that certain vocal p.a. has built in vibrato fx...

[puretube]

Of course, there is another Pure-Sine-Apparatus out there,
that is capable of both:
BBD-vibrato,
BBD-vibrato,
OTA-vibrato,
and combinations thereof:



of course it takes a quadrature oscillator to get there...


oh, yes: it does the so-called: "un*v*bed" - vibrato, too!


Mark: the "hyperboliced" triangle isn`t that suited for vibrato, IMHO...

Rodgre: thanx! 

Nasse: AFAIK, that is the feature of the BOSS pedal...

BYOC: stocking NOS applies to components, Pedals, & schematix...

[David]

Are there any vibrato kits available apart from the BYOC VB2 clone?

You may want to look at this:

http://members.shaw.ca/roma/vibrato.html

I remember breadboarding it once, but can't recall how it sounded...

It probably sounded pretty quiet -- as in nonfunctional.  I have never gotten any of that guy's stuff to work.

[Mark Hammer]

Mark: the "hyperboliced" triangle isn`t that suited for vibrato, IMHO...

I concur.  It is most useful at slower sweep speeds, particularly those too slow to be perceived as vibrato or "bubbly".  I think this is true of both allpass-produced vibrato, and pure time-based vibrato.

This is also why the original (six CA3094s) Small Stone sounds great at fast speeds; the 50uf smoothing cap to ground - missing in the Issue J - "re-sines" the hyperbolic waveform from the LFO when it starts to get up into faster speeds.  That's also why I recommend sticking a 10-22uf cap to ground on the comparable point on the Ross Phaser.  I find the fastest speeds much more listenable when the sharp, spiky "turnaround" of the hyperbolic waveform is rounded off.

Of course, at long slow sweep speeds, the hyperbolic sweep comes into its own and DOES improve the feel and dramatic quality of the sweep.  The nice thing about that smoothing cap is that it exerts less rounding as speed slows down.  A sort of automatic waveform-vs-speed adjustment.

[brett]

Hi

It probably sounded pretty quiet -- as in nonfunctional.  I have never gotten any of that guy's stuff to work.

But there's a standard Big Muff and a Fuzzface.  Ok, so FFs often don't work
But the principles of those circuits are simple enough, even if the implementation is rough (the 555 tremolo seems like a bad idea).
cheers

[dxm1]

It probably sounded pretty quiet -- as in nonfunctional.  I have never gotten any of that guy's stuff to work.

Hmm... That's possible, I guess, though I doubt that I'd have kept the link if it
were pure crap. It's true that I never bothered to move it to perf or pcb.

Guess I'll have to breadboard it again...

[Processaurus]

Sometimes I think it would be more prudent to invest in NOS components rather than stocks or mutual funds....I'm dead serious.

Another thing about the VB-2 that is total different from all BOSS pedals from that era (except for maybe the DF-2.  I am not familiar with that one) and earlier is the bypass.  Instead of the standard two transistor flipflop, it uses a single chip - the BA324.  It's a SIP5...pretty odd.  I suppose they had to use something different to get the bypass function and the ramp function from the same footswitch. 

It would be interesting to use the EA trem LFO with a BBD and RG's ramp mod to make a vibrato that is comparable to the VB-2.  Give it a much smaller buffer with just a dual op amp and I think you'd have a vibrato project that would be more accessible to the general DIY community that likes to source their own parts and make their own PCBs.

If I'm not mistaken (and I'm rushing here, eating lunch before I have tor un back to an all-day meeting in 2 minutes), RG's circuit is envelope driven, where the VB-2's ramping circuit is user-actuated via a footswitch, such that vibrato can be ramped up independent of actual playing level or pick intensity.  In some respects, they can achieve the same outcome, but in other respects a different way of incorporating modulation into one's playing.

Thats a coincidence day before yesterday I tried a vibrato with my small clone board (I got from BYOC, actually) with an EA trem's LFO feeding the top of the depth pot, sounded pretty good.  There is a little anomaly in that lfo at the bottom of the peak that is an abrupt change in direction though, you can see it on a scope.  A RC filter can mellow it out, I've done that on an EA.  For this use an RC filter would be especially nice, because it would reduce the waveform as it gets faster, so as you sped up the rate you wouldn't need to turn down the depth to get the same amount of pitch bending. 

Still its speed is limited with that lfo.  Also RGs vibra-matic mod of shorting out the LFO didn't quite work as well as i hoped when I tried it, it seems to delay the wave form, and then it starts quickly, rather than a steady ramp.  That was using it for the trem, though, I haven't tried it driving a BBD vibrato.

[Uma Floresta]

ANY phaser, chorus, or flanger can be teased into providing vibrato, simply by lifting the dry signal from the mixing stage.  Unfortunately, for the vast majority of commercial pedals, the way that electronic switching is accomplished the "effect mode" consists of turning a single FET on to connect the wet signal to the mixing stage, and "bypass" simply turns that same FET off so you are only left with dry signal.  In such circumstances, installing a "dry-lift" switch results in no signal whatsoever when you go from effect to bypass, using the stock system.  Of course, if you build your own pedal and use a mechanical (rather than FET) switch, you can still use a dry-lift to achieve vibrato since the DPDT or 3PDT will sidestep the entire circuit.

Just wondering, how would this be accomplished on a Deluxe Electric Mistress?

Schematic:

[Mark Hammer]

Probably the simplest way would be to break the connection between C18 and R40.  I was going to suggest grounding the link between C18 and R40 but that would result in too much of the signal going to the BBD being rolled off.  On the other hand, a muted wet signal used for vibrato might be interesting.

[Uma Floresta]

Probably the simplest way would be to break the connection between C18 and R40.  I was going to suggest grounding the link between C18 and R40 but that would result in too much of the signal going to the BBD being rolled off.  On the other hand, a muted wet signal used for vibrato might be interesting.

Thanks kindly, I'll give that a try and post the results. 

[bonkdav]

how hard is it to find the place to break the dry signal. i have a crappy phaser and a crappy flanger that arent worth 15 bucks between them. id really wanna try this. i think theres a lot of surface mount in them so i dont know how easy itll be. its an arion phaser and a danelectro phaser.
sorry if im jacking.

thanks
davis

[Uma Floresta]

Probably the simplest way would be to break the connection between C18 and R40.  I was going to suggest grounding the link between C18 and R40 but that would result in too much of the signal going to the BBD being rolled off.  On the other hand, a muted wet signal used for vibrato might be interesting.

Well, I opened it up and discovered the reissue is quite different from the original DEM. I've labeled the components below. All of the chips have been sanded to remove labels for some reason, except two. As far as I know, the SAD1024a should have 16 legs, right? The one in the reissue only has 14 -  What BBD chip are they using here??

While I was in there, I removed all the cheap ceramic caps and replaced with red film caps - the sound is much clearer now.

I'm having a hard time locating the connection I should break - a lot of the values are much different here, too. Any ideas?



By the way, the green .0033uF cap isn't original - it's part of a modification to make it sound more like the non-Deluxe EM.

For reference, here is the top of a reissue DEM pcb:



And here's the original DEM:

[newperson]

What is the difference in sound between the VB-2 and the CE-1 vibrato?