It works the first time! A great sense of pride, and a big thank you to all.

[sfr]

Finished up the Neovibe today.  It works the first time!  I still want mess with the junction of R35/36 for the best mix, and I think I might fool around with the light shield a bit more, and it needs a switch; but I plugged it in, and there it is, that cool 'Vibe sound! Right off the bat!

I remember when I couldn't even get a booster working on the first try.  Learning to check for errors as I go along and *before* I plug it in has certainly helped, but more than anything, the gracious nature of you fine folks here has helped me to understand what goes on in these mysterious boxes.  And for every stupid question I ask, there's a dozen more I'm now able to figure out myself because of this place. 

I'm a little underslept (12 hour night shift got done 6 hours ago) so maybe this is silly, but I'm just very proud of myself.

And thanks of course to R.G. for making the Neovibe circuit available to us all.

Now I need to find an enclosure - it *almost* looks like it might fit entirely in a crybaby shell (I have a couple of the older ones - are these deeper than the current ones) - anyone accomplished this? 

I'm sure I'll have more stupid questions when it comes to tweaking this build, but for now, thanks.

[R.G.]

If you have one of the new ones from GGG, I think it will fit if you remove the mounting boss that the crybaby board screws to. I think a rotary file bit in a drill or a dremel tool would do it. Seems to me that I remember doing this once.

[sfr]

Hrm . . . anyone know where to get a 200K pot for keying in the optimum blend at R35/36?  I haven't been able to find one.

[R.G.]

Having said that, how useful would it be to have a 'vibe layout intended to go in a crybaby shell and mount on the existing bosses?

This would likely be even a more densely populated board than the one now. DIY home methods like PNP would be tough.

[Barcode80]

I think a 250k would be plenty close enough to not notice much diff. in control.

[R.G.]

250K ought to work fine.

By the way, make that a 250K stereo audio taper. The rack and pinion reverse the effective pot rotation so a reverse audio is not needed.

[sfr]

250K ought to work fine.

By the way, make that a 250K stereo audio taper. The rack and pinion reverse the effective pot rotation so a reverse audio is not needed.

Are we talking two things here?

The 250K we're talking the trimmer for replacing R35/R36 (the resistors before the switch/volume pot)   And talking about going with regular audio taper because of teh rack/pinion reversing pot rotation, we're talking about the dual pot for the speed control?   Or am I missing something?

Rather than try and find a way to kludge a dual gang pot into the crybaby shell, I was planning on rigging up a small daughter board for the single pot setup from Geofex. (  http://geofex.com/circuits/ldrlfo.htm ), but I guess I screwed something up, because breadboarding it to see if I had the values worked out right, I managed to smoke a resistor! (first time I've ever done that!)  Trying to run that off the 18VDC supply before it hits the Neovibe board is probably not the best solution. 

From the above linked page:

With a single 10K pot for speed, the transistor varies the current in the LED from nearly off to a voltage set by the V+ voltage minus the LED 1 voltage and the Vbe of the transistor. You take that voltage, divide  by the max current you want in your LED, and insert that resistance into Re.

When we say "max current you want in your LED", does that mean LED1 in that circuit, or the LED of the Vactrol?  With an 18V DC supply, I'm thinking I need around 47 Ohms for Re, and I got good values of resistance across the Vactrol LDR leads until Re started smoking.  I think I'm putting like 6W across that resistor.  (Current should be about .38 amps if I've got 18V across 47Ohm, right?  So that means about 6.8 W?  I need a higher power resistor lower power supply or I'm totally calculating something wrong.)  Sorry, I'm not too bright.

[Peter Snowberg]

Congratulations!

[R.G.]

The 250K we're talking the trimmer for replacing R35/R36 (the resistors before the switch/volume pot)

... oops... sorry. My bad.

If it's a perfect mix you're after - The two resistors that do the mix could each be off by 5% max, probably less. There could be some gain error in one path versus the other, but probably nothing like a 2:1 difference. I would make each resistor 47K and use a 100K trimmer. That would tweak in even radically different paths. Frankly, if you can hear any phasing with 100K fixed 5% resistors in those positions, it would make a lot of sense to put in 91K resistors and a 20K trimpot. The smaller the range of the trimpot, the easier it is to make fine adjustments.

When we say "max current you want in your LED", does that mean LED1 in that circuit, or the LED of the Vactrol?  With an 18V DC supply, I'm thinking I need around 47 Ohms for Re, and I got good values of resistance across the Vactrol LDR leads until Re started smoking.  I think I'm putting like 6W across that resistor.  (Current should be about .38 amps if I've got 18V across 47Ohm, right?  So that means about 6.8 W?  I need a higher power resistor lower power supply or I'm totally calculating something wrong.)  Sorry, I'm not too bright.

You're bright enough, I'm not a good enough explainer.

The LED we want a max current in is the LED in the Vactrol. LED1 is already limited to the supply voltage minus one LED and one diode drop, divided by the 10K of the speed pot. When the 10K speed pot is turned all the way up, it goes up to the power supply less the drop of LED1. So its emitter is one diode drop lower than that. And the current in Re is (Vsupply - Vled1 - Vbe)/Re.

We want that current to be less than the max current for the LED in the vactrol. I forget the max spec, but let's just say it's 5ma for the purposes of this example. So if Vsupply is 18V, Vled1 = 2V, and Imax = 5ma, then Re = (18V - 2V -0.6V)/0.005 = 1925 ohms, or 1.925K. I'd use a 2.0K.

I do know from past experience that your vactrol is likely to need currents under 5ma for the best range on the univibe circuit. Vactrols should hit 250K resistance on each side of the centertap with something well under 1ma in the LED.

[sfr]

Hrm.  Tried those numbers below, was still getting too much current across the LED legs on the vactrol.  (8mA with the pot turned all of one way)  So it turns out my 18V supply is actually 18.35.  So I punched that in (18.35 - 2V -.6V)/0.005=3150.  Series a couple resistors to get 3.15 for Re.  Used a 10K pot.  Depending on the pot's position, I get .03 to 5.03mA across the LED lugs.   The resistance across the LDR lugs of the VT5C3/2 is about 5K or 6K minimum and about 5Meg maximum.  (interestingly enough, the resistance between the outside legs changes from 12K to 5Meg - I thought these would always be the same fairly high resistance?)  Does this seem right?  I'm not sure exactly what I'm looking for.  I figure since I won't get the full pot rotation in the crybaby shell, I should be able to key in on the resistance range I want, but it occurs to me I'm not even sure *what* I'm looking for for measurements on this thing.  I also wonder if I messed something up in the VT5C3/2 when I managed to smoke that low value resistor the first time I breadboarded this.

The LED we want a max current in is the LED in the Vactrol. LED1 is already limited to the supply voltage minus one LED and one diode drop, divided by the 10K of the speed pot. When the 10K speed pot is turned all the way up, it goes up to the power supply less the drop of LED1. So its emitter is one diode drop lower than that. And the current in Re is (Vsupply - Vled1 - Vbe)/Re.

We want that current to be less than the max current for the LED in the vactrol. I forget the max spec, but let's just say it's 5ma for the purposes of this example. So if Vsupply is 18V, Vled1 = 2V, and Imax = 5ma, then Re = (18V - 2V -0.6V)/0.005 = 1925 ohms, or 1.925K. I'd use a 2.0K.

I do know from past experience that your vactrol is likely to need currents under 5ma for the best range on the univibe circuit. Vactrols should hit 250K resistance on each side of the centertap with something well under 1ma in the LED.

[petemoore]

  RG: {quote] Having said that, how useful would it be to have a 'vibe layout intended to go in a crybaby shell and mount on the existing bosses?
  YES...Useful to be certain, but...
  I have a SS Phaser in a V847 shell w/speed control knob controlled by the treadle. Works good, I had to mess with the pots value, also I put in a 'ratio altering wedge' on the pot/string [done like an Ernie Ball Volume pedal method], to have a more controllable range sweep of speed. This took some diddling, and I would go as far as to say "having a limited range/fine tunable with the treadle", then a range switch to allow getting adjustments closer to the max available fast or slower settings would provide excellent, usable control of speed by foot.
  For extended lead solo's, even moderate speed changes illicit a more exciting/less repetitive sound, ability to change speed on the fly is quite usable for these applications...I went with Pete's 'stupid' method of adding a ziptie around the potshaft, creating a 'pointer' which can easily enough be controlled by foot, also the pointer makes a nice potshaft position indicator.
  An alternate I prefer though is to have the bypass switch 'next to' the treadle, this way the speed can be pre-set, then when the effect is switched on, it doesn't start with a quick-whip-warble [which can be cool too] until you get the toe up a bit...reversing the action so that 'slow' is toe down...yet another option.
  I'm now using switch on the side/ziptie indicator-foot controller. I have the ziptie protruding past the front of the pedal, set up so 1/4 turn allows adjustment between sorta fast to sorta slow, this allows EZ control, and fairly precise speed adjustment by sighting the ziptie pointer. A treadle IS better for more exciting/relaxing, quick speed up/speed down changes during a lead passage.

[sfr]

We want that current to be less than the max current for the LED in the vactrol. I forget the max spec, but let's just say it's 5ma for the purposes of this example.

Just out of curiosity I dug up the Vactrol data sheet.  (Reason I couldn't find it before was that I forgot the "L" in "VTL5C3/2", just as R.G. did in the "Single Pot LFO's..." article.  I guess I'm in good company)  Turns out the max. current for the LED in the Vactrol is 40mA, which is nice because otherwise, I surely would have messed it up in my earlier experiments.    (Data sheet here for those interested.)

I haven't had a chance to rig it up yet, but experimenting, it seems like with a 10K Linear pot, you should be able to get enough rotation out of a Crybaby shell to get a reasonably accurate speed control (relative to the values of the stock 100K dual pot).  I have two used shells, and in one of them has the rubber stop in the heel of the pedal ground a bit lower, for more sweep.  (One pedal gets about 180 degrees of rotation, the other gets a little shy of 270.)  Of course, you'll need to line up the rack and pinion so that at the pedal's max rotation you avoid the MOhm range of resistance out of the LDR.  The resistance values you get out of the LDR seem to follow something akin to a logarithmic taper relatively well, I'll see how the whole thing sounds when I get a chance to box it up.  (I still need to pick up a longer-shaft 10K pot for rigging up to the crybaby shell.) 

Since the minimum resistance I'm getting out of the LDR is around 5K, it seems like I may want to tweak R43 and R42, the resistors in series with the speed pot?  In practice I don't think this is an issue, as I don't think anything that low on the dial is particularly useful. 

R.G. :

As far as a Crybaby shell-fitting 'Vibe layout,  I think that would be cool, and very usefull.  I've had a lot of success with very tight spacing on etching - (out of experimentation, I used some P'n'P and copper clad to make "art" a while back [looks cool with a little coloured light behind it in a box]) and was able to get some very fine detail.  The only thing that gets me is drilling the holes in the very small pads on some layouts.  (The neovibe, and the Geofex layout for the ultraflanger)  but I think  that's a result of my drilling setup (hand-held dremel and bits in desperate need of replacement) than and fault of the board layouts.  If possible, adding the Single pot LFO mod from GEO on board would be handy as well - adding a bit more soldering work to the job seems easier than trying to alter the crybaby shell to hold a dual pot.   ( a 25mm dual doesn't fit in the frame opening, a 16mm needs the bracket modified to hold the smaller size throat, although with enough thread on the shaft, washers sort of work as a jury-rigged solution.)

The other concern I have is those 1000µF caps - depending on how I orient the board, they barely clear in my Crybaby shells, and I was at the music store the other day, and I swear that the Crybaby I saw had a shell that was shallower than the two I have here.  Going with axial caps, however, makes the board even larger, laterally.

[sfr]

It just occured to me - adding a 100K parallel resistor across the lugs of the LDR would limit the maximum resistance of the LDR to the 100K the pot normally shows, wouldn't it?

[sfr]

Well, I've got this all boxed up in a crybaby shell (tight fit, and the taller 1000µF electrolytic caps at C24/25 forced a less-than ideal orientation that makes for some long wiring runs in places) and working.  Enjoying it, but still some tweaking to do.

I swapped in two 47K resistors and the 100K trimmer to get the best mix and dialed it in as best I could.  Even at the best mix setting on the trim, it still doesn't sound as deep and phasey as some of the sound samples I've heard - I guess the next step is to play with the bulb trimmer (even at it's dimmest, I'm not sure if the bulb is as dim as it should be, I'm going to try adding some resistance here, maybe moving up to the next value trimmer) and playing with the alignment of the bulbs/photocells. 

Can any old photocell work, or is there a particular photocell I need for this project?  I ordered the "photocell for univibe/neovibe" from smallbear, but I'm going to need more, I want to try the photocells angled towards the bulb rather than flat against the board under the reflective shield.  Forgot to toss some in with my order from Smallbear a couple of days ago.

I got the single-gang pot vactrol-based speed control thing (from geofex  - http://geofex.com/circuits/ldrlfo.htm ) working, sort of.  Still fooling with parallel resistors to get the taper and the max value in a working range.  It's a tough balancing act between getting something close to the original taper and keeping a usuable range inside the sweep of the pedal.

The big problem so far seems to be that the minimum resistance on the Vactrol is almost 6K, and as such I loose the fastest speeds on this build using the Vactrol-based solution.  I pulled R42 and R43 and replaced them with jumpers.  (It appears that these are series resistors with the dual gang pot that set a minimum resistance value on the speed control?)  But since those where 4k7, with the minimum resistance of the Vactrol, I'm still a bit over the minimum resistance I had with those resistors in and just a pot, and still can't get the fastest speeds with the pedal all the way down.  There's actually a surprising bit of range on the fastest side of the dial that I've lost, compounded with the fact that the last little bit of turn on the pot is reserved for pressing the footswitch.

I originally used the Vactrol-based solution both for reliability and to keep from having to do too much modifiying to my crybaby shell.  Since I had to tweak the shell anyway to fit the pot for the Vactrol setup in there, I think I may just toss a  dual 100KA pot in there and be done with it.   But I have to ask - is there another way to get those faster speeds back that I'm just not seeing? I went to trouble of making this Vactrol-based thing work, I'd like to use it!  At this point I imagine I'd have to tweak/change something around Q11/Q12?