univibe oscillator substitution

[David Deen]

I'm sure this subject has been discussed in length before but I would like to go ahead and ask anyway. I've built a univibe clone once in the past but did not like how the speed control swing was so small compared to how much it affected the oscillator speed. I understand this is a relatively common problem.? I'm building my second clone now using the neovibe layout but alas, have run aground with the same haunting problem.

My question then is does anyone know of another good oscillator whose speed can be controlled by a pot where the pot has a good physical swing w/ respect to the range of allowed speed. Is there a good substitute for the variable speed LFO in the old style univibe?

What about the oscillators used in the Danelectro Chicken Salad, the Dunlop Rotovibe, or any conventional oscillators like the wien-bridge, another LF phase shift oscillator, or maybe a 555 timer implementation?

Any suggestions would be greatly appreciated.
Thanks.
-David

[Eb7+9]

Because the feeds to the bulb driver are AC coupled you can theoretically stick any kind of LFO waveform generator behind the cap and have it work good at medium to high speeds ... the ramp/comparator circuit of the MXR phase45 would do great ... but the transfer is limited at very low LFO speeds by the size of the coupling caps - and there's a limit on how big you can make those caps or their charging/discharging can easily suck enough base current from the bulb driver and shut it off during the cycle ...

Instead, you could directly DC-couple say the op-amp  LFO circuit and intensity control of the mxr45 to the bulb driver but need to have the reference voltage of that circuitry match the idling voltage of the bulb driver circuit ... in fact, the same would apply to any generalized op-amp waveform generator (S/H perhaps ?!) provided the common-mode (average) reference can be adjusted to match the idling requirements of the bulb driver circuit ...

Something to try for glide/track/hold enthusiasts ...

~jc

[RedHouse]

I don't have links at hand, but there are a couple good alternatives going  around the internet, one is on Roger Mayer's Voodoo-Vibe which uses the ICL8038 and the other uses the XR2206.

I've tinkered with both in the past and liked the XR2206, built a signal generator out of one I use with my scope.

I've been wanting to try the XR2206 with a univibe circuit myself lately but just haven't had the time.

[RedHouse]

Hey JC,

Another question about your LFO mod if you don't mind.

On the original schem there is a resistor which is absent in your mod schem, is it omitted because it's unnessessary?
(or is it a typo).

The 47k resistor right after the coupling cap, before the junction of the 100k/47k/Base:


isn't on this schem:

[RedHouse]

Bump...

[Dan N]

You know the univibe used an audio taper that went reverse due to the gearing in the controller. Doesn't smallbear have a dual reverse audio? Might give you better control of speed.

That said, a new lfo choice would be lovely! Carry on!

[Eb7+9]

the 47k resistor in the original circuit is inserted to prevent the fairly low input impedance of the bulb driver from loading the output of the oscillator through the coupling caps ... the Darlington arrangement effectively provides an even higher Zin so the original 47k is redundant in that case ...

to get a normal sweep response on the vibe LFO you need a reverse-log pot - Small-Bear started carrying those in dual-100k values a year or two back, I'm just about to try them out myself ... another reported way, if I recall correctly, is to use dual-250k LIN with a 330k across each side to simulate rough reverse-log - this gives you an extra 25~30k of resistance which will extend the slow range with another simple mod to the LFO ...

personally I like the response of the vibe LFO once it's properly tweaked - but so many have asked me over the years how to replace the whole thing with a constant amplitude circuit ... btw, someone a while back posted the link to a 1971 TI data sheet that had bulbs inside of multiple op-amps, but I didn't save it - and when I asked about it nobody replied ... if someone does remember seeing it maybe they can reply here and link again - it did look promising ...

I went on a search a while back after simulating the Boss VB-2 LFO and finding a piece-wise abnormality in its cycle ... and more links if you're interested looking into this :

http://www.discovercircuits.com/O/o-sine.htm
http://www.edn.com/archives/1997/061997/13di_01.htm
http://www.interq.or.jp/japan/se-inoue/e_ckt18.htm

this next EDN article by Michael Fisch is the one that appeals the most to me so far, when I get a chance I want to try adapting it for single +9v and +24v operation ... should be just a case of choosing appropriate diode voltages ...

http://www.edn.com/contents/images/61302di.pdf

jc

[RedHouse]

Thanks JC.

I am integrating an XR2206 into a board which originally started out based on your design (fit into a Wah case) a couple years ago. I imported a graphic (I believe from Moosapotamus old site) into Photoshop, spruced it up a bit, moved some parts around a bit, made it a normal PCB with the parts mounted through, and added your darlington-mod to it.

I recently resurected the project as a friend  and I were speaking of raising the bar on a univibe clone (better parts, LFO etc) I needed to re-do the PBC to be able to use those small "stacked" 1uF polyester caps which meant most traces needed to be moved around to make physical room for the caps. At this stage the board only faintly resembles your (and Moos's) original work (the LDR/Bulb is in the same place).

I decided it was a good time as any to take a shot at replacing the LFO section with one based on the XR2206 chip and use your darlington bulb driver-mod and hence the question if that resistor is needed. I'm nearly done and have yet to etch one and try it out.