UVICS - Univibe In a Crybaby Shell

[RedHouse]

Yes but my son's daycare has an age limit.
You might be a little too old  

I'll keep looking then.  

On the RM voodoo vibe, RM used 2.2uF caps ...but... (IIRC) he also changed the associated resistors so as to re-scale (impedance) of things, it comes out the same. Another thing to notice is RM used a .047 phase cap in place of the 470pF normally used in stage-3. His use of the compound transistor front end (ala Octavia) actually works well too, you can see he uses one PNP/NPN in the input section and the output section. Seems everything he did (many things anyway) has that configuration in it somewhere.

[Brymus]

OK, did some tinkering.

I think this works. Tie one end of a 100K to the emitter of Q12. Tie the other end to the base of an MPSA13 or a discrete darlington. Emitter of the darlington goes to ground through a 2.2K resistor. LEDs are stacked in series from the collectors of Q11/Q12 (power supply, that is) to the collector of the added darlington. You can use several LEDs in series until you run out of voltage. Two blues and more greens and reds should be possible.

The emitter resistor for the darlington sets the current swing and brightness swing. It goes to essentially zero, and up to about 5ma with the 2.2K resistor Bigger resistor is lower current. All the LEDs use the same current, so it's easy on the power supply. If you want to use the LEDs as bypass status indicators, short the base of the darlington to ground when the effect is bypassed. That's what the 100K in series with the base is for, to let you short the base to turn them off.

I'd do this rather than a whole new driver. It amounts to a driver circuit, but pre-biased and pre-set up to work for nominal LED currents.

Thanks for this RG
I tried it and of course it worked straight away.
But it wasn't very noticable at min speed.
Dim and not much fluctuation.
I swapped in some pots for the 2 resistors,used a MPSA13.
What I liked best was a 4K7 in place of the 100K and a 1K2 in place of the 2K2.
This is the brightest I can get it and still have enough turn off time for a good strobe effect at all speeds.

But by using such small value resistors am I letting too much current through ?

Oh and I also have the LED connected to Q12's emitter and the 4K7 to the collector,it seemed to work better this way.
But IDK why.

[R.G.]

This is the brightest I can get it and still have enough turn off time for a good strobe effect at all speeds.

Are you wanting it to turn them on and off or to ramp up and down?

[R.G.]

It turns out that if you use an Omron D2MQ-4L-1 basic switch (Mouser 653-D2MQ-4L-1, $2.44) have some RC airplane background, and do a little PCB work, you can integrate the cancel function for a univibe onto a PCB that fits into a Crybaby shell directly.

The switch goes directly onto the component side of the PCB. A music wire arm is bent to a right angle and soldered onto the PCB so it holds the lever down. A brass 2-56 pan head Phillips screw with a bit of music wire soldered co-axially into its tip goes into a loop in the end of the music wire holding the level down, and then vertically through the PCB, through a hole in the case drilled to match the PCB hole, and thence to the underside of the rocker. Using the 2-56 screw allows you to put nuts and lockwashers on the threaded part and adjust the trip point.

Rocking heel down pushes the rod, relieves the pressure on the lever and the switch makes on its NC contact, which is arranged to ground the cancel line. There are many, many approaches to this. Obvious ones are reed switches and Hall effect sensors with a magnet on the rocker; optoisolators interrupted by a tab attached to the rocker; one I've used several times before, using a dual pot when it's not needed, and just sensing DC on the pot wiper to activate something purely electronically. I liked the pushrod approach because it allows the switch to be PCB mounted, and have minimal changes and mechanical work on the shell; only one hole to drill, and that can be done with the PCB in place using the hole the wire goes through as a guide for drilling. The hole can be quite small - 0.065 to 0.080 should be fine.  You'll want to sand the end of the music wire to a nice smooth round end and possibly dip it in some epoxy for the contact end. Any of you who have done model airplanes will have no trouble with this at all.

Another that is probably a little hard to do right from a mechanical standpoint is to put a (switch/opto sensor/capacitive sensor) under the rubber pad on the rocker and have the thing un-cancel or un-bypass when there is a foot on the pedal. No magic - just application of standard technology to an ordinary problem.

And while we're at it, it's simple to add either CMOS bypass switching ( http://geofex.com/Article_Folders/cd4053/cd4053.htm from 2000, and old at the time ) or relay bypassing using a submini signal relay like the Omron G6 series (~ $3.00, Mouser). Using the full length of the available PCB space in a shell (http://geofex.com/FX_images/Crybaby Shell Measurements.pdf) leaves plenty of room to mount any of this, and still leave room for a 9V to 17V upverter with the LT1054.

It's also an obvious bit of switching to make the thing do either cancel or bypass once you've put the switching into the box. As usual, the problem is where to put all those controls. A programming jumper on the PCB seems like the obvious way to do it.

[Brymus]

This is the brightest I can get it and still have enough turn off time for a good strobe effect at all speeds.

Are you wanting it to turn them on and off or to ramp up and down?

Turn on and off,its more noticable that way.
I really like it the way I have it now.

[R.G.]

Cool. Whatever works.

Are you interested in trying out the single-pot control with a LED/LDR?

[Brymus]

Cool. Whatever works.

Are you interested in trying out the single-pot control with a LED/LDR?

I could give it a try,I like the way the sweep is on mine finally.
So I can always go back to it if I want.

What value LDR for the two LDR's ?
And what color LED or does it matter ?

[R.G.]

What I had in mind was sending you a VTL5C3/2 module.

I was interested in getting an end-user test on the circuit before I post the semi-final layouts.

[Brymus]

Sure that will work too 

I added the LEDs and baby board,but I made the mistake of using a different pair of LEDs.
They blink slightly different ,no matter, it does what I want it too.
Just a nice blue glow emminating from under the treadel in sync to the LFO.

[R.G.]

Help me. I can't stop. 

Just put up Son Of UVICS at GEO.

I wondered if I could get more stuff on the PCB. Just crossword puzzle stuff. So I converted most of the resistors to stand-up style and changed all those 1uF film caps back to 1uF electrolytic like the original Univibes used. Doing that, I got onto the PCB:
- stock univibe circuit
- three phase capacitors in each position, with room for jumper programming pins; this can also be taken off-board to a 4P3T selector switch if you like, or just left stock
- mixer/trim phase depth trimmer
- lamp driver trimmer pots
- single speed pot by use of LED/LDR module on the main PCB; for a blast from the past, see: http://www.geofex.com/FX_images/tubev1.jpg That little blob on the lower right of the PCB is a VTL5C3/2
- 9V upverter with LT1054 on the main PCB (no baby board)
- standard/improved AC input power supply
- DC input for 15-20Vdc inputs
- two methods of installing an external LED speed indicator
- LED kill for bypass indication on both external LED circuits
- all the usual resistor value diddling, darlington/non-darlington variations, etc.
- C4 compensation standard or feedback
The variations are done by selective population of the PCB, leaving off what you don't want to use. For instance, if you use 9V DC in and the upverter, you don't need the rectifier bridge, so its space is "reused" by a capacitor used by the LT1054. If you use AC input and the full wave bridge, you don't need the capacitors and  diodes that go with the LT1054, so you leave them off.

I thought about putting an electronic or relay bypass on the PCB, extending it down between the jacks and such, but decided to try to hold it down before I went for an SMD layout.   

This is a practical demonstration of more flexibility = more complexity. The population diagrams run to several pages (that haven't been written yet  )

So - what's the most useful version of this for the DIY community? Stock, allowing resistor swapping? Or full-blown? Or fuller-blown? How complex is enough? 

[Brymus]

Awesome !!!!
Dont stop!!

I personally,would rather see the upverter on a baby board, to leave more room for mods or extra stages.
But I also see the value of being able to daisy chain rather than have a dedicated wart for the vibe.
So I am sure alot of people will like that option.

I have a question about the regenerative feedback.
Is it possible to use more than one cap and resistor combo,to get feedback at different frequencies?
Or do the larger caps make the smaller ones redundant when using more than one ?

When I was testing it a couple of nights ago I noticed it had all the neighbors dogs howling (with certain cap values) even when it wasnt oscillating at a feq audible to me.
It was kinda funny.

[RedHouse]

Keep on rockin RG!

[R.G.]

I intend to.

[R.G.]

I personally,would rather see the upverter on a baby board, to leave more room for mods or extra stages.
But I also see the value of being able to daisy chain rather than have a dedicated wart for the vibe.
So I am sure alot of people will like that option.

Since I already have a chain of layouts with more stuff shoved onto them, Maybe I just put up the whole mess and let people pick the level they like.

I have a question about the regenerative feedback.
Is it possible to use more than one cap and resistor combo,to get feedback at different frequencies?
Or do the larger caps make the smaller ones redundant when using more than one ?

Not sure what you mean. Do you want to use multiple regen paths? to different places in the circuit? different RCs to select the regen peaks? something like that?

Positive feedback works on the same concepts as negative; feed back signal, and at frequencies where it reinforces, there is a peak. Where it opposes, there is a valley. If you feed back the output of a phaser to any point along the phase line, whatever frequencies reinforce make peaks. Because the phase line has a varying  - sometimes wildly! - phase response, the frequencies which are reinforced are wobbled around by the LFO just like the the notches are.

When I was testing it a couple of nights ago I noticed it had all the neighbors dogs howling (with certain cap values) even when it wasnt oscillating at a feq audible to me. It was kinda funny.

Our dogs and cats have me all figured out. When I pick up a guitar, they run and hide.      

[RedHouse]

Our dogs and cats have me all figured out. When I pick up a guitar, they run and hide.    

Are ya sure it's the HF oscillations they're running from?

On a more serious note, just an FIY on the up-vertor thing, y'all should know it can have issues supplying a vibe.

Yeah, I know the spec sheet says otherwise but in my experiments (and others) was found it can add some audible funk (noise) to the power suppply when it's at it's peak capacity (100mA). I use the standard T1-3/4 40mA bulb and that might have been an issue, I bet if one uses the Radio Shack 272-1139 it would be fine (see below).

The workarounds seem to be, first Bob Sweet (mojovibe) solved his problem by supplying Q10 (UVICS Q10/Q14) at the 9V level, this gained just enough elbow room for the up-convertor, while Mike Fuller (mini deja vibe, also using the LT1054) and he's using a 20mA "grain of wheat" bulb seems to squeeze by.

it's a heads-up to the builders on using the up-convertor schema in a vibe, not always a slam-dunk.

[R.G.]

Our dogs and cats have me all figured out. When I pick up a guitar, they run and hide.    

Are ya sure it's the HF oscillations they're running from?

I'm quite certain it's not. That was the point. 

On a more serious note, just an FIY on the up-vertor thing, y'all should know it can have issues supplying a vibe.

Yeah, I know the spec sheet says otherwise but in my experiments (and others) was found it can add some audible funk (noise) to the power suppply when it's at it's peak capacity (100mA). I use the standard T1-3/4 40mA bulb and that might have been an issue, I bet if one uses the Radio Shack 272-1139 it would be fine (see below).

The workarounds seem to be, first Bob Sweet (mojovibe) solved his problem by supplying Q10 (UVICS Q10/Q14) at the 9V level, this gained just enough elbow room for the up-convertor, while Mike Fuller (mini deja vibe, also using the LT1054) and he's using a 20mA "grain of wheat" bulb seems to squeeze by.

it's a heads-up to the builders on using the up-convertor schema in a vibe, not always a slam-dunk.

As with almost everything in nature, it's possible to get technical matters wrong. Nothing is a slam-dunk unless you know what you're doing, and maybe not then.

Everyone who has ever used a switching power supply for audio knows that issues of minimum load, switching noise and ripple, ground currents, overload, transient response, and bleedthrough are things which cannot be ignored. In most cases, these are layout issues more than circuit issues, and especially power and ground routing, loop area, and decoupling. As such, they remain mysteries to people who have not yet "graduated" to the more subtle issues of how layout affects performance. It's not just supplying a Univibe derivative, it's all audio circuits you have to be careful with. 

My early experiments with charge pump converters on a univibe clone were not successful, and so I just didn't publish them and didn't recommend them. It was so long ago that I just forgot about it. Then someone tried it with the LT1054, and apparently the higher current capability did the trick, at least with an external converter board.  I consider it an experimental issue until verified, because of the issues that any switching power supply can bring up. Accordingly, the converter circuit is an option, not a necessity, and it can be left out of the latest version of the board by leaving off the parts and reconnecting the stock parts. It's a good possibility for people who want to experiment, certainly until the layout is proven to sidestep the known issues with switching power supplies.

[RedHouse]

It's a good possibility for people who want to experiment

Yeah baby!  a lot of people ask about it on the forums.

[Brymus]

Not sure what you mean. Do you want to use multiple regen paths? to different places in the circuit? different RCs to select the regen peaks? something like that?

Positive feedback works on the same concepts as negative; feed back signal, and at frequencies where it reinforces, there is a peak. Where it opposes, there is a valley. If you feed back the output of a phaser to any point along the phase line, whatever frequencies reinforce make peaks. Because the phase line has a varying  - sometimes wildly! - phase response, the frequencies which are reinforced are wobbled around by the LFO just like the the notches are.

OK
I was trying to get more depth or phase.
I tried again yesterday after reading your post and while I didnt get the results I wanted they were interesting.
Using a 500K pot in series with a 1N - 220N cap from output to ?
I found that C12 was the (best) spot,for lack of better word.
With (I forget, should have taken notes ) one value going to the C12 R27 junction I could get the vibe effect canceled on the high E,B,G strings and the effect was more pronounced on the low E,A,D strings.
Another cap and the vibe effect could be canceled on all but the low E and it would seem doubled rate (vally's and peaks emphasized?) and more pronounced.
A very cool effect,just not sure how to use it.
Going to the other side of C12 just meant the resistance had to be oposite for the same effect,interesting.

Other spots it just seem to cancel out the effect or oscillate.

I read a post by JCM saying for maximum depth to change one of the caps to 470uf but I didnt understand which one he meant.
I think he meant C22 (on UVICS)
Does that sound right ? ( I would like to know before pulling the PCB yet again)
I would like to get more depth at times without losing the abilty to sound the way it does now.

I did use my UVICS again at last practice,after we finished our Zeppelin we did "Breath" and "Bridge of Sighs" very fun.

I also used it instead of my Ibanez PM7 on "The Rover",I really like it for that.
The PM7 is very noticable,the UVIVS way more subtle and I can sweep the treadel to get more phasey on the breaks.
Also on "Kashmir" during the bridge and outro I used it to help simulate the keyboard melodies.

[R.G.]

I was trying to get more depth or phase.
I tried again yesterday after reading your post and while I didnt get the results I wanted they were interesting.
Using a 500K pot in series with a 1N - 220N cap from output to ?
I found that C12 was the (best) spot,for lack of better word.
With (I forget, should have taken notes ) one value going to the C12 R27 junction I could get the vibe effect canceled on the high E,B,G strings and the effect was more pronounced on the low E,A,D strings.
Another cap and the vibe effect could be canceled on all but the low E and it would seem doubled rate (vally's and peaks emphasized?) and more pronounced.
A very cool effect,just not sure how to use it.
Going to the other side of C12 just meant the resistance had to be oposite for the same effect,interesting.

Other spots it just seem to cancel out the effect or oscillate.

I read a post by JCM saying for maximum depth to change one of the caps to 470uf but I didnt understand which one he meant.
I think he meant C22 (on UVICS)
Does that sound right ? ( I would like to know before pulling the PCB yet again)
I would like to get more depth at times without losing the abilty to sound the way it does now.

I see. I think you need... Mark Hammer.

Mark, are you listening? Didn't you do some detailed looking at the differences in where you applied feedback in a phase line at one time? Wasn't that on a univibe?

I did use my UVICS again at last practice,after we finished our Zeppelin we did "Breath" and "Bridge of Sighs" very fun.

I also used it instead of my Ibanez PM7 on "The Rover",I really like it for that.
The PM7 is very noticable,the UVIVS way more subtle and I can sweep the treadel to get more phasey on the breaks.
Also on "Kashmir" during the bridge and outro I used it to help simulate the keyboard melodies.

Sweet!

[Brymus]

OK the vactrol arrived yesterday afternoon,Thanx RG
So hopefully this week I will switch over to that and a wah pot.
Once thats through I will post a build report thread with,pics,maybe a vid.

I havent had a need to mess with it lately,it works great and I like it alot.

We did do floyd's Breath at our last show,but while the music was good the singer didnt practice that song so I wont be posting the vid.
But yeah it worked great ,made the sound fit the song.